System

Abstract

A system is provided.SOLUTION: A system comprising: means for generating a variety of personas using a generative AI model; means for conducting questionnaires on products in the generated personas; means for aggregating questionnaire results and generating statistical information; means for analyzing improvement points of products based on the generated statistical information; and means for displaying analytical results on user terminals.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 market research methods have the problem of making it difficult to quickly and economically gather diverse user perspectives. Furthermore, traditional methods require a lot of time and cost, slowing down product development. Furthermore, opinions tend to be biased toward specific user segments, making it difficult to incorporate new perspectives and diverse opinions. This makes it difficult to develop products that truly meet market demands. [Means for solving the problem]

[0005] The present invention provides a system for generating diverse personas using a generative AI model. This system conducts product-related surveys with the generated personas, collates the results, and generates statistical information. It also includes a means for analyzing areas for improvement of the product based on the generated statistical information and displaying the analysis results on a user terminal. Furthermore, by providing a means for the user to design survey items and generate a survey format, and a means for generating responses from the generated personas using natural language processing technology, it is possible to economically collect diverse perspectives and develop products that quickly and effectively meet market demands.

[0006] A "generative AI model" is an artificial intelligence model that generates data or information that meets specific conditions based on given parameters.

[0007] A "persona" is a virtual character that represents a specific user group and has attributes such as age, gender, occupation, and hobbies.

[0008] A "survey" is a set of questions designed to gather opinions or opinions on a particular subject.

[0009] "Statistical information" is numerical information obtained by analyzing collected data, and indicates overall trends and characteristics.

[0010] "Natural language processing technology" is a general term for technology that processes human language using computers to understand, generate, translate, etc.

[0011] A "user terminal" is an electronic device that allows a user to input information and receive output results.

[0012] "Survey format" refers to the layout and format of questions when conducting a survey, and includes multiple choice, free description, and other formats.

[0013] "Aggregation" is the process of consolidating multiple pieces of data into one.

[0014] "Analysis" is the process of examining collected data and information in detail and finding meaning and trends within it. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results. The program processing of this system is explained below in natural language.

[0037] Preparing for Persona Generation

[0038] Data collection and classification

[0039] server

[0040] The server collects user data from internal databases and external market research data sources. The collected data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, attribute information such as "female in her 20s, student, hobby is fashion" may be collected.

[0041] Data Preprocessing

[0042] server

[0043] The collected data is cleaned to remove incomplete or inconsistent data, and then converted into a normalized and consistent format.

[0044] Persona Generation

[0045] AI model parameter set

[0046] server

[0047] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[0048] Persona Generation

[0049] server

[0050] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[0051] Product survey design

[0052] Setting up questionnaire items

[0053] User

[0054] A user in the product development department designs a survey item to inquire about specific features of a new product, such as, "What do you think about the design of the new smartwatch?"

[0055] Creating a survey format

[0056] server

[0057] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0058] Conducting a survey using personas

[0059] Sending a survey

[0060] server

[0061] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0062] Response collection

[0063] server

[0064] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0065] Compilation and analysis of survey results

[0066] Data aggregation

[0067] server

[0068] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[0069] Trend and outlier detection

[0070] server

[0071] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0072] Feedback of results

[0073] Results display

[0074] Terminal

[0075] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0076] Improvement suggestions

[0077] server

[0078] Based on the analysis results, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[0079] The present invention is a system that collects diverse viewpoints quickly and economically through the above processing steps, and supports effective product development.

[0080] The processing flow will be explained below.

[0081] Step 1:

[0082] Data collection

[0083] server

[0084] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[0085] Step 2:

[0086] Data Preprocessing

[0087] server

[0088] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes the data to make it consistent, including filling in missing values ​​and standardizing the data format.

[0089] Step 3:

[0090] Preparing the parameter set

[0091] server

[0092] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[0093] Step 4:

[0094] Launching the AI ​​model and generating personas

[0095] server

[0096] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. Specifically, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" is generated.

[0097] Step 5:

[0098] Setting up questionnaire items

[0099] User

[0100] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0101] Step 6:

[0102] Creating a survey format

[0103] server

[0104] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice and open-ended question formats.

[0105] Step 7:

[0106] Sending a survey

[0107] server

[0108] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0109] Step 8:

[0110] Response collection

[0111] server

[0112] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0113] Step 9:

[0114] Data aggregation

[0115] server

[0116] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[0117] Step 10:

[0118] Trend and outlier detection

[0119] server

[0120] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0121] Step 11:

[0122] Results display

[0123] Terminal

[0124] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0125] Step 12:

[0126] Improvement suggestions

[0127] server

[0128] Based on the analysis results, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[0129] Example 1

[0130] 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."

[0131] With conventional questionnaire surveys, it was difficult to collect customer opinions quickly and economically, making it difficult to develop products that reflected diverse perspectives.In addition, the time and effort required to design questionnaire items, collect responses, and analyze them made it difficult to respond quickly to market needs.

[0132] 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.

[0133] In this invention, the server includes means for generating various personas using a generative AI model, means for collecting user data from an internal database or external data sources and classifying and preprocessing the data, means for conducting a product-related survey with the generated personas, means for aggregating the survey results and generating statistical information, means for analyzing product improvements based on the generated statistical information, and means for displaying the analysis results on a user terminal. This makes it possible to quickly and economically collect diverse customer opinions and realize efficient product improvements.

[0134] A "generative AI model" is a model that uses artificial intelligence technology to generate diverse virtual characters (personas) based on specific purposes and requirements.

[0135] A "persona" is a fictional user with specific attributes and characteristics, used to clarify the target users of a product.

[0136] A "server" is a computer system for collecting, processing, storing, analyzing, etc. data.

[0137] "User data" refers to data that includes people's attribute information such as age, gender, occupation, and hobbies.

[0138] "Data source" refers to the information source from which the necessary data is collected, and includes internal databases and external market research databases.

[0139] "Data cleaning" is the process of improving data quality by correcting or removing duplicates, incompleteness, and inconsistencies.

[0140] "Data preprocessing" refers to a series of steps that convert raw data into a format suitable for subsequent analysis or model training.

[0141] A "questionnaire" is a research tool used to gather opinions and evaluations from respondents through specific questions.

[0142] "Natural language processing technology" is a technology that allows computers to analyze, understand, and generate human language.

[0143] "Statistical information" refers to information such as averages, percentages, and trends calculated based on collected data.

[0144] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[0145] "User terminal" refers to a computer or mobile device that a user uses to access the system, display results, and perform operations.

[0146] "Statistical analysis tools" are software and algorithms used to analyze collected data and extract statistics and trends.

[0147] This invention relates to a system that generates various personas using a generative AI model, conducts product-related surveys using those personas, and analyzes areas for improvement of the product based on the survey results. This system operates in cooperation with a server, terminals, and users.

[0148] Data collection and classification

[0149] server

[0150] The server collects user data from internal databases and external market research data sources. Specifically, it retrieves data from external APIs via HTTP requests and stores it in its own database. This data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, it collects attribute information such as "female in her 20s, student, hobby is fashion."

[0151] Data Preprocessing

[0152] server

[0153] The server cleans the collected data, removing incomplete or inconsistent data, removing duplicates and filling in missing data, and normalizing and encoding categorical data to convert it into a consistent format.

[0154] AI model parameter set

[0155] server

[0156] The server creates the parameter set required for the generative AI model based on the preprocessed data. Specifically, it sets the hyperparameters (e.g., learning rate, number of epochs) required for persona generation and prepares a dataset that reflects the characteristics of the target user group.

[0157] Persona Generation

[0158] server

[0159] The server uses the generative AI model to generate various personas. Specifically, it inputs characteristics such as "25-year-old female, hobby is fitness" into the AI ​​model and generates a persona of "25-year-old fitness enthusiast, occupation is marketing."

[0160] Survey design

[0161] User

[0162] A user (a product development department employee) designs a survey to inquire about specific features of a new product, such as "What do you think about the battery life of the new smartwatch?"

[0163] Creating a survey format

[0164] server

[0165] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including formats with multiple choice options (e.g., very good, good, average, bad, very bad) and open-ended questions.

[0166] Sending surveys and collecting responses

[0167] server

[0168] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0169] server

[0170] The server collects the survey responses from each persona and stores the data in a database. Specifically, the responses are classified and stored for each persona.

[0171] Compilation and analysis of survey results

[0172] server

[0173] The server aggregates all survey results and generates statistics, such as response rates and average scores, which are displayed in graphs and tables.

[0174] server

[0175] The server uses statistical analysis tools to detect key trends and outliers in the survey results, such as discovering that users in their 30s are dissatisfied with the prices.

[0176] Display of analysis results and improvement suggestions

[0177] Terminal

[0178] The device displays the aggregated and analyzed results obtained from the server to the user. For example, the dashboard might show a result such as, "The design of the new smartwatch has been well-received, but improvements in battery life are needed."

[0179] server

[0180] Based on the analysis results, the server provides specific improvement suggestions to the product development department. For example, it may generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[0181] Examples and prompts

[0182] Specific examples

[0183] For example, consider the case where the following survey is conducted on a female persona in her 20s.

[0184] Example personas:

[0185] Age: 25

[0186] Gender: Female

[0187] Occupation: Student

[0188] Hobbies: Fashion

[0189] Survey question examples:

[0190] What do you think of the new smartwatch design?

[0191] Very good

[0192] good

[0193] usually

[0194] bad

[0195] Very bad

[0196] Example prompts

[0197] "Please tell us your thoughts on the design of new products. For example, what is your opinion on the design of the new smartwatch?"

[0198] By using the above-mentioned methods, this system can quickly and economically collect diverse customer opinions and realize efficient product improvements.

[0199] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0200] Step 1:

[0201] Data collection and classification

[0202] server

[0203] The server collects user data from an internal database and external market research data sources. This process uses API requests to retrieve data and stores it in the internal database. The input data is user data obtained from external data sources, and the output is data organized and categorized in the internal database. Specifically, the server uses an API key to access the market research site and retrieves data in JSON format.

[0204] Step 2:

[0205] Data Preprocessing

[0206] server

[0207] The server cleans the collected data and removes incomplete or inconsistent data. The input is raw data stored in a database, and the output is preprocessed, consistent data. Specific operations include imputing missing values, removing duplicate data, normalizing data, and encoding categorical data.

[0208] Step 3:

[0209] AI model parameter set

[0210] server

[0211] The server creates the parameter set required for the generative AI model based on the preprocessed data. The input is a normalized dataset, and the output is the hyperparameters to be applied to the AI ​​model. Specifically, it sets parameters such as the learning rate, number of epochs, and batch size to reflect the characteristics of the target user group in the model.

[0212] Step 4:

[0213] Persona Generation

[0214] server

[0215] The server uses a generative AI model to generate various personas. The input is the AI ​​model and parameter set feature data, and the output is a list of generated personas. Specifically, feature data is input into the model to generate a detailed persona such as "25-year-old fitness enthusiast, occupation: marketing."

[0216] Step 5:

[0217] Survey design

[0218] User

[0219] A user (a product development department employee) designs a questionnaire to inquire about specific features of a new product. The input is a list of proposed questions based on the user's needs and product features, and the output is the designed questionnaire items. Specific actions include designing questions such as "What do you think about the battery life of the new smartwatch?"

[0220] Step 6:

[0221] Creating a survey format

[0222] server

[0223] The server automatically generates a questionnaire format based on the questionnaire items set by the user. The input is the questionnaire questions set by the user, and the output is a format that includes options and free-form questions. Specifically, the server automatically arranges the options and completes the format.

[0224] Step 7:

[0225] Sending surveys and collecting responses

[0226] server

[0227] The server distributes questionnaires to the generated personas and collects responses. The input is the generated persona and the questionnaire format, and the output is the response data from each persona. Specifically, it uses natural language processing technology based on the persona's attribute information to generate realistic responses and save them in a response database.

[0228] Step 8:

[0229] Compilation and analysis of survey results

[0230] server

[0231] The server aggregates all survey results and generates statistical information. The input is the collected response data, and the output is the results of statistical information and trend analysis. Specific operations include calculating response ratios and average scores and displaying them in graphs and tables.

[0232] Step 9:

[0233] Display of analysis results and improvement suggestions

[0234] Terminal

[0235] The device displays the aggregated and analyzed results obtained from the server to the user. The input is the statistical information sent from the server, and the output is the result displayed visually to the user. Specifically, the result "The design of the new smartwatch has been well received, but improvements in battery life are needed" is displayed on the dashboard.

[0236] server

[0237] Based on the analysis results, the server provides specific improvement suggestions to the product development department. The input is statistical information and trend analysis results, and the output is specific improvement suggestions. The specific operation is to generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[0238] (Application example 1)

[0239] 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."

[0240] Gathering feedback quickly and efficiently from a diverse user base and identifying areas for product improvement is time-consuming and costly using conventional methods. Online shopping sites, in particular, need a system that can accurately collect diverse user opinions, statistically analyze them, and use them to improve products.

[0241] 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.

[0242] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting product-related questionnaires to the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, and means for setting and sending questionnaire items from a smartphone and displaying responses. This enables the online shopping site to efficiently collect feedback from various perspectives and use it to improve products.

[0243] plaintext

[0244] A "generative AI model" is an artificial intelligence model that learns from data and generates a virtual persona with characteristics that suit a specific purpose.

[0245] A "persona" is a virtual character that represents a specific user group and mimics feedback about a product based on their attributes and behavior.

[0246] A "survey" is a survey in the form of questions that are used to gather responses to specific questions and understand users' opinions and preferences.

[0247] "Aggregation" is the process of organizing and classifying data such as survey results and generating statistical information.

[0248] "Statistical information" is numerical information calculated based on collected data, and indicates trends and characteristics of the data.

[0249] "Areas for improvement in the product" are areas of the product that require improvement in performance, design, specifications, etc., identified based on user feedback.

[0250] A "user terminal" is a device such as a smartphone or computer operated by a user, which generates and receives information.

[0251] "Natural language processing technology" is a technology for analyzing and generating sentences entered by users and answers generated by personas, and is part of artificial intelligence.

[0252] A "smartphone" is a type of mobile phone that can connect to the Internet and use multifunctional applications.

[0253] "Survey items" are specific questions set for users to answer in a questionnaire.

[0254] The system of this invention generates various personas related to an online shopping site, conducts product-related surveys with these personas, and aggregates and analyzes the results to identify areas for improvement in the products. An embodiment of this system is shown below.

[0255] System Configuration

[0256] The system consists of a server, user terminals, and smartphone devices, and uses software such as a database management system (MySQL), a generative AI model (OpenAI GPT-4), statistical analysis tools (Python Pandas, NumPy), and natural language processing technology (SpaCy, NLTK).

[0257] Program processing

[0258] 1. Data collection and classification

[0259] The server collects user data from internal databases and external market research data sources. This data is categorized based on attributes such as age, gender, occupation, and hobbies. As a result, data with diverse user characteristics is collected.

[0260] 2. Data Preprocessing

[0261] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes and converts it into a consistent format, preparing the data for input into the generative AI model.

[0262] 3. Persona Generation

[0263] The server sets the parameters of the generative AI model based on the preprocessed data, which generates a variety of personas that represent specific user demographics. For example, a persona might be generated such as "35 years old, male, hobby is fishing, occupation is engineer."

[0264] 4. Setting up the questionnaire items and creating the format

[0265] The user (a product development department employee) sets up a questionnaire about specific features of a new product. For example, they set up a question such as, "What do you think about the design of the new smartwatch?" The server automatically generates a questionnaire format based on the questionnaire items set by the user.

[0266] 5. Conducting surveys and collecting responses

[0267] The server distributes questionnaires to the generated personas. Each persona generates realistic answers using natural language processing technology. For example, a response such as "The design of this smartwatch is sophisticated, and I want to share it with my friends on social media" may be generated. The server collects these responses and stores them in a database.

[0268] 6. Calculation and analysis of results

[0269] The server aggregates all survey results and generates statistics. For example, the aggregated result may be "80% of personas are satisfied with the new features." Furthermore, statistical analysis tools are used to detect trends and outliers in the survey results.

[0270] 7. Feedback of results

[0271] The server then displays the results of the analysis on the user's device. For example, it may display a result such as, "The design is highly rated, but battery life needs improvement." It also provides specific improvement suggestions to the product development department based on the identified areas for improvement.

[0272] Prompt Sentence Examples

[0273] "What do female users in their 20s think about new smartphone designs?"

[0274] This system makes it possible to efficiently collect feedback from a variety of perspectives and quickly and accurately improve the products offered on the online shopping site, thereby improving customer satisfaction and supporting the development of competitive products.

[0275] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0276] plaintext

[0277] Step 1: Data collection and classification

[0278] The server collects user data from internal databases and external market research data sources. The input includes user attribute data such as age, gender, occupation, and hobbies. Based on this data, the server categorizes the data based on user characteristics. The output is user data by category.

[0279] Step 2: Data Preprocessing

[0280] The server cleans the collected data, removing incomplete or inconsistent data, then normalizes and converts the data into a consistent format. The input includes classified user data. The output is preprocessed, cleaned data.

[0281] Step 3: Generate personas

[0282] The server sets the parameters of the generative AI model based on the preprocessed data, which generates diverse personas that represent specific user demographics. The inputs include the preprocessed data and model parameters. The output is a diverse set of generated personas.

[0283] Step 4: Setting up the questionnaire and creating the format

[0284] The user sets up questionnaire items about specific features of the new product. The server automatically generates a questionnaire format based on the questionnaire items set up by the user. The input includes the questionnaire items set up by the user. The output is the generated questionnaire format.

[0285] Step 5: Conduct the survey and collect responses

[0286] The server distributes surveys to the generated personas. Each persona generates realistic responses using natural language processing techniques. The inputs include the generated personas and the survey format. As an output, the personas' survey responses are collected and stored in a database.

[0287] Step 6: Compile and analyze the results

[0288] The server aggregates all survey results and generates statistics. It also uses statistical analysis tools to detect trends and outliers in the survey results. Inputs include persona survey responses. Outputs include statistics and trend analysis results.

[0289] Step 7: Feedback on results

[0290] The server displays the aggregated and analyzed results on the user's device. Based on the analysis results, it provides specific improvement proposals to the product development department. The input includes statistical information and trend analysis results. The output is improvement proposals that are displayed on the user's device.

[0291] 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.

[0292] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results and emotion recognition. The program processing of this system is explained below in natural language.

[0293] Preparing for Persona Generation

[0294] Data collection and classification

[0295] server

[0296] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, attribute information such as "female in her 30s, working in finance, running as a hobby" is collected.

[0297] Data Preprocessing

[0298] server

[0299] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[0300] Persona Generation

[0301] AI model parameter set

[0302] server

[0303] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[0304] Persona Generation

[0305] server

[0306] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[0307] Product survey design

[0308] Setting up questionnaire items

[0309] User

[0310] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0311] Creating a survey format

[0312] server

[0313] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0314] Conducting a survey using personas

[0315] Sending a survey

[0316] server

[0317] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0318] Response collection

[0319] server

[0320] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0321] Analysis by emotion engine

[0322] Conducting sentiment analysis

[0323] server

[0324] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[0325] Aggregating emotion data

[0326] server

[0327] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[0328] Aggregation and analysis of survey results and sentiment data

[0329] Data aggregation

[0330] server

[0331] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[0332] Trend and outlier detection

[0333] server

[0334] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0335] Feedback of results

[0336] Results display

[0337] Terminal

[0338] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0339] Emotional Feedback

[0340] Terminal

[0341] Based on the analysis results of the emotion engine, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[0342] Improvement suggestions

[0343] server

[0344] Based on the analysis results and sentiment data, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[0345] The present invention is a system that collects a variety of viewpoints and emotion data quickly and economically through the above processing steps, and supports effective product development.

[0346] The processing flow will be explained below.

[0347] Step 1:

[0348] Data collection

[0349] server

[0350] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[0351] Step 2:

[0352] Data Preprocessing

[0353] server

[0354] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[0355] Step 3:

[0356] Preparing the parameter set

[0357] server

[0358] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[0359] Step 4:

[0360] Launching the AI ​​model and generating personas

[0361] server

[0362] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. For example, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" can be generated.

[0363] Step 5:

[0364] Setting up questionnaire items

[0365] User

[0366] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0367] Step 6:

[0368] Creating a survey format

[0369] server

[0370] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0371] Step 7:

[0372] Sending a survey

[0373] server

[0374] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0375] Step 8:

[0376] Response collection

[0377] server

[0378] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0379] Step 9:

[0380] Conducting sentiment analysis

[0381] server

[0382] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[0383] Step 10:

[0384] Aggregating emotion data

[0385] server

[0386] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[0387] Step 11:

[0388] Data aggregation

[0389] server

[0390] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[0391] Step 12:

[0392] Trend and outlier detection

[0393] server

[0394] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0395] Step 13:

[0396] Results display

[0397] Terminal

[0398] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0399] Step 14:

[0400] Emotional Feedback

[0401] Terminal

[0402] Based on the results of the emotion analysis obtained from the server, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[0403] Step 15:

[0404] Improvement suggestions

[0405] server

[0406] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[0407] Example 2

[0408] 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."

[0409] Conventional product improvement support systems have struggled to quickly and efficiently collect diverse user perspectives and emotional data and accurately analyze product improvements. Furthermore, collecting survey data and providing feedback through emotional analysis is a time-consuming and labor-intensive process. To address these challenges, the present invention provides a system that uses a generative AI model to generate diverse personas and efficiently collect and analyze survey data.

[0410] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user data from an internal database and an external data source and cleaning and normalizing the data, means for creating a parameter set for a generative AI model based on the collected and preprocessed data and generating a persona, means for conducting a product-related survey with the generated persona, means for collecting survey results and emotional data and storing the same in a database, means for generating statistical information based on the collected survey results and emotional data and detecting trends and outliers, means for analyzing product improvements based on the generated statistical information and emotional data, and means for displaying the analysis results on a user terminal. This enables rapid and efficient collection and analysis of diverse perspectives and emotional data.

[0411] An "internal database" is a collection of data managed within the system, and is a storage device for storing various user information, past survey results, etc.

[0412] "External data sources" are information sources for obtaining data from outside the system, including market research data and publicly available information.

[0413] "User data" refers to data that includes the attribute information of a specific user, such as age, gender, occupation, hobbies, and purchasing history.

[0414] "Cleaning" is the process of removing incomplete or inconsistent data from collected data and preparing it in a format suitable for data analysis.

[0415] "Normalization" is the process of converting collected data into a consistent format and maintaining consistency between the data.

[0416] A "generative AI model" is an algorithm that uses artificial intelligence (AI) technology to generate diverse personas, taking a specific set of parameters as input.

[0417] A "parameter set" refers to a set of data or conditions that are input into a generative AI model, which then generates a specific persona based on these parameters.

[0418] A "persona" is a virtual character that simulates a specific user profile and has attribute information such as age, gender, occupation, and hobbies.

[0419] A "questionnaire" is a survey method used to solicit opinions on specific questions, and is conducted in a multiple choice or free description format.

[0420] "Emotional data" refers to emotional information analyzed based on questionnaire responses, and includes emotional states such as satisfaction and dissatisfaction.

[0421] "Statistical information" refers to data that shows aggregated results and trends calculated from collected data, and includes various trends and outliers.

[0422] "Trend" refers to a general pattern of tendency or fluctuation found in a large amount of collected data.

[0423] An "outlier" is data that shows significantly different values ​​or patterns compared to other data, and may indicate a particular issue or problem.

[0424] This invention is a system that generates various personas using a generative AI model, conducts product-related surveys using these personas, and analyzes product improvements based on the results and sentiment analysis. The main roles of this system are played by the server, users, and terminals.

[0425] 1. Data collection and classification

[0426] Server Data Collection

[0427] The server collects user data from an internal database and external data sources. The internal database includes existing customer information, purchase history, survey results, etc. Market research data and public data are also used as external data sources. The server uses a data collection API to obtain attribute information such as "female in her 30s, working in finance, hobby is running."

[0428] Data classification process

[0429] The server categorizes the collected data into categories such as age, gender, occupation, hobbies, and purchasing history. This categorization creates groups of users with specific attributes, which are used for analysis and persona generation.

[0430] 2. Data Preprocessing

[0431] Data Cleaning

[0432] The server removes incomplete or inconsistent data from the collected data, for example, deleting records with missing occupation data, and normalizing data formats if they are not standardized.

[0433] Data normalization

[0434] The server converts the data into a consistent format, for example by aligning all age data to "decades."

[0435] 3. Parameter set of the AI ​​model

[0436] Creating a Parameter Set

[0437] The server uses the preprocessed data to create a parameter set for the generative AI model, which includes data corresponding to a specific user profile, such as "35 years old, male, engineer, fishing hobby."

[0438] 4. Persona Generation

[0439] Persona generation process

[0440] The server uses a generative AI model to generate various personas, resulting in a specific persona such as "35 years old, male, engineer, fishing hobby." The generated personas are stored in an internal database.

[0441] 5. Setting up the questionnaire items

[0442] Survey design

[0443] The user (a product development department employee) sets specific questions about the product, for example, designing a questionnaire item such as "What do you think about the design of the new smartwatch?"

[0444] 6. Creating a survey format

[0445] Generate a questionnaire format

[0446] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice questions and open-ended questions, and stores the format in electronic format.

[0447] 7. Conduct a survey using personas

[0448] Survey distribution processing

[0449] The server then sends a questionnaire to the persona, which uses natural language processing technology to generate realistic responses based on its attributes. For example, a scenario might be generated in which a 35-year-old male engineer gives positive feedback on the design of a new smartwatch.

[0450] Collecting responses

[0451] The server collects the persona's responses to the questionnaire and stores them in a database for later analysis.

[0452] 8. Conducting sentiment analysis

[0453] Emotional Data Analysis

[0454] The server performs sentiment analysis based on the collected survey responses, using an emotion engine to identify emotional data such as "the persona is satisfied based on the responses" or "they are surprised."

[0455] 9. Data Collection and Analysis

[0456] Data aggregation process

[0457] The server statistically aggregates the survey results and sentiment data, generating aggregated results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design."

[0458] Trend and outlier detection

[0459] The server uses the statistics to detect key trends and outliers, such as finding that certain age groups are unhappy with prices or that certain interests are responding positively to new features.

[0460] 10. Feedback of results

[0461] Display the results on the user's device

[0462] The server displays the analysis results on the user's device, where the user can see results such as "Male users in their 40s are very satisfied with the new features" and "They rate the design highly, but are often dissatisfied with the battery life."

[0463] 11. Improvement suggestions

[0464] Proposal for improvement

[0465] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life" or "improve the finer details of the design," which are displayed on the user's device.

[0466] Through the above processing, this system quickly and efficiently collects diverse viewpoints and emotional data, supporting effective product development.

[0467] Prompt Sentence Examples

[0468] Below is an example of an input prompt for a generative AI model.

[0469] Generate feedback for the design of a new smartwatch based on the demographic information "female in her 30s, working in finance, running as a hobby."

[0470] Example of feedback: Positive about the design, but dissatisfied with the battery life.

[0471] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0472] Step 1: Data collection and classification

[0473] The server collects user data from an internal database and external data sources. As input, customer information from the internal database and external market research data are provided to the server. The server obtains this data via API and stores the data obtained from each data source together. As output, user data categorized by age, gender, occupation, hobbies, purchase history, etc. is obtained.

[0474] Step 2: Data cleaning

[0475] The server removes incomplete or inconsistent data from the data collected in step 1. User data collected in various formats is provided to the server as input. The server performs processing such as deleting records with missing values ​​and imputing some missing values. The output is cleaned user data.

[0476] Step 3: Data normalization

[0477] The server normalizes the cleaned data and converts it into a consistent format. As input, the cleaned user data is provided to the server. For example, the server converts age data into a unified format (e.g., decade units) to align all data formats. As output, the server obtains user data in a consistent format.

[0478] Step 4: Creating a parameter set for the AI ​​model

[0479] The server creates a parameter set for the generative AI model based on the preprocessed data. Normalized user data is provided to the server as input. The server extracts parameters corresponding to each persona (e.g., "35 years old, male, engineer, hobby is fishing") and creates a parameter set to input into the generative AI model. The parameter set for the generative AI model is obtained as output.

[0480] Step 5: Persona Generation

[0481] The server uses a generative AI model to generate various personas. A set of parameters is input to the generative AI model. The server runs the generative AI model to generate various personas (e.g., "35 years old, male, engineer, hobby is fishing") and stores them in an internal database. The generated personas are obtained as output.

[0482] Step 6: Setting up the survey items

[0483] The user sets specific questions about the product. As input, the user is provided with basic information for setting up the questionnaire items. The user designs specific questionnaire items such as "What do you think about the design of the new smartwatch?". As output, the set questionnaire items are obtained.

[0484] Step 7: Generate the survey format

[0485] The server automatically generates a questionnaire based on the questionnaire items set by the user. As input, the set questionnaire items are provided to the server. The server creates an electronic format containing multiple choice and open-ended questions. As output, the generated questionnaire is obtained.

[0486] Step 8: Send out the survey

[0487] The server distributes the questionnaire to the generated personas. As input, the server is provided with the questionnaire format and persona data. The server sends the questionnaire to each persona and collects the responses. As output, the persona's questionnaire responses are obtained.

[0488] Step 9: Generate and collect answers

[0489] The server generates the answers of the generated persona using natural language processing technology. As input, the persona data and the questionnaire form are provided to the server. The server uses natural language processing technology to generate realistic answers and stores them in a database. As output, the server obtains the questionnaire answers generated by natural language processing.

[0490] Step 10: Sentiment Analysis

[0491] The server performs sentiment analysis based on the collected survey responses. The persona's survey responses are provided to the server as input. The server uses an emotion engine to identify emotions such as "the persona is satisfied" or "surprised" based on the responses. The analyzed emotional data is obtained as output.

[0492] Step 11: Data collection and analysis

[0493] The server statistically aggregates the survey results and emotional data. The survey responses and emotional data are provided to the server as input. The server statistically analyzes this data and obtains results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design." The output is aggregated statistical information.

[0494] Step 12: Detect trends and outliers

[0495] The server detects key trends and outliers from the collected data. As input, aggregated statistics are provided to the server. The server uses statistical analysis tools to determine results such as "certain age groups are dissatisfied with the price" or "certain hobbies are responding positively to new features." As output, the detected trends and outliers are provided.

[0496] Step 13: Feedback and display of results

[0497] The server displays the analysis results on the user's device. The analysis results and emotional data are provided to the server as input. The server sends this to the user's device, which then displays the results, such as "Male users in their 40s are very satisfied with the new function." The analysis results are displayed on the user's device as output.

[0498] Step 14: Present improvement suggestions

[0499] The server provides specific improvement suggestions to the product development department based on the analysis results and emotion data. Detailed analysis results and emotion data are provided to the server as input. The server generates specific suggestions, such as "focus on improving battery life" or "improve the finer details of the design," and sends them to the user's device. Specific improvement suggestions are obtained as output.

[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] While conventional persona generation systems can analyze product improvements based on statistical information, they have the problem of being unable to obtain information that reflects actual customer behavior and emotions in real time. This can sometimes make it difficult to accurately evaluate the effects of product improvements, creating a need for more accurate and immediate data collection and analysis.

[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 various personas using a generative AI model, means for conducting a product-related questionnaire for the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, means for capturing customer behavior and facial expressions using smart glasses, and means for analyzing customer emotions in real time based on the captured data. This makes it possible to collect data reflecting customer behavior and emotions in real time and instantly identify areas for improvement in products and services.

[0505] A "generative AI model" is an artificial intelligence technology used to automatically generate diverse personas.

[0506] A "persona" is a virtual user profile with specific attributes and behavioral patterns.

[0507] A "questionnaire" is a method for asking questions to subjects and collecting their responses.

[0508] "Statistical information" is information presented in the form of numbers, graphs, etc. obtained by aggregating and analyzing multiple data.

[0509] "Analysis" is the process of evaluating collected data and extracting meaningful information.

[0510] A "user terminal" is a device such as a computer or smartphone that is used to display results and data.

[0511] "Smart glasses" are wearable devices that the wearer uses to display external information and collect data through sensors.

[0512] "Behavioral capture" is a method of observing customer behavior and recording it as digital information.

[0513] "Facial expression capture" is a method of observing a customer's facial expressions and recording them digitally.

[0514] "Emotion analysis" is a technology that identifies and evaluates the emotions expressed by customers based on captured facial expression data.

[0515] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[0516] This invention is a system that uses smart glasses in brick-and-mortar stores to analyze customer behavior and facial expressions in real time and utilizes generative AI models based on that data. By analyzing customer emotions and behavior, this system can instantly identify product evaluations and in-store improvements.

[0517] Hardware and Software Configuration

[0518] To implement this system, the following hardware and software are used.

[0519] Hardware

[0520] Smart glasses: Wearable devices for capturing customer behavior and facial expressions.

[0521] Server: Collects, analyzes and manages data.

[0522] User terminal: The device that displays and manipulates data (e.g., PC, smartphone).

[0523] software

[0524] Generative AI models: Used to generate diverse personas.

[0525] Sentiment analysis engine: Technology that analyzes customer sentiment in real time based on captured data.

[0526] Natural language processing techniques: Used to generate answers for personas.

[0527] Specific operating procedures for the system

[0528] 1. Data collection: The server captures customer behavior and facial expression data in real time through the smart glasses.

[0529] 2. Data Preprocessing: The server preprocesses the captured data and converts it into a format that can be analyzed by the sentiment analysis engine.

[0530] 3. Sentiment analysis: The server uses a sentiment analysis engine to analyze customer sentiment in real time.

[0531] 4. Persona generation: The server uses a generative AI model to generate a persona with specific attributes and behavioral patterns.

[0532] 5. Conducting a survey: The server conducts a survey about the product to the generated personas and collects responses.

[0533] 6. Response analysis: The server aggregates the collected survey results and generates statistical information.

[0534] 7. Analysis of Improvement Points: The server analyzes the improvement points of the product based on the generated statistical information.

[0535] 8. Displaying results: The server displays the analysis results on the user's terminal in real time.

[0536] Specific examples

[0537] For example, when a customer in a store picks up a new smartwatch, their behavior and facial expression data are captured by the smart glasses. If the emotion analysis engine detects emotions such as "surprise" or "satisfaction," the server can instantly generate information such as "this smartwatch is popular among men in their 20s and 30s" and display it on the user's device.

[0538] Prompt Sentence Examples

[0539] Below are examples of prompts that can be used to conduct a survey on the generated personas.

[0540] Age: 30, Gender: Female, Purchasing history: Unknown, Hobbies: Fashion

[0541] In this way, the system of the present invention makes it possible to analyze customer sentiment and behavior in real time, evaluate products, and identify areas for improvement.

[0542] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0543] Step 1:

[0544] The server captures customer behavior and facial expression data in real time through the smart glasses. When customers browse and pick up products in the store, the video data is collected by the camera in the smart glasses. The input is the video data, and the output is the captured raw data.

[0545] Step 2:

[0546] The server preprocesses the captured video data and converts it into a format that can be analyzed by the emotion analysis engine. Specifically, it converts the video data to grayscale, trims unnecessary parts, and extracts facial features. The input is the captured video data, and the output is preprocessed facial data.

[0547] Step 3:

[0548] The server performs emotion analysis using the preprocessed facial data. It uses an emotion analysis engine to identify emotions such as happiness, surprise, sadness, etc. based on the data. The input is the preprocessed facial data, and the output is the identified emotion data.

[0549] Step 4:

[0550] The server inputs customer data into a generative AI model to generate various personas. For example, attribute data such as a customer's age, gender, purchasing history, and hobbies are input to create a virtual user profile. The input is customer attribute data, and the output is the generated persona.

[0551] Step 5:

[0552] The server then conducts a product-related questionnaire for the generated persona. The questionnaire is designed to collect customer preferences and opinions using natural language processing technology. The input is the persona and the questionnaire items, and the output is the persona's responses to the questionnaire.

[0553] Step 6:

[0554] The server aggregates the collected survey results and generates statistical information. It analyzes the response data of each persona and extracts common trends and outliers. The input is the survey response data, and the output is statistical information.

[0555] Step 7:

[0556] The server analyzes product improvements based on the generated statistical information. It identifies trends and outliers to determine which parts of the product or service should be improved. The input is statistical information, and the output is the identification of areas for improvement.

[0557] Step 8:

[0558] The server displays the analysis results in real time on the user's device. The user's device receives information on product evaluations and areas for improvement and displays it in a format that is intuitively easy for the user to understand. The input is the analysis results, and the output is the data displayed on the user's device.

[0559] 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.

[0560] 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.

[0561] 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.

[0562] [Second embodiment]

[0563] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0564] 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.

[0565] 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).

[0566] 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.

[0567] 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.

[0568] 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).

[0569] 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.

[0570] 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.

[0571] 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.

[0572] 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.

[0573] 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.

[0574] 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."

[0575] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results. The program processing of this system is explained below in natural language.

[0576] Preparing for Persona Generation

[0577] Data collection and classification

[0578] server

[0579] The server collects user data from internal databases and external market research data sources. The collected data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, attribute information such as "female in her 20s, student, hobby is fashion" may be collected.

[0580] Data Preprocessing

[0581] server

[0582] The collected data is cleaned to remove incomplete or inconsistent data, and then converted into a normalized and consistent format.

[0583] Persona Generation

[0584] AI model parameter set

[0585] server

[0586] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[0587] Persona Generation

[0588] server

[0589] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[0590] Product survey design

[0591] Setting up questionnaire items

[0592] User

[0593] A user in the product development department designs a survey item to inquire about specific features of a new product, such as, "What do you think about the design of the new smartwatch?"

[0594] Creating a survey format

[0595] server

[0596] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0597] Conducting a survey using personas

[0598] Sending a survey

[0599] server

[0600] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0601] Response collection

[0602] server

[0603] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0604] Compilation and analysis of survey results

[0605] Data aggregation

[0606] server

[0607] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[0608] Trend and outlier detection

[0609] server

[0610] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0611] Feedback of results

[0612] Results display

[0613] Terminal

[0614] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0615] Improvement suggestions

[0616] server

[0617] Based on the analysis results, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[0618] The present invention is a system that collects diverse viewpoints quickly and economically through the above processing steps, and supports effective product development.

[0619] The processing flow will be explained below.

[0620] Step 1:

[0621] Data collection

[0622] server

[0623] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[0624] Step 2:

[0625] Data Preprocessing

[0626] server

[0627] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes the data to make it consistent, including filling in missing values ​​and standardizing the data format.

[0628] Step 3:

[0629] Preparing the parameter set

[0630] server

[0631] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[0632] Step 4:

[0633] Launching the AI ​​model and generating personas

[0634] server

[0635] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. Specifically, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" is generated.

[0636] Step 5:

[0637] Setting up questionnaire items

[0638] User

[0639] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0640] Step 6:

[0641] Creating a survey format

[0642] server

[0643] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice and open-ended question formats.

[0644] Step 7:

[0645] Sending a survey

[0646] server

[0647] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0648] Step 8:

[0649] Response collection

[0650] server

[0651] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0652] Step 9:

[0653] Data aggregation

[0654] server

[0655] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[0656] Step 10:

[0657] Trend and outlier detection

[0658] server

[0659] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0660] Step 11:

[0661] Results display

[0662] Terminal

[0663] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0664] Step 12:

[0665] Improvement suggestions

[0666] server

[0667] Based on the analysis results, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[0668] Example 1

[0669] 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."

[0670] With conventional questionnaire surveys, it was difficult to collect customer opinions quickly and economically, making it difficult to develop products that reflected diverse perspectives.In addition, the time and effort required to design questionnaire items, collect responses, and analyze them made it difficult to respond quickly to market needs.

[0671] 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.

[0672] In this invention, the server includes means for generating various personas using a generative AI model, means for collecting user data from an internal database or external data sources and classifying and preprocessing the data, means for conducting a product-related survey with the generated personas, means for aggregating the survey results and generating statistical information, means for analyzing product improvements based on the generated statistical information, and means for displaying the analysis results on a user terminal. This makes it possible to quickly and economically collect diverse customer opinions and realize efficient product improvements.

[0673] A "generative AI model" is a model that uses artificial intelligence technology to generate diverse virtual characters (personas) based on specific purposes and requirements.

[0674] A "persona" is a fictional user with specific attributes and characteristics, used to clarify the target users of a product.

[0675] A "server" is a computer system for collecting, processing, storing, analyzing, etc. data.

[0676] "User data" refers to data that includes people's attribute information such as age, gender, occupation, and hobbies.

[0677] "Data source" refers to the information source from which the necessary data is collected, and includes internal databases and external market research databases.

[0678] "Data cleaning" is the process of improving data quality by correcting or removing duplicates, incompleteness, and inconsistencies.

[0679] "Data preprocessing" refers to a series of steps that convert raw data into a format suitable for subsequent analysis or model training.

[0680] A "questionnaire" is a research tool used to gather opinions and evaluations from respondents through specific questions.

[0681] "Natural language processing technology" is a technology that allows computers to analyze, understand, and generate human language.

[0682] "Statistical information" refers to information such as averages, percentages, and trends calculated based on collected data.

[0683] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[0684] "User terminal" refers to a computer or mobile device that a user uses to access the system, display results, and perform operations.

[0685] "Statistical analysis tools" are software and algorithms used to analyze collected data and extract statistics and trends.

[0686] This invention relates to a system that generates various personas using a generative AI model, conducts product-related surveys using those personas, and analyzes areas for improvement of the product based on the survey results. This system operates in cooperation with a server, terminals, and users.

[0687] Data collection and classification

[0688] server

[0689] The server collects user data from internal databases and external market research data sources. Specifically, it retrieves data from external APIs via HTTP requests and stores it in its own database. This data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, it collects attribute information such as "female in her 20s, student, hobby is fashion."

[0690] Data Preprocessing

[0691] server

[0692] The server cleans the collected data, removing incomplete or inconsistent data, removing duplicates and filling in missing data, and normalizing and encoding categorical data to convert it into a consistent format.

[0693] AI model parameter set

[0694] server

[0695] The server creates the parameter set required for the generative AI model based on the preprocessed data. Specifically, it sets the hyperparameters (e.g., learning rate, number of epochs) required for persona generation and prepares a dataset that reflects the characteristics of the target user group.

[0696] Persona Generation

[0697] server

[0698] The server uses the generative AI model to generate various personas. Specifically, it inputs characteristics such as "25-year-old female, hobby is fitness" into the AI ​​model and generates a persona of "25-year-old fitness enthusiast, occupation is marketing."

[0699] Survey design

[0700] User

[0701] A user (a product development department employee) designs a survey to inquire about specific features of a new product, such as "What do you think about the battery life of the new smartwatch?"

[0702] Creating a survey format

[0703] server

[0704] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including formats with multiple choice options (e.g., very good, good, average, bad, very bad) and open-ended questions.

[0705] Sending surveys and collecting responses

[0706] server

[0707] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0708] server

[0709] The server collects the survey responses from each persona and stores the data in a database. Specifically, the responses are classified and stored for each persona.

[0710] Compilation and analysis of survey results

[0711] server

[0712] The server aggregates all survey results and generates statistics, such as response rates and average scores, which are displayed in graphs and tables.

[0713] server

[0714] The server uses statistical analysis tools to detect key trends and outliers in the survey results, such as discovering that users in their 30s are dissatisfied with the prices.

[0715] Display of analysis results and improvement suggestions

[0716] Terminal

[0717] The device displays the aggregated and analyzed results obtained from the server to the user. For example, the dashboard might show a result such as, "The design of the new smartwatch has been well-received, but improvements in battery life are needed."

[0718] server

[0719] Based on the analysis results, the server provides specific improvement suggestions to the product development department. For example, it may generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[0720] Examples and prompts

[0721] Specific examples

[0722] For example, consider the case where the following survey is conducted on a female persona in her 20s.

[0723] Example personas:

[0724] Age: 25

[0725] Gender: Female

[0726] Occupation: Student

[0727] Hobbies: Fashion

[0728] Survey question examples:

[0729] What do you think of the new smartwatch design?

[0730] Very good

[0731] good

[0732] usually

[0733] bad

[0734] Very bad

[0735] Example prompts

[0736] "Please tell us your thoughts on the design of new products. For example, what is your opinion on the design of the new smartwatch?"

[0737] By using the above-mentioned methods, this system can quickly and economically collect diverse customer opinions and realize efficient product improvements.

[0738] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0739] Step 1:

[0740] Data collection and classification

[0741] server

[0742] The server collects user data from an internal database and external market research data sources. This process uses API requests to retrieve data and stores it in the internal database. The input data is user data obtained from external data sources, and the output is data organized and categorized in the internal database. Specifically, the server uses an API key to access the market research site and retrieves data in JSON format.

[0743] Step 2:

[0744] Data Preprocessing

[0745] server

[0746] The server cleans the collected data and removes incomplete or inconsistent data. The input is raw data stored in a database, and the output is preprocessed, consistent data. Specific operations include imputing missing values, removing duplicate data, normalizing data, and encoding categorical data.

[0747] Step 3:

[0748] AI model parameter set

[0749] server

[0750] The server creates the parameter set required for the generative AI model based on the preprocessed data. The input is a normalized dataset, and the output is the hyperparameters to be applied to the AI ​​model. Specifically, it sets parameters such as the learning rate, number of epochs, and batch size to reflect the characteristics of the target user group in the model.

[0751] Step 4:

[0752] Persona Generation

[0753] server

[0754] The server uses a generative AI model to generate various personas. The input is the AI ​​model and parameter set feature data, and the output is a list of generated personas. Specifically, feature data is input into the model to generate a detailed persona such as "25-year-old fitness enthusiast, occupation: marketing."

[0755] Step 5:

[0756] Survey design

[0757] User

[0758] A user (a product development department employee) designs a questionnaire to inquire about specific features of a new product. The input is a list of proposed questions based on the user's needs and product features, and the output is the designed questionnaire items. Specific actions include designing questions such as "What do you think about the battery life of the new smartwatch?"

[0759] Step 6:

[0760] Creating a survey format

[0761] server

[0762] The server automatically generates a questionnaire format based on the questionnaire items set by the user. The input is the questionnaire questions set by the user, and the output is a format that includes options and free-form questions. Specifically, the server automatically arranges the options and completes the format.

[0763] Step 7:

[0764] Sending surveys and collecting responses

[0765] server

[0766] The server distributes questionnaires to the generated personas and collects responses. The input is the generated persona and the questionnaire format, and the output is the response data from each persona. Specifically, it uses natural language processing technology based on the persona's attribute information to generate realistic responses and save them in a response database.

[0767] Step 8:

[0768] Compilation and analysis of survey results

[0769] server

[0770] The server aggregates all survey results and generates statistical information. The input is the collected response data, and the output is the results of statistical information and trend analysis. Specific operations include calculating response ratios and average scores and displaying them in graphs and tables.

[0771] Step 9:

[0772] Display of analysis results and improvement suggestions

[0773] Terminal

[0774] The device displays the aggregated and analyzed results obtained from the server to the user. The input is the statistical information sent from the server, and the output is the result displayed visually to the user. Specifically, the result "The design of the new smartwatch has been well received, but improvements in battery life are needed" is displayed on the dashboard.

[0775] server

[0776] Based on the analysis results, the server provides specific improvement suggestions to the product development department. The input is statistical information and trend analysis results, and the output is specific improvement suggestions. The specific operation is to generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[0777] (Application example 1)

[0778] 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."

[0779] Gathering feedback quickly and efficiently from a diverse user base and identifying areas for product improvement is time-consuming and costly using conventional methods. Online shopping sites, in particular, need a system that can accurately collect diverse user opinions, statistically analyze them, and use them to improve products.

[0780] 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.

[0781] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting product-related questionnaires to the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, and means for setting and sending questionnaire items from a smartphone and displaying responses. This enables the online shopping site to efficiently collect feedback from various perspectives and use it to improve products.

[0782] plaintext

[0783] A "generative AI model" is an artificial intelligence model that learns from data and generates a virtual persona with characteristics that suit a specific purpose.

[0784] A "persona" is a virtual character that represents a specific user group and mimics feedback about a product based on their attributes and behavior.

[0785] A "survey" is a survey in the form of questions that are used to gather responses to specific questions and understand users' opinions and preferences.

[0786] "Aggregation" is the process of organizing and classifying data such as survey results and generating statistical information.

[0787] "Statistical information" is numerical information calculated based on collected data, and indicates trends and characteristics of the data.

[0788] "Areas for improvement in the product" are areas of the product that require improvement in performance, design, specifications, etc., identified based on user feedback.

[0789] A "user terminal" is a device such as a smartphone or computer operated by a user, which generates and receives information.

[0790] "Natural language processing technology" is a technology for analyzing and generating sentences entered by users and answers generated by personas, and is part of artificial intelligence.

[0791] A "smartphone" is a type of mobile phone that can connect to the Internet and use multifunctional applications.

[0792] "Survey items" are specific questions set for users to answer in a questionnaire.

[0793] The system of this invention generates various personas related to an online shopping site, conducts product-related surveys with these personas, and aggregates and analyzes the results to identify areas for improvement in the products. An embodiment of this system is shown below.

[0794] System Configuration

[0795] The system consists of a server, user terminals, and smartphone devices, and uses software such as a database management system (MySQL), a generative AI model (OpenAI GPT-4), statistical analysis tools (Python Pandas, NumPy), and natural language processing technology (SpaCy, NLTK).

[0796] Program processing

[0797] 1. Data collection and classification

[0798] The server collects user data from internal databases and external market research data sources. This data is categorized based on attributes such as age, gender, occupation, and hobbies. As a result, data with diverse user characteristics is collected.

[0799] 2. Data Preprocessing

[0800] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes and converts it into a consistent format, preparing the data for input into the generative AI model.

[0801] 3. Persona Generation

[0802] The server sets the parameters of the generative AI model based on the preprocessed data, which generates a variety of personas that represent specific user demographics. For example, a persona might be generated such as "35 years old, male, hobby is fishing, occupation is engineer."

[0803] 4. Setting up the questionnaire items and creating the format

[0804] The user (a product development department employee) sets up a questionnaire about specific features of a new product. For example, they set up a question such as, "What do you think about the design of the new smartwatch?" The server automatically generates a questionnaire format based on the questionnaire items set by the user.

[0805] 5. Conducting surveys and collecting responses

[0806] The server distributes questionnaires to the generated personas. Each persona generates realistic answers using natural language processing technology. For example, a response such as "The design of this smartwatch is sophisticated, and I want to share it with my friends on social media" may be generated. The server collects these responses and stores them in a database.

[0807] 6. Calculation and analysis of results

[0808] The server aggregates all survey results and generates statistics. For example, the aggregated result may be "80% of personas are satisfied with the new features." Furthermore, statistical analysis tools are used to detect trends and outliers in the survey results.

[0809] 7. Feedback of results

[0810] The server then displays the results of the analysis on the user's device. For example, it may display a result such as, "The design is highly rated, but battery life needs improvement." It also provides specific improvement suggestions to the product development department based on the identified areas for improvement.

[0811] Prompt Sentence Examples

[0812] "What do female users in their 20s think about new smartphone designs?"

[0813] This system makes it possible to efficiently collect feedback from a variety of perspectives and quickly and accurately improve the products offered on the online shopping site, thereby improving customer satisfaction and supporting the development of competitive products.

[0814] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0815] plaintext

[0816] Step 1: Data collection and classification

[0817] The server collects user data from internal databases and external market research data sources. The input includes user attribute data such as age, gender, occupation, and hobbies. Based on this data, the server categorizes the data based on user characteristics. The output is user data by category.

[0818] Step 2: Data Preprocessing

[0819] The server cleans the collected data, removing incomplete or inconsistent data, then normalizes and converts the data into a consistent format. The input includes classified user data. The output is preprocessed, cleaned data.

[0820] Step 3: Generate personas

[0821] The server sets the parameters of the generative AI model based on the preprocessed data, which generates diverse personas that represent specific user demographics. The inputs include the preprocessed data and model parameters. The output is a diverse set of generated personas.

[0822] Step 4: Setting up the questionnaire and creating the format

[0823] The user sets up questionnaire items about specific features of the new product. The server automatically generates a questionnaire format based on the questionnaire items set up by the user. The input includes the questionnaire items set up by the user. The output is the generated questionnaire format.

[0824] Step 5: Conduct the survey and collect responses

[0825] The server distributes surveys to the generated personas. Each persona generates realistic responses using natural language processing techniques. The inputs include the generated personas and the survey format. As an output, the personas' survey responses are collected and stored in a database.

[0826] Step 6: Compile and analyze the results

[0827] The server aggregates all survey results and generates statistics. It also uses statistical analysis tools to detect trends and outliers in the survey results. Inputs include persona survey responses. Outputs include statistics and trend analysis results.

[0828] Step 7: Feedback on results

[0829] The server displays the aggregated and analyzed results on the user's device. Based on the analysis results, it provides specific improvement proposals to the product development department. The input includes statistical information and trend analysis results. The output is improvement proposals that are displayed on the user's device.

[0830] 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.

[0831] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results and emotion recognition. The program processing of this system is explained below in natural language.

[0832] Preparing for Persona Generation

[0833] Data collection and classification

[0834] server

[0835] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, attribute information such as "female in her 30s, working in finance, running as a hobby" is collected.

[0836] Data Preprocessing

[0837] server

[0838] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[0839] Persona Generation

[0840] AI model parameter set

[0841] server

[0842] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[0843] Persona Generation

[0844] server

[0845] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[0846] Product survey design

[0847] Setting up questionnaire items

[0848] User

[0849] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0850] Creating a survey format

[0851] server

[0852] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0853] Conducting a survey using personas

[0854] Sending a survey

[0855] server

[0856] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0857] Response collection

[0858] server

[0859] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0860] Analysis by emotion engine

[0861] Conducting sentiment analysis

[0862] server

[0863] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[0864] Aggregating emotion data

[0865] server

[0866] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[0867] Aggregation and analysis of survey results and sentiment data

[0868] Data aggregation

[0869] server

[0870] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[0871] Trend and outlier detection

[0872] server

[0873] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0874] Feedback of results

[0875] Results display

[0876] Terminal

[0877] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0878] Emotional Feedback

[0879] Terminal

[0880] Based on the analysis results of the emotion engine, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[0881] Improvement suggestions

[0882] server

[0883] Based on the analysis results and sentiment data, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[0884] The present invention is a system that collects a variety of viewpoints and emotion data quickly and economically through the above processing steps, and supports effective product development.

[0885] The processing flow will be explained below.

[0886] Step 1:

[0887] Data collection

[0888] server

[0889] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[0890] Step 2:

[0891] Data Preprocessing

[0892] server

[0893] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[0894] Step 3:

[0895] Preparing the parameter set

[0896] server

[0897] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[0898] Step 4:

[0899] Launching the AI ​​model and generating personas

[0900] server

[0901] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. For example, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" can be generated.

[0902] Step 5:

[0903] Setting up questionnaire items

[0904] User

[0905] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[0906] Step 6:

[0907] Creating a survey format

[0908] server

[0909] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[0910] Step 7:

[0911] Sending a survey

[0912] server

[0913] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[0914] Step 8:

[0915] Response collection

[0916] server

[0917] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[0918] Step 9:

[0919] Conducting sentiment analysis

[0920] server

[0921] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[0922] Step 10:

[0923] Aggregating emotion data

[0924] server

[0925] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[0926] Step 11:

[0927] Data aggregation

[0928] server

[0929] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[0930] Step 12:

[0931] Trend and outlier detection

[0932] server

[0933] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[0934] Step 13:

[0935] Results display

[0936] Terminal

[0937] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[0938] Step 14:

[0939] Emotional Feedback

[0940] Terminal

[0941] Based on the results of the emotion analysis obtained from the server, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[0942] Step 15:

[0943] Improvement suggestions

[0944] server

[0945] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[0946] Example 2

[0947] 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."

[0948] Conventional product improvement support systems have struggled to quickly and efficiently collect diverse user perspectives and emotional data and accurately analyze product improvements. Furthermore, collecting survey data and providing feedback through emotional analysis is a time-consuming and labor-intensive process. To address these challenges, the present invention provides a system that uses a generative AI model to generate diverse personas and efficiently collect and analyze survey data.

[0949] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user data from an internal database and an external data source and cleaning and normalizing the data, means for creating a parameter set for a generative AI model based on the collected and preprocessed data and generating a persona, means for conducting a product-related survey with the generated persona, means for collecting survey results and emotional data and storing the same in a database, means for generating statistical information based on the collected survey results and emotional data and detecting trends and outliers, means for analyzing product improvements based on the generated statistical information and emotional data, and means for displaying the analysis results on a user terminal. This enables rapid and efficient collection and analysis of diverse perspectives and emotional data.

[0950] An "internal database" is a collection of data managed within the system, and is a storage device for storing various user information, past survey results, etc.

[0951] "External data sources" are information sources for obtaining data from outside the system, including market research data and publicly available information.

[0952] "User data" refers to data that includes the attribute information of a specific user, such as age, gender, occupation, hobbies, and purchasing history.

[0953] "Cleaning" is the process of removing incomplete or inconsistent data from collected data and preparing it in a format suitable for data analysis.

[0954] "Normalization" is the process of converting collected data into a consistent format and maintaining consistency between the data.

[0955] A "generative AI model" is an algorithm that uses artificial intelligence (AI) technology to generate diverse personas, taking a specific set of parameters as input.

[0956] A "parameter set" refers to a set of data or conditions that are input into a generative AI model, which then generates a specific persona based on these parameters.

[0957] A "persona" is a virtual character that simulates a specific user profile and has attribute information such as age, gender, occupation, and hobbies.

[0958] A "questionnaire" is a survey method used to solicit opinions on specific questions, and is conducted in a multiple choice or free description format.

[0959] "Emotional data" refers to emotional information analyzed based on questionnaire responses, and includes emotional states such as satisfaction and dissatisfaction.

[0960] "Statistical information" refers to data that shows aggregated results and trends calculated from collected data, and includes various trends and outliers.

[0961] "Trend" refers to a general pattern of tendency or fluctuation found in a large amount of collected data.

[0962] An "outlier" is data that shows significantly different values ​​or patterns compared to other data, and may indicate a particular issue or problem.

[0963] This invention is a system that generates various personas using a generative AI model, conducts product-related surveys using these personas, and analyzes product improvements based on the results and sentiment analysis. The main roles of this system are played by the server, users, and terminals.

[0964] 1. Data collection and classification

[0965] Server Data Collection

[0966] The server collects user data from an internal database and external data sources. The internal database includes existing customer information, purchase history, survey results, etc. Market research data and public data are also used as external data sources. The server uses a data collection API to obtain attribute information such as "female in her 30s, working in finance, hobby is running."

[0967] Data classification process

[0968] The server categorizes the collected data into categories such as age, gender, occupation, hobbies, and purchasing history. This categorization creates groups of users with specific attributes, which are used for analysis and persona generation.

[0969] 2. Data Preprocessing

[0970] Data Cleaning

[0971] The server removes incomplete or inconsistent data from the collected data, for example, deleting records with missing occupation data, and normalizing data formats if they are not standardized.

[0972] Data normalization

[0973] The server converts the data into a consistent format, for example by aligning all age data to "decades."

[0974] 3. Parameter set of the AI ​​model

[0975] Creating a Parameter Set

[0976] The server uses the preprocessed data to create a parameter set for the generative AI model, which includes data corresponding to a specific user profile, such as "35 years old, male, engineer, fishing hobby."

[0977] 4. Persona Generation

[0978] Persona generation process

[0979] The server uses a generative AI model to generate various personas, resulting in a specific persona such as "35 years old, male, engineer, fishing hobby." The generated personas are stored in an internal database.

[0980] 5. Setting up the questionnaire items

[0981] Survey design

[0982] The user (a product development department employee) sets specific questions about the product, for example, designing a questionnaire item such as "What do you think about the design of the new smartwatch?"

[0983] 6. Creating a survey format

[0984] Generate a questionnaire format

[0985] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice questions and open-ended questions, and stores the format in electronic format.

[0986] 7. Conduct a survey using personas

[0987] Survey distribution processing

[0988] The server then sends a questionnaire to the persona, which uses natural language processing technology to generate realistic responses based on its attributes. For example, a scenario might be generated in which a 35-year-old male engineer gives positive feedback on the design of a new smartwatch.

[0989] Collecting responses

[0990] The server collects the persona's responses to the questionnaire and stores them in a database for later analysis.

[0991] 8. Conducting sentiment analysis

[0992] Emotional Data Analysis

[0993] The server performs sentiment analysis based on the collected survey responses, using an emotion engine to identify emotional data such as "the persona is satisfied based on the responses" or "they are surprised."

[0994] 9. Data Collection and Analysis

[0995] Data aggregation process

[0996] The server statistically aggregates the survey results and sentiment data, generating aggregated results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design."

[0997] Trend and outlier detection

[0998] The server uses the statistics to detect key trends and outliers, such as finding that certain age groups are unhappy with prices or that certain interests are responding positively to new features.

[0999] 10. Feedback of results

[1000] Display the results on the user's device

[1001] The server displays the analysis results on the user's device, where the user can see results such as "Male users in their 40s are very satisfied with the new features" and "They rate the design highly, but are often dissatisfied with the battery life."

[1002] 11. Improvement suggestions

[1003] Proposal for improvement

[1004] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life" or "improve the finer details of the design," which are displayed on the user's device.

[1005] Through the above processing, this system quickly and efficiently collects diverse viewpoints and emotional data, supporting effective product development.

[1006] Prompt Sentence Examples

[1007] Below is an example of an input prompt for a generative AI model.

[1008] Generate feedback for the design of a new smartwatch based on the demographic information "female in her 30s, working in finance, running as a hobby."

[1009] Example of feedback: Positive about the design, but dissatisfied with the battery life.

[1010] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1011] Step 1: Data collection and classification

[1012] The server collects user data from an internal database and external data sources. As input, customer information from the internal database and external market research data are provided to the server. The server obtains this data via API and stores the data obtained from each data source together. As output, user data categorized by age, gender, occupation, hobbies, purchase history, etc. is obtained.

[1013] Step 2: Data cleaning

[1014] The server removes incomplete or inconsistent data from the data collected in step 1. User data collected in various formats is provided to the server as input. The server performs processing such as deleting records with missing values ​​and imputing some missing values. The output is cleaned user data.

[1015] Step 3: Data normalization

[1016] The server normalizes the cleaned data and converts it into a consistent format. As input, the cleaned user data is provided to the server. For example, the server converts age data into a unified format (e.g., decade units) to align all data formats. As output, the server obtains user data in a consistent format.

[1017] Step 4: Creating a parameter set for the AI ​​model

[1018] The server creates a parameter set for the generative AI model based on the preprocessed data. Normalized user data is provided to the server as input. The server extracts parameters corresponding to each persona (e.g., "35 years old, male, engineer, hobby is fishing") and creates a parameter set to input into the generative AI model. The parameter set for the generative AI model is obtained as output.

[1019] Step 5: Persona Generation

[1020] The server uses a generative AI model to generate various personas. A set of parameters is input to the generative AI model. The server runs the generative AI model to generate various personas (e.g., "35 years old, male, engineer, hobby is fishing") and stores them in an internal database. The generated personas are obtained as output.

[1021] Step 6: Setting up the survey items

[1022] The user sets specific questions about the product. As input, the user is provided with basic information for setting up the questionnaire items. The user designs specific questionnaire items such as "What do you think about the design of the new smartwatch?". As output, the set questionnaire items are obtained.

[1023] Step 7: Generate the survey format

[1024] The server automatically generates a questionnaire based on the questionnaire items set by the user. As input, the set questionnaire items are provided to the server. The server creates an electronic format containing multiple choice and open-ended questions. As output, the generated questionnaire is obtained.

[1025] Step 8: Send out the survey

[1026] The server distributes the questionnaire to the generated personas. As input, the server is provided with the questionnaire format and persona data. The server sends the questionnaire to each persona and collects the responses. As output, the persona's questionnaire responses are obtained.

[1027] Step 9: Generate and collect answers

[1028] The server generates the answers of the generated persona using natural language processing technology. As input, the persona data and the questionnaire form are provided to the server. The server uses natural language processing technology to generate realistic answers and stores them in a database. As output, the server obtains the questionnaire answers generated by natural language processing.

[1029] Step 10: Sentiment Analysis

[1030] The server performs sentiment analysis based on the collected survey responses. The persona's survey responses are provided to the server as input. The server uses an emotion engine to identify emotions such as "the persona is satisfied" or "surprised" based on the responses. The analyzed emotional data is obtained as output.

[1031] Step 11: Data collection and analysis

[1032] The server statistically aggregates the survey results and emotional data. The survey responses and emotional data are provided to the server as input. The server statistically analyzes this data and obtains results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design." The output is aggregated statistical information.

[1033] Step 12: Detect trends and outliers

[1034] The server detects key trends and outliers from the collected data. As input, aggregated statistics are provided to the server. The server uses statistical analysis tools to determine results such as "certain age groups are dissatisfied with the price" or "certain hobbies are responding positively to new features." As output, the detected trends and outliers are provided.

[1035] Step 13: Feedback and display of results

[1036] The server displays the analysis results on the user's device. The analysis results and emotional data are provided to the server as input. The server sends this to the user's device, which then displays the results, such as "Male users in their 40s are very satisfied with the new function." The analysis results are displayed on the user's device as output.

[1037] Step 14: Present improvement suggestions

[1038] The server provides specific improvement suggestions to the product development department based on the analysis results and emotion data. Detailed analysis results and emotion data are provided to the server as input. The server generates specific suggestions, such as "focus on improving battery life" or "improve the finer details of the design," and sends them to the user's device. Specific improvement suggestions are obtained as output.

[1039] (Application example 2)

[1040] 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."

[1041] While conventional persona generation systems can analyze product improvements based on statistical information, they have the problem of being unable to obtain information that reflects actual customer behavior and emotions in real time. This can sometimes make it difficult to accurately evaluate the effects of product improvements, creating a need for more accurate and immediate data collection and analysis.

[1042] 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.

[1043] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting a product-related questionnaire for the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, means for capturing customer behavior and facial expressions using smart glasses, and means for analyzing customer emotions in real time based on the captured data. This makes it possible to collect data reflecting customer behavior and emotions in real time and instantly identify areas for improvement in products and services.

[1044] A "generative AI model" is an artificial intelligence technology used to automatically generate diverse personas.

[1045] A "persona" is a virtual user profile with specific attributes and behavioral patterns.

[1046] A "questionnaire" is a method for asking questions to subjects and collecting their responses.

[1047] "Statistical information" is information presented in the form of numbers, graphs, etc. obtained by aggregating and analyzing multiple data.

[1048] "Analysis" is the process of evaluating collected data and extracting meaningful information.

[1049] A "user terminal" is a device such as a computer or smartphone that is used to display results and data.

[1050] "Smart glasses" are wearable devices that the wearer uses to display external information and collect data through sensors.

[1051] "Behavioral capture" is a method of observing customer behavior and recording it as digital information.

[1052] "Facial expression capture" is a method of observing a customer's facial expressions and recording them digitally.

[1053] "Emotion analysis" is a technology that identifies and evaluates the emotions expressed by customers based on captured facial expression data.

[1054] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[1055] This invention is a system that uses smart glasses in brick-and-mortar stores to analyze customer behavior and facial expressions in real time and utilizes generative AI models based on that data. By analyzing customer emotions and behavior, this system can instantly identify product evaluations and in-store improvements.

[1056] Hardware and Software Configuration

[1057] To implement this system, the following hardware and software are used.

[1058] Hardware

[1059] Smart glasses: Wearable devices for capturing customer behavior and facial expressions.

[1060] Server: Collects, analyzes and manages data.

[1061] User terminal: The device that displays and manipulates data (e.g., PC, smartphone).

[1062] software

[1063] Generative AI models: Used to generate diverse personas.

[1064] Sentiment analysis engine: Technology that analyzes customer sentiment in real time based on captured data.

[1065] Natural language processing techniques: Used to generate answers for personas.

[1066] Specific operating procedures for the system

[1067] 1. Data collection: The server captures customer behavior and facial expression data in real time through the smart glasses.

[1068] 2. Data Preprocessing: The server preprocesses the captured data and converts it into a format that can be analyzed by the sentiment analysis engine.

[1069] 3. Sentiment analysis: The server uses a sentiment analysis engine to analyze customer sentiment in real time.

[1070] 4. Persona generation: The server uses a generative AI model to generate a persona with specific attributes and behavioral patterns.

[1071] 5. Conducting a survey: The server conducts a survey about the product to the generated personas and collects responses.

[1072] 6. Response analysis: The server aggregates the collected survey results and generates statistical information.

[1073] 7. Analysis of Improvement Points: The server analyzes the improvement points of the product based on the generated statistical information.

[1074] 8. Displaying results: The server displays the analysis results on the user's terminal in real time.

[1075] Specific examples

[1076] For example, when a customer in a store picks up a new smartwatch, their behavior and facial expression data are captured by the smart glasses. If the emotion analysis engine detects emotions such as "surprise" or "satisfaction," the server can instantly generate information such as "this smartwatch is popular among men in their 20s and 30s" and display it on the user's device.

[1077] Prompt Sentence Examples

[1078] Below are examples of prompts that can be used to conduct a survey on the generated personas.

[1079] Age: 30, Gender: Female, Purchasing history: Unknown, Hobbies: Fashion

[1080] In this way, the system of the present invention makes it possible to analyze customer sentiment and behavior in real time, evaluate products, and identify areas for improvement.

[1081] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1082] Step 1:

[1083] The server captures customer behavior and facial expression data in real time through the smart glasses. When customers browse and pick up products in the store, the video data is collected by the camera in the smart glasses. The input is the video data, and the output is the captured raw data.

[1084] Step 2:

[1085] The server preprocesses the captured video data and converts it into a format that can be analyzed by the emotion analysis engine. Specifically, it converts the video data to grayscale, trims unnecessary parts, and extracts facial features. The input is the captured video data, and the output is preprocessed facial data.

[1086] Step 3:

[1087] The server performs emotion analysis using the preprocessed facial data. It uses an emotion analysis engine to identify emotions such as happiness, surprise, sadness, etc. based on the data. The input is the preprocessed facial data, and the output is the identified emotion data.

[1088] Step 4:

[1089] The server inputs customer data into a generative AI model to generate various personas. For example, attribute data such as a customer's age, gender, purchasing history, and hobbies are input to create a virtual user profile. The input is customer attribute data, and the output is the generated persona.

[1090] Step 5:

[1091] The server then conducts a product-related questionnaire for the generated persona. The questionnaire is designed to collect customer preferences and opinions using natural language processing technology. The input is the persona and the questionnaire items, and the output is the persona's responses to the questionnaire.

[1092] Step 6:

[1093] The server aggregates the collected survey results and generates statistical information. It analyzes the response data of each persona and extracts common trends and outliers. The input is the survey response data, and the output is statistical information.

[1094] Step 7:

[1095] The server analyzes product improvements based on the generated statistical information. It identifies trends and outliers to determine which parts of the product or service should be improved. The input is statistical information, and the output is the identification of areas for improvement.

[1096] Step 8:

[1097] The server displays the analysis results in real time on the user's device. The user's device receives information on product evaluations and areas for improvement and displays it in a format that is intuitively easy for the user to understand. The input is the analysis results, and the output is the data displayed on the user's device.

[1098] 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.

[1099] 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.

[1100] 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.

[1101] [Third embodiment]

[1102] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1103] 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.

[1104] 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).

[1105] 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.

[1106] 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.

[1107] 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).

[1108] 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.

[1109] 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.

[1110] 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.

[1111] 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.

[1112] 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.

[1113] 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."

[1114] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results. The program processing of this system is explained below in natural language.

[1115] Preparing for Persona Generation

[1116] Data collection and classification

[1117] server

[1118] The server collects user data from internal databases and external market research data sources. The collected data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, attribute information such as "female in her 20s, student, hobby is fashion" may be collected.

[1119] Data Preprocessing

[1120] server

[1121] The collected data is cleaned to remove incomplete or inconsistent data, and then converted into a normalized and consistent format.

[1122] Persona Generation

[1123] AI model parameter set

[1124] server

[1125] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[1126] Persona Generation

[1127] server

[1128] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[1129] Product survey design

[1130] Setting up questionnaire items

[1131] User

[1132] A user in the product development department designs a survey item to inquire about specific features of a new product, such as, "What do you think about the design of the new smartwatch?"

[1133] Creating a survey format

[1134] server

[1135] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1136] Conducting a survey using personas

[1137] Sending a survey

[1138] server

[1139] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1140] Response collection

[1141] server

[1142] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1143] Compilation and analysis of survey results

[1144] Data aggregation

[1145] server

[1146] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[1147] Trend and outlier detection

[1148] server

[1149] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1150] Feedback of results

[1151] Results display

[1152] Terminal

[1153] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1154] Improvement suggestions

[1155] server

[1156] Based on the analysis results, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[1157] The present invention is a system that collects diverse viewpoints quickly and economically through the above processing steps, and supports effective product development.

[1158] The processing flow will be explained below.

[1159] Step 1:

[1160] Data collection

[1161] server

[1162] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[1163] Step 2:

[1164] Data Preprocessing

[1165] server

[1166] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes the data to make it consistent, including filling in missing values ​​and standardizing the data format.

[1167] Step 3:

[1168] Preparing the parameter set

[1169] server

[1170] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[1171] Step 4:

[1172] Launching the AI ​​model and generating personas

[1173] server

[1174] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. Specifically, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" is generated.

[1175] Step 5:

[1176] Setting up questionnaire items

[1177] User

[1178] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1179] Step 6:

[1180] Creating a survey format

[1181] server

[1182] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice and open-ended question formats.

[1183] Step 7:

[1184] Sending a survey

[1185] server

[1186] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1187] Step 8:

[1188] Response collection

[1189] server

[1190] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1191] Step 9:

[1192] Data aggregation

[1193] server

[1194] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[1195] Step 10:

[1196] Trend and outlier detection

[1197] server

[1198] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1199] Step 11:

[1200] Results display

[1201] Terminal

[1202] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1203] Step 12:

[1204] Improvement suggestions

[1205] server

[1206] Based on the analysis results, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[1207] Example 1

[1208] 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."

[1209] With conventional questionnaire surveys, it was difficult to collect customer opinions quickly and economically, making it difficult to develop products that reflected diverse perspectives.In addition, the time and effort required to design questionnaire items, collect responses, and analyze them made it difficult to respond quickly to market needs.

[1210] 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.

[1211] In this invention, the server includes means for generating various personas using a generative AI model, means for collecting user data from an internal database or external data sources and classifying and preprocessing the data, means for conducting a product-related survey with the generated personas, means for aggregating the survey results and generating statistical information, means for analyzing product improvements based on the generated statistical information, and means for displaying the analysis results on a user terminal. This makes it possible to quickly and economically collect diverse customer opinions and realize efficient product improvements.

[1212] A "generative AI model" is a model that uses artificial intelligence technology to generate diverse virtual characters (personas) based on specific purposes and requirements.

[1213] A "persona" is a fictional user with specific attributes and characteristics, used to clarify the target users of a product.

[1214] A "server" is a computer system for collecting, processing, storing, analyzing, etc. data.

[1215] "User data" refers to data that includes people's attribute information such as age, gender, occupation, and hobbies.

[1216] "Data source" refers to the information source from which the necessary data is collected, and includes internal databases and external market research databases.

[1217] "Data cleaning" is the process of improving data quality by correcting or removing duplicates, incompleteness, and inconsistencies.

[1218] "Data preprocessing" refers to a series of steps that convert raw data into a format suitable for subsequent analysis or model training.

[1219] A "questionnaire" is a research tool used to gather opinions and evaluations from respondents through specific questions.

[1220] "Natural language processing technology" is a technology that allows computers to analyze, understand, and generate human language.

[1221] "Statistical information" refers to information such as averages, percentages, and trends calculated based on collected data.

[1222] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[1223] "User terminal" refers to a computer or mobile device that a user uses to access the system, display results, and perform operations.

[1224] "Statistical analysis tools" are software and algorithms used to analyze collected data and extract statistics and trends.

[1225] This invention relates to a system that generates various personas using a generative AI model, conducts product-related surveys using those personas, and analyzes areas for improvement of the product based on the survey results. This system operates in cooperation with a server, terminals, and users.

[1226] Data collection and classification

[1227] server

[1228] The server collects user data from internal databases and external market research data sources. Specifically, it retrieves data from external APIs via HTTP requests and stores it in its own database. This data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, it collects attribute information such as "female in her 20s, student, hobby is fashion."

[1229] Data Preprocessing

[1230] server

[1231] The server cleans the collected data, removing incomplete or inconsistent data, removing duplicates and filling in missing data, and normalizing and encoding categorical data to convert it into a consistent format.

[1232] AI model parameter set

[1233] server

[1234] The server creates the parameter set required for the generative AI model based on the preprocessed data. Specifically, it sets the hyperparameters (e.g., learning rate, number of epochs) required for persona generation and prepares a dataset that reflects the characteristics of the target user group.

[1235] Persona Generation

[1236] server

[1237] The server uses the generative AI model to generate various personas. Specifically, it inputs characteristics such as "25-year-old female, hobby is fitness" into the AI ​​model and generates a persona of "25-year-old fitness enthusiast, occupation is marketing."

[1238] Survey design

[1239] User

[1240] A user (a product development department employee) designs a survey to inquire about specific features of a new product, such as "What do you think about the battery life of the new smartwatch?"

[1241] Creating a survey format

[1242] server

[1243] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including formats with multiple choice options (e.g., very good, good, average, bad, very bad) and open-ended questions.

[1244] Sending surveys and collecting responses

[1245] server

[1246] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1247] server

[1248] The server collects the survey responses from each persona and stores the data in a database. Specifically, the responses are classified and stored for each persona.

[1249] Compilation and analysis of survey results

[1250] server

[1251] The server aggregates all survey results and generates statistics, such as response rates and average scores, which are displayed in graphs and tables.

[1252] server

[1253] The server uses statistical analysis tools to detect key trends and outliers in the survey results, such as discovering that users in their 30s are dissatisfied with the prices.

[1254] Display of analysis results and improvement suggestions

[1255] Terminal

[1256] The device displays the aggregated and analyzed results obtained from the server to the user. For example, the dashboard might show a result such as, "The design of the new smartwatch has been well-received, but improvements in battery life are needed."

[1257] server

[1258] Based on the analysis results, the server provides specific improvement suggestions to the product development department. For example, it may generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[1259] Examples and prompts

[1260] Specific examples

[1261] For example, consider the case where the following survey is conducted on a female persona in her 20s.

[1262] Example personas:

[1263] Age: 25

[1264] Gender: Female

[1265] Occupation: Student

[1266] Hobbies: Fashion

[1267] Survey question examples:

[1268] What do you think of the new smartwatch design?

[1269] Very good

[1270] good

[1271] usually

[1272] bad

[1273] Very bad

[1274] Example prompts

[1275] "Please tell us your thoughts on the design of new products. For example, what is your opinion on the design of the new smartwatch?"

[1276] By using the above-mentioned methods, this system can quickly and economically collect diverse customer opinions and realize efficient product improvements.

[1277] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1278] Step 1:

[1279] Data collection and classification

[1280] server

[1281] The server collects user data from an internal database and external market research data sources. This process uses API requests to retrieve data and stores it in the internal database. The input data is user data obtained from external data sources, and the output is data organized and categorized in the internal database. Specifically, the server uses an API key to access the market research site and retrieves data in JSON format.

[1282] Step 2:

[1283] Data Preprocessing

[1284] server

[1285] The server cleans the collected data and removes incomplete or inconsistent data. The input is raw data stored in a database, and the output is preprocessed, consistent data. Specific operations include imputing missing values, removing duplicate data, normalizing data, and encoding categorical data.

[1286] Step 3:

[1287] AI model parameter set

[1288] server

[1289] The server creates the parameter set required for the generative AI model based on the preprocessed data. The input is a normalized dataset, and the output is the hyperparameters to be applied to the AI ​​model. Specifically, it sets parameters such as the learning rate, number of epochs, and batch size to reflect the characteristics of the target user group in the model.

[1290] Step 4:

[1291] Persona Generation

[1292] server

[1293] The server uses a generative AI model to generate various personas. The input is the AI ​​model and parameter set feature data, and the output is a list of generated personas. Specifically, feature data is input into the model to generate a detailed persona such as "25-year-old fitness enthusiast, occupation: marketing."

[1294] Step 5:

[1295] Survey design

[1296] User

[1297] A user (a product development department employee) designs a questionnaire to inquire about specific features of a new product. The input is a list of proposed questions based on the user's needs and product features, and the output is the designed questionnaire items. Specific actions include designing questions such as "What do you think about the battery life of the new smartwatch?"

[1298] Step 6:

[1299] Creating a survey format

[1300] server

[1301] The server automatically generates a questionnaire format based on the questionnaire items set by the user. The input is the questionnaire questions set by the user, and the output is a format that includes options and free-form questions. Specifically, the server automatically arranges the options and completes the format.

[1302] Step 7:

[1303] Sending surveys and collecting responses

[1304] server

[1305] The server distributes questionnaires to the generated personas and collects responses. The input is the generated persona and the questionnaire format, and the output is the response data from each persona. Specifically, it uses natural language processing technology based on the persona's attribute information to generate realistic responses and save them in a response database.

[1306] Step 8:

[1307] Compilation and analysis of survey results

[1308] server

[1309] The server aggregates all survey results and generates statistical information. The input is the collected response data, and the output is the results of statistical information and trend analysis. Specific operations include calculating response ratios and average scores and displaying them in graphs and tables.

[1310] Step 9:

[1311] Display of analysis results and improvement suggestions

[1312] Terminal

[1313] The device displays the aggregated and analyzed results obtained from the server to the user. The input is the statistical information sent from the server, and the output is the result displayed visually to the user. Specifically, the result "The design of the new smartwatch has been well received, but improvements in battery life are needed" is displayed on the dashboard.

[1314] server

[1315] Based on the analysis results, the server provides specific improvement suggestions to the product development department. The input is statistical information and trend analysis results, and the output is specific improvement suggestions. The specific operation is to generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[1316] (Application example 1)

[1317] 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."

[1318] Gathering feedback quickly and efficiently from a diverse user base and identifying areas for product improvement is time-consuming and costly using conventional methods. Online shopping sites, in particular, need a system that can accurately collect diverse user opinions, statistically analyze them, and use them to improve products.

[1319] 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.

[1320] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting product-related questionnaires to the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, and means for setting and sending questionnaire items from a smartphone and displaying responses. This enables the online shopping site to efficiently collect feedback from various perspectives and use it to improve products.

[1321] plaintext

[1322] A "generative AI model" is an artificial intelligence model that learns from data and generates a virtual persona with characteristics that suit a specific purpose.

[1323] A "persona" is a virtual character that represents a specific user group and mimics feedback about a product based on their attributes and behavior.

[1324] A "survey" is a survey in the form of questions that are used to gather responses to specific questions and understand users' opinions and preferences.

[1325] "Aggregation" is the process of organizing and classifying data such as survey results and generating statistical information.

[1326] "Statistical information" is numerical information calculated based on collected data, and indicates trends and characteristics of the data.

[1327] "Areas for improvement in the product" are areas of the product that require improvement in performance, design, specifications, etc., identified based on user feedback.

[1328] A "user terminal" is a device such as a smartphone or computer operated by a user, which generates and receives information.

[1329] "Natural language processing technology" is a technology for analyzing and generating sentences entered by users and answers generated by personas, and is part of artificial intelligence.

[1330] A "smartphone" is a type of mobile phone that can connect to the Internet and use multifunctional applications.

[1331] "Survey items" are specific questions set for users to answer in a questionnaire.

[1332] The system of this invention generates various personas related to an online shopping site, conducts product-related surveys with these personas, and aggregates and analyzes the results to identify areas for improvement in the products. An embodiment of this system is shown below.

[1333] System Configuration

[1334] The system consists of a server, user terminals, and smartphone devices, and uses software such as a database management system (MySQL), a generative AI model (OpenAI GPT-4), statistical analysis tools (Python Pandas, NumPy), and natural language processing technology (SpaCy, NLTK).

[1335] Program processing

[1336] 1. Data collection and classification

[1337] The server collects user data from internal databases and external market research data sources. This data is categorized based on attributes such as age, gender, occupation, and hobbies. As a result, data with diverse user characteristics is collected.

[1338] 2. Data Preprocessing

[1339] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes and converts it into a consistent format, preparing the data for input into the generative AI model.

[1340] 3. Persona Generation

[1341] The server sets the parameters of the generative AI model based on the preprocessed data, which generates a variety of personas that represent specific user demographics. For example, a persona might be generated such as "35 years old, male, hobby is fishing, occupation is engineer."

[1342] 4. Setting up the questionnaire items and creating the format

[1343] The user (a product development department employee) sets up a questionnaire about specific features of a new product. For example, they set up a question such as, "What do you think about the design of the new smartwatch?" The server automatically generates a questionnaire format based on the questionnaire items set by the user.

[1344] 5. Conducting surveys and collecting responses

[1345] The server distributes questionnaires to the generated personas. Each persona generates realistic answers using natural language processing technology. For example, a response such as "The design of this smartwatch is sophisticated, and I want to share it with my friends on social media" may be generated. The server collects these responses and stores them in a database.

[1346] 6. Calculation and analysis of results

[1347] The server aggregates all survey results and generates statistics. For example, the aggregated result may be "80% of personas are satisfied with the new features." Furthermore, statistical analysis tools are used to detect trends and outliers in the survey results.

[1348] 7. Feedback of results

[1349] The server then displays the results of the analysis on the user's device. For example, it may display a result such as, "The design is highly rated, but battery life needs improvement." It also provides specific improvement suggestions to the product development department based on the identified areas for improvement.

[1350] Prompt Sentence Examples

[1351] "What do female users in their 20s think about new smartphone designs?"

[1352] This system makes it possible to efficiently collect feedback from a variety of perspectives and quickly and accurately improve the products offered on the online shopping site, thereby improving customer satisfaction and supporting the development of competitive products.

[1353] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1354] plaintext

[1355] Step 1: Data collection and classification

[1356] The server collects user data from internal databases and external market research data sources. The input includes user attribute data such as age, gender, occupation, and hobbies. Based on this data, the server categorizes the data based on user characteristics. The output is user data by category.

[1357] Step 2: Data Preprocessing

[1358] The server cleans the collected data, removing incomplete or inconsistent data, then normalizes and converts the data into a consistent format. The input includes classified user data. The output is preprocessed, cleaned data.

[1359] Step 3: Generate personas

[1360] The server sets the parameters of the generative AI model based on the preprocessed data, which generates diverse personas that represent specific user demographics. The inputs include the preprocessed data and model parameters. The output is a diverse set of generated personas.

[1361] Step 4: Setting up the questionnaire and creating the format

[1362] The user sets up questionnaire items about specific features of the new product. The server automatically generates a questionnaire format based on the questionnaire items set up by the user. The input includes the questionnaire items set up by the user. The output is the generated questionnaire format.

[1363] Step 5: Conduct the survey and collect responses

[1364] The server distributes surveys to the generated personas. Each persona generates realistic responses using natural language processing techniques. The inputs include the generated personas and the survey format. As an output, the personas' survey responses are collected and stored in a database.

[1365] Step 6: Compile and analyze the results

[1366] The server aggregates all survey results and generates statistics. It also uses statistical analysis tools to detect trends and outliers in the survey results. Inputs include persona survey responses. Outputs include statistics and trend analysis results.

[1367] Step 7: Feedback on results

[1368] The server displays the aggregated and analyzed results on the user's device. Based on the analysis results, it provides specific improvement proposals to the product development department. The input includes statistical information and trend analysis results. The output is improvement proposals that are displayed on the user's device.

[1369] 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.

[1370] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results and emotion recognition. The program processing of this system is explained below in natural language.

[1371] Preparing for Persona Generation

[1372] Data collection and classification

[1373] server

[1374] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, attribute information such as "female in her 30s, working in finance, running as a hobby" is collected.

[1375] Data Preprocessing

[1376] server

[1377] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[1378] Persona Generation

[1379] AI model parameter set

[1380] server

[1381] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[1382] Persona Generation

[1383] server

[1384] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[1385] Product survey design

[1386] Setting up questionnaire items

[1387] User

[1388] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1389] Creating a survey format

[1390] server

[1391] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1392] Conducting a survey using personas

[1393] Sending a survey

[1394] server

[1395] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1396] Response collection

[1397] server

[1398] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1399] Analysis by emotion engine

[1400] Conducting sentiment analysis

[1401] server

[1402] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[1403] Aggregating emotion data

[1404] server

[1405] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[1406] Aggregation and analysis of survey results and sentiment data

[1407] Data aggregation

[1408] server

[1409] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[1410] Trend and outlier detection

[1411] server

[1412] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1413] Feedback of results

[1414] Results display

[1415] Terminal

[1416] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1417] Emotional Feedback

[1418] Terminal

[1419] Based on the analysis results of the emotion engine, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[1420] Improvement suggestions

[1421] server

[1422] Based on the analysis results and sentiment data, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[1423] The present invention is a system that collects a variety of viewpoints and emotion data quickly and economically through the above processing steps, and supports effective product development.

[1424] The processing flow will be explained below.

[1425] Step 1:

[1426] Data collection

[1427] server

[1428] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[1429] Step 2:

[1430] Data Preprocessing

[1431] server

[1432] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[1433] Step 3:

[1434] Preparing the parameter set

[1435] server

[1436] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[1437] Step 4:

[1438] Launching the AI ​​model and generating personas

[1439] server

[1440] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. For example, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" can be generated.

[1441] Step 5:

[1442] Setting up questionnaire items

[1443] User

[1444] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1445] Step 6:

[1446] Creating a survey format

[1447] server

[1448] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1449] Step 7:

[1450] Sending a survey

[1451] server

[1452] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1453] Step 8:

[1454] Response collection

[1455] server

[1456] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1457] Step 9:

[1458] Conducting sentiment analysis

[1459] server

[1460] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[1461] Step 10:

[1462] Aggregating emotion data

[1463] server

[1464] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[1465] Step 11:

[1466] Data aggregation

[1467] server

[1468] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[1469] Step 12:

[1470] Trend and outlier detection

[1471] server

[1472] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1473] Step 13:

[1474] Results display

[1475] Terminal

[1476] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1477] Step 14:

[1478] Emotional Feedback

[1479] Terminal

[1480] Based on the results of the emotion analysis obtained from the server, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[1481] Step 15:

[1482] Improvement suggestions

[1483] server

[1484] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[1485] Example 2

[1486] 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."

[1487] Conventional product improvement support systems have struggled to quickly and efficiently collect diverse user perspectives and emotional data and accurately analyze product improvements. Furthermore, collecting survey data and providing feedback through emotional analysis is a time-consuming and labor-intensive process. To address these challenges, the present invention provides a system that uses a generative AI model to generate diverse personas and efficiently collect and analyze survey data.

[1488] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user data from an internal database and an external data source and cleaning and normalizing the data, means for creating a parameter set for a generative AI model based on the collected and preprocessed data and generating a persona, means for conducting a product-related survey with the generated persona, means for collecting survey results and emotional data and storing the same in a database, means for generating statistical information based on the collected survey results and emotional data and detecting trends and outliers, means for analyzing product improvements based on the generated statistical information and emotional data, and means for displaying the analysis results on a user terminal. This enables rapid and efficient collection and analysis of diverse perspectives and emotional data.

[1489] An "internal database" is a collection of data managed within the system, and is a storage device for storing various user information, past survey results, etc.

[1490] "External data sources" are information sources for obtaining data from outside the system, including market research data and publicly available information.

[1491] "User data" refers to data that includes the attribute information of a specific user, such as age, gender, occupation, hobbies, and purchasing history.

[1492] "Cleaning" is the process of removing incomplete or inconsistent data from collected data and preparing it in a format suitable for data analysis.

[1493] "Normalization" is the process of converting collected data into a consistent format and maintaining consistency between the data.

[1494] A "generative AI model" is an algorithm that uses artificial intelligence (AI) technology to generate diverse personas, taking a specific set of parameters as input.

[1495] A "parameter set" refers to a set of data or conditions that are input into a generative AI model, which then generates a specific persona based on these parameters.

[1496] A "persona" is a virtual character that simulates a specific user profile and has attribute information such as age, gender, occupation, and hobbies.

[1497] A "questionnaire" is a survey method used to solicit opinions on specific questions, and is conducted in a multiple choice or free description format.

[1498] "Emotional data" refers to emotional information analyzed based on questionnaire responses, and includes emotional states such as satisfaction and dissatisfaction.

[1499] "Statistical information" refers to data that shows aggregated results and trends calculated from collected data, and includes various trends and outliers.

[1500] "Trend" refers to a general pattern of tendency or fluctuation found in a large amount of collected data.

[1501] An "outlier" is data that shows significantly different values ​​or patterns compared to other data, and may indicate a particular issue or problem.

[1502] This invention is a system that generates various personas using a generative AI model, conducts product-related surveys using these personas, and analyzes product improvements based on the results and sentiment analysis. The main roles of this system are played by the server, users, and terminals.

[1503] 1. Data collection and classification

[1504] Server Data Collection

[1505] The server collects user data from an internal database and external data sources. The internal database includes existing customer information, purchase history, survey results, etc. Market research data and public data are also used as external data sources. The server uses a data collection API to obtain attribute information such as "female in her 30s, working in finance, hobby is running."

[1506] Data classification process

[1507] The server categorizes the collected data into categories such as age, gender, occupation, hobbies, and purchasing history. This categorization creates groups of users with specific attributes, which are used for analysis and persona generation.

[1508] 2. Data Preprocessing

[1509] Data Cleaning

[1510] The server removes incomplete or inconsistent data from the collected data, for example, deleting records with missing occupation data, and normalizing data formats if they are not standardized.

[1511] Data normalization

[1512] The server converts the data into a consistent format, for example by aligning all age data to "decades."

[1513] 3. Parameter set of the AI ​​model

[1514] Creating a Parameter Set

[1515] The server uses the preprocessed data to create a parameter set for the generative AI model, which includes data corresponding to a specific user profile, such as "35 years old, male, engineer, fishing hobby."

[1516] 4. Persona Generation

[1517] Persona generation process

[1518] The server uses a generative AI model to generate various personas, resulting in a specific persona such as "35 years old, male, engineer, fishing hobby." The generated personas are stored in an internal database.

[1519] 5. Setting up the questionnaire items

[1520] Survey design

[1521] The user (a product development department employee) sets specific questions about the product, for example, designing a questionnaire item such as "What do you think about the design of the new smartwatch?"

[1522] 6. Creating a survey format

[1523] Generate a questionnaire format

[1524] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice questions and open-ended questions, and stores the format in electronic format.

[1525] 7. Conduct a survey using personas

[1526] Survey distribution processing

[1527] The server then sends a questionnaire to the persona, which uses natural language processing technology to generate realistic responses based on its attributes. For example, a scenario might be generated in which a 35-year-old male engineer gives positive feedback on the design of a new smartwatch.

[1528] Collecting responses

[1529] The server collects the persona's responses to the questionnaire and stores them in a database for later analysis.

[1530] 8. Conducting sentiment analysis

[1531] Emotional Data Analysis

[1532] The server performs sentiment analysis based on the collected survey responses, using an emotion engine to identify emotional data such as "the persona is satisfied based on the responses" or "they are surprised."

[1533] 9. Data Collection and Analysis

[1534] Data aggregation process

[1535] The server statistically aggregates the survey results and sentiment data, generating aggregated results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design."

[1536] Trend and outlier detection

[1537] The server uses the statistics to detect key trends and outliers, such as finding that certain age groups are unhappy with prices or that certain interests are responding positively to new features.

[1538] 10. Feedback of results

[1539] Display the results on the user's device

[1540] The server displays the analysis results on the user's device, where the user can see results such as "Male users in their 40s are very satisfied with the new features" and "They rate the design highly, but are often dissatisfied with the battery life."

[1541] 11. Improvement suggestions

[1542] Proposal for improvement

[1543] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life" or "improve the finer details of the design," which are displayed on the user's device.

[1544] Through the above processing, this system quickly and efficiently collects diverse viewpoints and emotional data, supporting effective product development.

[1545] Prompt Sentence Examples

[1546] Below is an example of an input prompt for a generative AI model.

[1547] Generate feedback for the design of a new smartwatch based on the demographic information "female in her 30s, working in finance, running as a hobby."

[1548] Example of feedback: Positive about the design, but dissatisfied with the battery life.

[1549] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1550] Step 1: Data collection and classification

[1551] The server collects user data from an internal database and external data sources. As input, customer information from the internal database and external market research data are provided to the server. The server obtains this data via API and stores the data obtained from each data source together. As output, user data categorized by age, gender, occupation, hobbies, purchase history, etc. is obtained.

[1552] Step 2: Data cleaning

[1553] The server removes incomplete or inconsistent data from the data collected in step 1. User data collected in various formats is provided to the server as input. The server performs processing such as deleting records with missing values ​​and imputing some missing values. The output is cleaned user data.

[1554] Step 3: Data normalization

[1555] The server normalizes the cleaned data and converts it into a consistent format. As input, the cleaned user data is provided to the server. For example, the server converts age data into a unified format (e.g., decade units) to align all data formats. As output, the server obtains user data in a consistent format.

[1556] Step 4: Creating a parameter set for the AI ​​model

[1557] The server creates a parameter set for the generative AI model based on the preprocessed data. Normalized user data is provided to the server as input. The server extracts parameters corresponding to each persona (e.g., "35 years old, male, engineer, hobby is fishing") and creates a parameter set to input into the generative AI model. The parameter set for the generative AI model is obtained as output.

[1558] Step 5: Persona Generation

[1559] The server uses a generative AI model to generate various personas. A set of parameters is input to the generative AI model. The server runs the generative AI model to generate various personas (e.g., "35 years old, male, engineer, hobby is fishing") and stores them in an internal database. The generated personas are obtained as output.

[1560] Step 6: Setting up the survey items

[1561] The user sets specific questions about the product. As input, the user is provided with basic information for setting up the questionnaire items. The user designs specific questionnaire items such as "What do you think about the design of the new smartwatch?". As output, the set questionnaire items are obtained.

[1562] Step 7: Generate the survey format

[1563] The server automatically generates a questionnaire based on the questionnaire items set by the user. As input, the set questionnaire items are provided to the server. The server creates an electronic format containing multiple choice and open-ended questions. As output, the generated questionnaire is obtained.

[1564] Step 8: Send out the survey

[1565] The server distributes the questionnaire to the generated personas. As input, the server is provided with the questionnaire format and persona data. The server sends the questionnaire to each persona and collects the responses. As output, the persona's questionnaire responses are obtained.

[1566] Step 9: Generate and collect answers

[1567] The server generates the answers of the generated persona using natural language processing technology. As input, the persona data and the questionnaire form are provided to the server. The server uses natural language processing technology to generate realistic answers and stores them in a database. As output, the server obtains the questionnaire answers generated by natural language processing.

[1568] Step 10: Sentiment Analysis

[1569] The server performs sentiment analysis based on the collected survey responses. The persona's survey responses are provided to the server as input. The server uses an emotion engine to identify emotions such as "the persona is satisfied" or "surprised" based on the responses. The analyzed emotional data is obtained as output.

[1570] Step 11: Data collection and analysis

[1571] The server statistically aggregates the survey results and emotional data. The survey responses and emotional data are provided to the server as input. The server statistically analyzes this data and obtains results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design." The output is aggregated statistical information.

[1572] Step 12: Detect trends and outliers

[1573] The server detects key trends and outliers from the collected data. As input, aggregated statistics are provided to the server. The server uses statistical analysis tools to determine results such as "certain age groups are dissatisfied with the price" or "certain hobbies are responding positively to new features." As output, the detected trends and outliers are provided.

[1574] Step 13: Feedback and display of results

[1575] The server displays the analysis results on the user's device. The analysis results and emotional data are provided to the server as input. The server sends this to the user's device, which then displays the results, such as "Male users in their 40s are very satisfied with the new function." The analysis results are displayed on the user's device as output.

[1576] Step 14: Present improvement suggestions

[1577] The server provides specific improvement suggestions to the product development department based on the analysis results and emotion data. Detailed analysis results and emotion data are provided to the server as input. The server generates specific suggestions, such as "focus on improving battery life" or "improve the finer details of the design," and sends them to the user's device. Specific improvement suggestions are obtained as output.

[1578] (Application example 2)

[1579] 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."

[1580] While conventional persona generation systems can analyze product improvements based on statistical information, they have the problem of being unable to obtain information that reflects actual customer behavior and emotions in real time. This can sometimes make it difficult to accurately evaluate the effects of product improvements, creating a need for more accurate and immediate data collection and analysis.

[1581] 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.

[1582] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting a product-related questionnaire for the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, means for capturing customer behavior and facial expressions using smart glasses, and means for analyzing customer emotions in real time based on the captured data. This makes it possible to collect data reflecting customer behavior and emotions in real time and instantly identify areas for improvement in products and services.

[1583] A "generative AI model" is an artificial intelligence technology used to automatically generate diverse personas.

[1584] A "persona" is a virtual user profile with specific attributes and behavioral patterns.

[1585] A "questionnaire" is a method for asking questions to subjects and collecting their responses.

[1586] "Statistical information" is information presented in the form of numbers, graphs, etc. obtained by aggregating and analyzing multiple data.

[1587] "Analysis" is the process of evaluating collected data and extracting meaningful information.

[1588] A "user terminal" is a device such as a computer or smartphone that is used to display results and data.

[1589] "Smart glasses" are wearable devices that the wearer uses to display external information and collect data through sensors.

[1590] "Behavioral capture" is a method of observing customer behavior and recording it as digital information.

[1591] "Facial expression capture" is a method of observing a customer's facial expressions and recording them digitally.

[1592] "Emotion analysis" is a technology that identifies and evaluates the emotions expressed by customers based on captured facial expression data.

[1593] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[1594] This invention is a system that uses smart glasses in brick-and-mortar stores to analyze customer behavior and facial expressions in real time and utilizes generative AI models based on that data. By analyzing customer emotions and behavior, this system can instantly identify product evaluations and in-store improvements.

[1595] Hardware and Software Configuration

[1596] To implement this system, the following hardware and software are used.

[1597] Hardware

[1598] Smart glasses: Wearable devices for capturing customer behavior and facial expressions.

[1599] Server: Collects, analyzes and manages data.

[1600] User terminal: The device that displays and manipulates data (e.g., PC, smartphone).

[1601] software

[1602] Generative AI models: Used to generate diverse personas.

[1603] Sentiment analysis engine: Technology that analyzes customer sentiment in real time based on captured data.

[1604] Natural language processing techniques: Used to generate answers for personas.

[1605] Specific operating procedures for the system

[1606] 1. Data collection: The server captures customer behavior and facial expression data in real time through the smart glasses.

[1607] 2. Data Preprocessing: The server preprocesses the captured data and converts it into a format that can be analyzed by the sentiment analysis engine.

[1608] 3. Sentiment analysis: The server uses a sentiment analysis engine to analyze customer sentiment in real time.

[1609] 4. Persona generation: The server uses a generative AI model to generate a persona with specific attributes and behavioral patterns.

[1610] 5. Conducting a survey: The server conducts a survey about the product to the generated personas and collects responses.

[1611] 6. Response analysis: The server aggregates the collected survey results and generates statistical information.

[1612] 7. Analysis of Improvement Points: The server analyzes the improvement points of the product based on the generated statistical information.

[1613] 8. Displaying results: The server displays the analysis results on the user's terminal in real time.

[1614] Specific examples

[1615] For example, when a customer in a store picks up a new smartwatch, their behavior and facial expression data are captured by the smart glasses. If the emotion analysis engine detects emotions such as "surprise" or "satisfaction," the server can instantly generate information such as "this smartwatch is popular among men in their 20s and 30s" and display it on the user's device.

[1616] Prompt Sentence Examples

[1617] Below are examples of prompts that can be used to conduct a survey on the generated personas.

[1618] Age: 30, Gender: Female, Purchasing history: Unknown, Hobbies: Fashion

[1619] In this way, the system of the present invention makes it possible to analyze customer sentiment and behavior in real time, evaluate products, and identify areas for improvement.

[1620] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1621] Step 1:

[1622] The server captures customer behavior and facial expression data in real time through the smart glasses. When customers browse and pick up products in the store, the video data is collected by the camera in the smart glasses. The input is the video data, and the output is the captured raw data.

[1623] Step 2:

[1624] The server preprocesses the captured video data and converts it into a format that can be analyzed by the emotion analysis engine. Specifically, it converts the video data to grayscale, trims unnecessary parts, and extracts facial features. The input is the captured video data, and the output is preprocessed facial data.

[1625] Step 3:

[1626] The server performs emotion analysis using the preprocessed facial data. It uses an emotion analysis engine to identify emotions such as happiness, surprise, sadness, etc. based on the data. The input is the preprocessed facial data, and the output is the identified emotion data.

[1627] Step 4:

[1628] The server inputs customer data into a generative AI model to generate various personas. For example, attribute data such as a customer's age, gender, purchasing history, and hobbies are input to create a virtual user profile. The input is customer attribute data, and the output is the generated persona.

[1629] Step 5:

[1630] The server then conducts a product-related questionnaire for the generated persona. The questionnaire is designed to collect customer preferences and opinions using natural language processing technology. The input is the persona and the questionnaire items, and the output is the persona's responses to the questionnaire.

[1631] Step 6:

[1632] The server aggregates the collected survey results and generates statistical information. It analyzes the response data of each persona and extracts common trends and outliers. The input is the survey response data, and the output is statistical information.

[1633] Step 7:

[1634] The server analyzes product improvements based on the generated statistical information. It identifies trends and outliers to determine which parts of the product or service should be improved. The input is statistical information, and the output is the identification of areas for improvement.

[1635] Step 8:

[1636] The server displays the analysis results in real time on the user's device. The user's device receives information on product evaluations and areas for improvement and displays it in a format that is intuitively easy for the user to understand. The input is the analysis results, and the output is the data displayed on the user's device.

[1637] 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.

[1638] 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.

[1639] 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.

[1640] [Fourth embodiment]

[1641] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1642] 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.

[1643] 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).

[1644] 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.

[1645] 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.

[1646] 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).

[1647] 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.

[1648] 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.

[1649] 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.

[1650] 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.

[1651] 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.

[1652] 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.

[1653] 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."

[1654] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results. The program processing of this system is explained below in natural language.

[1655] Preparing for Persona Generation

[1656] Data collection and classification

[1657] server

[1658] The server collects user data from internal databases and external market research data sources. The collected data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, attribute information such as "female in her 20s, student, hobby is fashion" may be collected.

[1659] Data Preprocessing

[1660] server

[1661] The collected data is cleaned to remove incomplete or inconsistent data, and then converted into a normalized and consistent format.

[1662] Persona Generation

[1663] AI model parameter set

[1664] server

[1665] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[1666] Persona Generation

[1667] server

[1668] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[1669] Product survey design

[1670] Setting up questionnaire items

[1671] User

[1672] A user in the product development department designs a survey item to inquire about specific features of a new product, such as, "What do you think about the design of the new smartwatch?"

[1673] Creating a survey format

[1674] server

[1675] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1676] Conducting a survey using personas

[1677] Sending a survey

[1678] server

[1679] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1680] Response collection

[1681] server

[1682] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1683] Compilation and analysis of survey results

[1684] Data aggregation

[1685] server

[1686] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[1687] Trend and outlier detection

[1688] server

[1689] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1690] Feedback of results

[1691] Results display

[1692] Terminal

[1693] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1694] Improvement suggestions

[1695] server

[1696] Based on the analysis results, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[1697] The present invention is a system that collects diverse viewpoints quickly and economically through the above processing steps, and supports effective product development.

[1698] The processing flow will be explained below.

[1699] Step 1:

[1700] Data collection

[1701] server

[1702] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[1703] Step 2:

[1704] Data Preprocessing

[1705] server

[1706] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes the data to make it consistent, including filling in missing values ​​and standardizing the data format.

[1707] Step 3:

[1708] Preparing the parameter set

[1709] server

[1710] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[1711] Step 4:

[1712] Launching the AI ​​model and generating personas

[1713] server

[1714] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. Specifically, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" is generated.

[1715] Step 5:

[1716] Setting up questionnaire items

[1717] User

[1718] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1719] Step 6:

[1720] Creating a survey format

[1721] server

[1722] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice and open-ended question formats.

[1723] Step 7:

[1724] Sending a survey

[1725] server

[1726] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1727] Step 8:

[1728] Response collection

[1729] server

[1730] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1731] Step 9:

[1732] Data aggregation

[1733] server

[1734] All survey results are aggregated to generate statistics, such as "80% of personas are satisfied with the new features."

[1735] Step 10:

[1736] Trend and outlier detection

[1737] server

[1738] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1739] Step 11:

[1740] Results display

[1741] Terminal

[1742] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1743] Step 12:

[1744] Improvement suggestions

[1745] server

[1746] Based on the analysis results, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[1747] Example 1

[1748] 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."

[1749] With conventional questionnaire surveys, it was difficult to collect customer opinions quickly and economically, making it difficult to develop products that reflected diverse perspectives.In addition, the time and effort required to design questionnaire items, collect responses, and analyze them made it difficult to respond quickly to market needs.

[1750] 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.

[1751] In this invention, the server includes means for generating various personas using a generative AI model, means for collecting user data from an internal database or external data sources and classifying and preprocessing the data, means for conducting a product-related survey with the generated personas, means for aggregating the survey results and generating statistical information, means for analyzing product improvements based on the generated statistical information, and means for displaying the analysis results on a user terminal. This makes it possible to quickly and economically collect diverse customer opinions and realize efficient product improvements.

[1752] A "generative AI model" is a model that uses artificial intelligence technology to generate diverse virtual characters (personas) based on specific purposes and requirements.

[1753] A "persona" is a fictional user with specific attributes and characteristics, used to clarify the target users of a product.

[1754] A "server" is a computer system for collecting, processing, storing, analyzing, etc. data.

[1755] "User data" refers to data that includes people's attribute information such as age, gender, occupation, and hobbies.

[1756] "Data source" refers to the information source from which the necessary data is collected, and includes internal databases and external market research databases.

[1757] "Data cleaning" is the process of improving data quality by correcting or removing duplicates, incompleteness, and inconsistencies.

[1758] "Data preprocessing" refers to a series of steps that convert raw data into a format suitable for subsequent analysis or model training.

[1759] A "questionnaire" is a research tool used to gather opinions and evaluations from respondents through specific questions.

[1760] "Natural language processing technology" is a technology that allows computers to analyze, understand, and generate human language.

[1761] "Statistical information" refers to information such as averages, percentages, and trends calculated based on collected data.

[1762] A "database" is a system for efficiently storing, searching, and managing large amounts of data.

[1763] "User terminal" refers to a computer or mobile device that a user uses to access the system, display results, and perform operations.

[1764] "Statistical analysis tools" are software and algorithms used to analyze collected data and extract statistics and trends.

[1765] This invention relates to a system that generates various personas using a generative AI model, conducts product-related surveys using those personas, and analyzes areas for improvement of the product based on the survey results. This system operates in cooperation with a server, terminals, and users.

[1766] Data collection and classification

[1767] server

[1768] The server collects user data from internal databases and external market research data sources. Specifically, it retrieves data from external APIs via HTTP requests and stores it in its own database. This data is categorized based on attributes such as age, gender, occupation, and hobbies. For example, it collects attribute information such as "female in her 20s, student, hobby is fashion."

[1769] Data Preprocessing

[1770] server

[1771] The server cleans the collected data, removing incomplete or inconsistent data, removing duplicates and filling in missing data, and normalizing and encoding categorical data to convert it into a consistent format.

[1772] AI model parameter set

[1773] server

[1774] The server creates the parameter set required for the generative AI model based on the preprocessed data. Specifically, it sets the hyperparameters (e.g., learning rate, number of epochs) required for persona generation and prepares a dataset that reflects the characteristics of the target user group.

[1775] Persona Generation

[1776] server

[1777] The server uses the generative AI model to generate various personas. Specifically, it inputs characteristics such as "25-year-old female, hobby is fitness" into the AI ​​model and generates a persona of "25-year-old fitness enthusiast, occupation is marketing."

[1778] Survey design

[1779] User

[1780] A user (a product development department employee) designs a survey to inquire about specific features of a new product, such as "What do you think about the battery life of the new smartwatch?"

[1781] Creating a survey format

[1782] server

[1783] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including formats with multiple choice options (e.g., very good, good, average, bad, very bad) and open-ended questions.

[1784] Sending surveys and collecting responses

[1785] server

[1786] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1787] server

[1788] The server collects the survey responses from each persona and stores the data in a database. Specifically, the responses are classified and stored for each persona.

[1789] Compilation and analysis of survey results

[1790] server

[1791] The server aggregates all survey results and generates statistics, such as response rates and average scores, which are displayed in graphs and tables.

[1792] server

[1793] The server uses statistical analysis tools to detect key trends and outliers in the survey results, such as discovering that users in their 30s are dissatisfied with the prices.

[1794] Display of analysis results and improvement suggestions

[1795] Terminal

[1796] The device displays the aggregated and analyzed results obtained from the server to the user. For example, the dashboard might show a result such as, "The design of the new smartwatch has been well-received, but improvements in battery life are needed."

[1797] server

[1798] Based on the analysis results, the server provides specific improvement suggestions to the product development department. For example, it may generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[1799] Examples and prompts

[1800] Specific examples

[1801] For example, consider the case where the following survey is conducted on a female persona in her 20s.

[1802] Example personas:

[1803] Age: 25

[1804] Gender: Female

[1805] Occupation: Student

[1806] Hobbies: Fashion

[1807] Survey question examples:

[1808] What do you think of the new smartwatch design?

[1809] Very good

[1810] good

[1811] usually

[1812] bad

[1813] Very bad

[1814] Example prompts

[1815] "Please tell us your thoughts on the design of new products. For example, what is your opinion on the design of the new smartwatch?"

[1816] By using the above-mentioned methods, this system can quickly and economically collect diverse customer opinions and realize efficient product improvements.

[1817] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1818] Step 1:

[1819] Data collection and classification

[1820] server

[1821] The server collects user data from an internal database and external market research data sources. This process uses API requests to retrieve data and stores it in the internal database. The input data is user data obtained from external data sources, and the output is data organized and categorized in the internal database. Specifically, the server uses an API key to access the market research site and retrieves data in JSON format.

[1822] Step 2:

[1823] Data Preprocessing

[1824] server

[1825] The server cleans the collected data and removes incomplete or inconsistent data. The input is raw data stored in a database, and the output is preprocessed, consistent data. Specific operations include imputing missing values, removing duplicate data, normalizing data, and encoding categorical data.

[1826] Step 3:

[1827] AI model parameter set

[1828] server

[1829] The server creates the parameter set required for the generative AI model based on the preprocessed data. The input is a normalized dataset, and the output is the hyperparameters to be applied to the AI ​​model. Specifically, it sets parameters such as the learning rate, number of epochs, and batch size to reflect the characteristics of the target user group in the model.

[1830] Step 4:

[1831] Persona Generation

[1832] server

[1833] The server uses a generative AI model to generate various personas. The input is the AI ​​model and parameter set feature data, and the output is a list of generated personas. Specifically, feature data is input into the model to generate a detailed persona such as "25-year-old fitness enthusiast, occupation: marketing."

[1834] Step 5:

[1835] Survey design

[1836] User

[1837] A user (a product development department employee) designs a questionnaire to inquire about specific features of a new product. The input is a list of proposed questions based on the user's needs and product features, and the output is the designed questionnaire items. Specific actions include designing questions such as "What do you think about the battery life of the new smartwatch?"

[1838] Step 6:

[1839] Creating a survey format

[1840] server

[1841] The server automatically generates a questionnaire format based on the questionnaire items set by the user. The input is the questionnaire questions set by the user, and the output is a format that includes options and free-form questions. Specifically, the server automatically arranges the options and completes the format.

[1842] Step 7:

[1843] Sending surveys and collecting responses

[1844] server

[1845] The server distributes questionnaires to the generated personas and collects responses. The input is the generated persona and the questionnaire format, and the output is the response data from each persona. Specifically, it uses natural language processing technology based on the persona's attribute information to generate realistic responses and save them in a response database.

[1846] Step 8:

[1847] Compilation and analysis of survey results

[1848] server

[1849] The server aggregates all survey results and generates statistical information. The input is the collected response data, and the output is the results of statistical information and trend analysis. Specific operations include calculating response ratios and average scores and displaying them in graphs and tables.

[1850] Step 9:

[1851] Display of analysis results and improvement suggestions

[1852] Terminal

[1853] The device displays the aggregated and analyzed results obtained from the server to the user. The input is the statistical information sent from the server, and the output is the result displayed visually to the user. Specifically, the result "The design of the new smartwatch has been well received, but improvements in battery life are needed" is displayed on the dashboard.

[1854] server

[1855] Based on the analysis results, the server provides specific improvement suggestions to the product development department. The input is statistical information and trend analysis results, and the output is specific improvement suggestions. The specific operation is to generate an improvement suggestion such as "focus on extending battery life" and notify the user's device.

[1856] (Application example 1)

[1857] 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."

[1858] Gathering feedback quickly and efficiently from a diverse user base and identifying areas for product improvement is time-consuming and costly using conventional methods. Online shopping sites, in particular, need a system that can accurately collect diverse user opinions, statistically analyze them, and use them to improve products.

[1859] 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.

[1860] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting product-related questionnaires to the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, and means for setting and sending questionnaire items from a smartphone and displaying responses. This enables the online shopping site to efficiently collect feedback from various perspectives and use it to improve products.

[1861] plaintext

[1862] A "generative AI model" is an artificial intelligence model that learns from data and generates a virtual persona with characteristics that suit a specific purpose.

[1863] A "persona" is a virtual character that represents a specific user group and mimics feedback about a product based on their attributes and behavior.

[1864] A "survey" is a survey in the form of questions that are used to gather responses to specific questions and understand users' opinions and preferences.

[1865] "Aggregation" is the process of organizing and classifying data such as survey results and generating statistical information.

[1866] "Statistical information" is numerical information calculated based on collected data, and indicates trends and characteristics of the data.

[1867] "Areas for improvement in the product" are areas of the product that require improvement in performance, design, specifications, etc., identified based on user feedback.

[1868] A "user terminal" is a device such as a smartphone or computer operated by a user, which generates and receives information.

[1869] "Natural language processing technology" is a technology for analyzing and generating sentences entered by users and answers generated by personas, and is part of artificial intelligence.

[1870] A "smartphone" is a type of mobile phone that can connect to the Internet and use multifunctional applications.

[1871] "Survey items" are specific questions set for users to answer in a questionnaire.

[1872] The system of this invention generates various personas related to an online shopping site, conducts product-related surveys with these personas, and aggregates and analyzes the results to identify areas for improvement in the products. An embodiment of this system is shown below.

[1873] System Configuration

[1874] The system consists of a server, user terminals, and smartphone devices, and uses software such as a database management system (MySQL), a generative AI model (OpenAI GPT-4), statistical analysis tools (Python Pandas, NumPy), and natural language processing technology (SpaCy, NLTK).

[1875] Program processing

[1876] 1. Data collection and classification

[1877] The server collects user data from internal databases and external market research data sources. This data is categorized based on attributes such as age, gender, occupation, and hobbies. As a result, data with diverse user characteristics is collected.

[1878] 2. Data Preprocessing

[1879] The server cleans the collected data, removing incomplete or inconsistent data, and normalizes and converts it into a consistent format, preparing the data for input into the generative AI model.

[1880] 3. Persona Generation

[1881] The server sets the parameters of the generative AI model based on the preprocessed data, which generates a variety of personas that represent specific user demographics. For example, a persona might be generated such as "35 years old, male, hobby is fishing, occupation is engineer."

[1882] 4. Setting up the questionnaire items and creating the format

[1883] The user (a product development department employee) sets up a questionnaire about specific features of a new product. For example, they set up a question such as, "What do you think about the design of the new smartwatch?" The server automatically generates a questionnaire format based on the questionnaire items set by the user.

[1884] 5. Conducting surveys and collecting responses

[1885] The server distributes questionnaires to the generated personas. Each persona generates realistic answers using natural language processing technology. For example, a response such as "The design of this smartwatch is sophisticated, and I want to share it with my friends on social media" may be generated. The server collects these responses and stores them in a database.

[1886] 6. Calculation and analysis of results

[1887] The server aggregates all survey results and generates statistics. For example, the aggregated result may be "80% of personas are satisfied with the new features." Furthermore, statistical analysis tools are used to detect trends and outliers in the survey results.

[1888] 7. Feedback of results

[1889] The server then displays the results of the analysis on the user's device. For example, it may display a result such as, "The design is highly rated, but battery life needs improvement." It also provides specific improvement suggestions to the product development department based on the identified areas for improvement.

[1890] Prompt Sentence Examples

[1891] "What do female users in their 20s think about new smartphone designs?"

[1892] This system makes it possible to efficiently collect feedback from a variety of perspectives and quickly and accurately improve the products offered on the online shopping site, thereby improving customer satisfaction and supporting the development of competitive products.

[1893] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1894] plaintext

[1895] Step 1: Data collection and classification

[1896] The server collects user data from internal databases and external market research data sources. The input includes user attribute data such as age, gender, occupation, and hobbies. Based on this data, the server categorizes the data based on user characteristics. The output is user data by category.

[1897] Step 2: Data Preprocessing

[1898] The server cleans the collected data, removing incomplete or inconsistent data, then normalizes and converts the data into a consistent format. The input includes classified user data. The output is preprocessed, cleaned data.

[1899] Step 3: Generate personas

[1900] The server sets the parameters of the generative AI model based on the preprocessed data, which generates diverse personas that represent specific user demographics. The inputs include the preprocessed data and model parameters. The output is a diverse set of generated personas.

[1901] Step 4: Setting up the questionnaire and creating the format

[1902] The user sets up questionnaire items about specific features of the new product. The server automatically generates a questionnaire format based on the questionnaire items set up by the user. The input includes the questionnaire items set up by the user. The output is the generated questionnaire format.

[1903] Step 5: Conduct the survey and collect responses

[1904] The server distributes surveys to the generated personas. Each persona generates realistic responses using natural language processing techniques. The inputs include the generated personas and the survey format. As an output, the personas' survey responses are collected and stored in a database.

[1905] Step 6: Compile and analyze the results

[1906] The server aggregates all survey results and generates statistics. It also uses statistical analysis tools to detect trends and outliers in the survey results. Inputs include persona survey responses. Outputs include statistics and trend analysis results.

[1907] Step 7: Feedback on results

[1908] The server displays the aggregated and analyzed results on the user's device. Based on the analysis results, it provides specific improvement proposals to the product development department. The input includes statistical information and trend analysis results. The output is improvement proposals that are displayed on the user's device.

[1909] 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.

[1910] This invention relates to a system that uses a generative AI model to generate various personas, uses these personas to conduct product-related surveys, and analyzes areas for improvement of the product based on the survey results and emotion recognition. The program processing of this system is explained below in natural language.

[1911] Preparing for Persona Generation

[1912] Data collection and classification

[1913] server

[1914] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, attribute information such as "female in her 30s, working in finance, running as a hobby" is collected.

[1915] Data Preprocessing

[1916] server

[1917] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[1918] Persona Generation

[1919] AI model parameter set

[1920] server

[1921] The server creates a set of parameters for the generative AI model based on the preprocessed data, which then generates a persona that reflects the characteristics of the target user group.

[1922] Persona Generation

[1923] server

[1924] The server uses a generative AI model to generate a variety of personas, such as a "35-year-old, male, hobby: fishing, occupation: engineer" with a rich personality.

[1925] Product survey design

[1926] Setting up questionnaire items

[1927] User

[1928] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1929] Creating a survey format

[1930] server

[1931] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1932] Conducting a survey using personas

[1933] Sending a survey

[1934] server

[1935] A questionnaire is sent to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1936] Response collection

[1937] server

[1938] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1939] Analysis by emotion engine

[1940] Conducting sentiment analysis

[1941] server

[1942] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[1943] Aggregating emotion data

[1944] server

[1945] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[1946] Aggregation and analysis of survey results and sentiment data

[1947] Data aggregation

[1948] server

[1949] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[1950] Trend and outlier detection

[1951] server

[1952] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[1953] Feedback of results

[1954] Results display

[1955] Terminal

[1956] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[1957] Emotional Feedback

[1958] Terminal

[1959] Based on the analysis results of the emotion engine, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[1960] Improvement suggestions

[1961] server

[1962] Based on the analysis results and sentiment data, the system provides specific improvement suggestions to the product development department. For example, it generates specific suggestions such as "focus on improving battery life" and displays them on the user's device.

[1963] The present invention is a system that collects a variety of viewpoints and emotion data quickly and economically through the above processing steps, and supports effective product development.

[1964] The processing flow will be explained below.

[1965] Step 1:

[1966] Data collection

[1967] server

[1968] The server collects various user data from internal databases and external market research data sources. Key data includes age, gender, occupation, hobbies, and purchasing history. For example, data such as "female in her 30s, working in finance, running as a hobby" may be collected.

[1969] Step 2:

[1970] Data Preprocessing

[1971] server

[1972] The collected data is cleaned to remove incomplete or inconsistent data, and then normalized and converted into a consistent format, specifically by filling in missing values ​​and standardizing the data format.

[1973] Step 3:

[1974] Preparing the parameter set

[1975] server

[1976] The server creates a set of parameters for the generative AI model based on the preprocessed data, including settings to specify the characteristics of the target user demographic. For example, the target age could be set to "25-35 years old" and the area of ​​interest to be "technology."

[1977] Step 4:

[1978] Launching the AI ​​model and generating personas

[1979] server

[1980] The server launches the generative AI model and inputs the prepared parameter set, which generates a variety of personas. For example, a persona such as "28 years old, male, hobby is gaming, occupation is engineer" can be generated.

[1981] Step 5:

[1982] Setting up questionnaire items

[1983] User

[1984] A user in the product development department designs a survey item to inquire about specific features of a new product, such as "What do you think about the design of the new smartwatch?"

[1985] Step 6:

[1986] Creating a survey format

[1987] server

[1988] The server automatically generates a questionnaire format based on the questionnaire items set by the user, which includes multiple-choice questions and open-ended questions.

[1989] Step 7:

[1990] Sending a survey

[1991] server

[1992] The server distributes questionnaires to the generated personas, and each persona generates realistic answers using natural language processing technology based on its characteristics.

[1993] Step 8:

[1994] Response collection

[1995] server

[1996] The server collects the survey responses from each persona and stores them in a database. For example, specific responses such as "A 35-year-old male engineer gives positive feedback on the design of a smartwatch" are collected.

[1997] Step 9:

[1998] Conducting sentiment analysis

[1999] server

[2000] The server uses an emotion engine to perform emotion analysis on the collected survey responses. For example, it may determine that the persona is feeling surprised or satisfied based on the responses.

[2001] Step 10:

[2002] Aggregating emotion data

[2003] server

[2004] Emotional data analyzed by the emotion engine is also aggregated along with the survey data, resulting in an emotional evaluation such as "75% of personas are satisfied with the new feature."

[2005] Step 11:

[2006] Data aggregation

[2007] server

[2008] All survey results and sentiment data are aggregated to generate statistics, such as "80% of personas are satisfied with the new feature."

[2009] Step 12:

[2010] Trend and outlier detection

[2011] server

[2012] The server uses statistical analysis tools to detect major trends and outliers in the survey results, such as finding that "many users in their 30s are dissatisfied with the prices."

[2013] Step 13:

[2014] Results display

[2015] Terminal

[2016] The aggregated and analyzed results obtained from the server are displayed on the user's device. For example, a result such as "The design is highly rated, but battery life needs improvement" may be displayed.

[2017] Step 14:

[2018] Emotional Feedback

[2019] Terminal

[2020] Based on the results of the emotion analysis obtained from the server, emotional feedback is displayed on the user's device. For example, an emotional evaluation such as "The majority of personas are satisfied with the new function" is displayed.

[2021] Step 15:

[2022] Improvement suggestions

[2023] server

[2024] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life," and displays the suggestions on the user's device.

[2025] Example 2

[2026] 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."

[2027] Conventional product improvement support systems have struggled to quickly and efficiently collect diverse user perspectives and emotional data and accurately analyze product improvements. Furthermore, collecting survey data and providing feedback through emotional analysis is a time-consuming and labor-intensive process. To address these challenges, the present invention provides a system that uses a generative AI model to generate diverse personas and efficiently collect and analyze survey data.

[2028] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user data from an internal database and an external data source and cleaning and normalizing the data, means for creating a parameter set for a generative AI model based on the collected and preprocessed data and generating a persona, means for conducting a product-related survey with the generated persona, means for collecting survey results and emotional data and storing the same in a database, means for generating statistical information based on the collected survey results and emotional data and detecting trends and outliers, means for analyzing product improvements based on the generated statistical information and emotional data, and means for displaying the analysis results on a user terminal. This enables rapid and efficient collection and analysis of diverse perspectives and emotional data.

[2029] An "internal database" is a collection of data managed within the system, and is a storage device for storing various user information, past survey results, etc.

[2030] "External data sources" are information sources for obtaining data from outside the system, including market research data and publicly available information.

[2031] "User data" refers to data that includes the attribute information of a specific user, such as age, gender, occupation, hobbies, and purchasing history.

[2032] "Cleaning" is the process of removing incomplete or inconsistent data from collected data and preparing it in a format suitable for data analysis.

[2033] "Normalization" is the process of converting collected data into a consistent format and maintaining consistency between the data.

[2034] A "generative AI model" is an algorithm that uses artificial intelligence (AI) technology to generate diverse personas, taking a specific set of parameters as input.

[2035] A "parameter set" refers to a set of data or conditions that are input into a generative AI model, which then generates a specific persona based on these parameters.

[2036] A "persona" is a virtual character that simulates a specific user profile and has attribute information such as age, gender, occupation, and hobbies.

[2037] A "questionnaire" is a survey method used to solicit opinions on specific questions, and is conducted in a multiple choice or free description format.

[2038] "Emotional data" refers to emotional information analyzed based on questionnaire responses, and includes emotional states such as satisfaction and dissatisfaction.

[2039] "Statistical information" refers to data that shows aggregated results and trends calculated from collected data, and includes various trends and outliers.

[2040] "Trend" refers to a general pattern of tendency or fluctuation found in a large amount of collected data.

[2041] An "outlier" is data that shows significantly different values ​​or patterns compared to other data, and may indicate a particular issue or problem.

[2042] This invention is a system that generates various personas using a generative AI model, conducts product-related surveys using these personas, and analyzes product improvements based on the results and sentiment analysis. The main roles of this system are played by the server, users, and terminals.

[2043] 1. Data collection and classification

[2044] Server Data Collection

[2045] The server collects user data from an internal database and external data sources. The internal database includes existing customer information, purchase history, survey results, etc. Market research data and public data are also used as external data sources. The server uses a data collection API to obtain attribute information such as "female in her 30s, working in finance, hobby is running."

[2046] Data classification process

[2047] The server categorizes the collected data into categories such as age, gender, occupation, hobbies, and purchasing history. This categorization creates groups of users with specific attributes, which are used for analysis and persona generation.

[2048] 2. Data Preprocessing

[2049] Data Cleaning

[2050] The server removes incomplete or inconsistent data from the collected data, for example, deleting records with missing occupation data, and normalizing data formats if they are not standardized.

[2051] Data normalization

[2052] The server converts the data into a consistent format, for example by aligning all age data to "decades."

[2053] 3. Parameter set of the AI ​​model

[2054] Creating a Parameter Set

[2055] The server uses the preprocessed data to create a parameter set for the generative AI model, which includes data corresponding to a specific user profile, such as "35 years old, male, engineer, fishing hobby."

[2056] 4. Persona Generation

[2057] Persona generation process

[2058] The server uses a generative AI model to generate various personas, resulting in a specific persona such as "35 years old, male, engineer, fishing hobby." The generated personas are stored in an internal database.

[2059] 5. Setting up the questionnaire items

[2060] Survey design

[2061] The user (a product development department employee) sets specific questions about the product, for example, designing a questionnaire item such as "What do you think about the design of the new smartwatch?"

[2062] 6. Creating a survey format

[2063] Generate a questionnaire format

[2064] The server automatically generates a questionnaire format based on the questionnaire items set by the user, including multiple choice questions and open-ended questions, and stores the format in electronic format.

[2065] 7. Conduct a survey using personas

[2066] Survey distribution processing

[2067] The server then sends a questionnaire to the persona, which uses natural language processing technology to generate realistic responses based on its attributes. For example, a scenario might be generated in which a 35-year-old male engineer gives positive feedback on the design of a new smartwatch.

[2068] Collecting responses

[2069] The server collects the persona's responses to the questionnaire and stores them in a database for later analysis.

[2070] 8. Conducting sentiment analysis

[2071] Emotional Data Analysis

[2072] The server performs sentiment analysis based on the collected survey responses, using an emotion engine to identify emotional data such as "the persona is satisfied based on the responses" or "they are surprised."

[2073] 9. Data Collection and Analysis

[2074] Data aggregation process

[2075] The server statistically aggregates the survey results and sentiment data, generating aggregated results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design."

[2076] Trend and outlier detection

[2077] The server uses the statistics to detect key trends and outliers, such as finding that certain age groups are unhappy with prices or that certain interests are responding positively to new features.

[2078] 10. Feedback of results

[2079] Display the results on the user's device

[2080] The server displays the analysis results on the user's device, where the user can see results such as "Male users in their 40s are very satisfied with the new features" and "They rate the design highly, but are often dissatisfied with the battery life."

[2081] 11. Improvement suggestions

[2082] Proposal for improvement

[2083] Based on the analysis results and sentiment data, the server provides specific improvement suggestions to the product development department, such as "focus on improving battery life" or "improve the finer details of the design," which are displayed on the user's device.

[2084] Through the above processing, this system quickly and efficiently collects diverse viewpoints and emotional data, supporting effective product development.

[2085] Prompt Sentence Examples

[2086] Below is an example of an input prompt for a generative AI model.

[2087] Generate feedback for the design of a new smartwatch based on the demographic information "female in her 30s, working in finance, running as a hobby."

[2088] Example of feedback: Positive about the design, but dissatisfied with the battery life.

[2089] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2090] Step 1: Data collection and classification

[2091] The server collects user data from an internal database and external data sources. As input, customer information from the internal database and external market research data are provided to the server. The server obtains this data via API and stores the data obtained from each data source together. As output, user data categorized by age, gender, occupation, hobbies, purchase history, etc. is obtained.

[2092] Step 2: Data cleaning

[2093] The server removes incomplete or inconsistent data from the data collected in step 1. User data collected in various formats is provided to the server as input. The server performs processing such as deleting records with missing values ​​and imputing some missing values. The output is cleaned user data.

[2094] Step 3: Data normalization

[2095] The server normalizes the cleaned data and converts it into a consistent format. As input, the cleaned user data is provided to the server. For example, the server converts age data into a unified format (e.g., decade units) to align all data formats. As output, the server obtains user data in a consistent format.

[2096] Step 4: Creating a parameter set for the AI ​​model

[2097] The server creates a parameter set for the generative AI model based on the preprocessed data. Normalized user data is provided to the server as input. The server extracts parameters corresponding to each persona (e.g., "35 years old, male, engineer, hobby is fishing") and creates a parameter set to input into the generative AI model. The parameter set for the generative AI model is obtained as output.

[2098] Step 5: Persona Generation

[2099] The server uses a generative AI model to generate various personas. A set of parameters is input to the generative AI model. The server runs the generative AI model to generate various personas (e.g., "35 years old, male, engineer, hobby is fishing") and stores them in an internal database. The generated personas are obtained as output.

[2100] Step 6: Setting up the survey items

[2101] The user sets specific questions about the product. As input, the user is provided with basic information for setting up the questionnaire items. The user designs specific questionnaire items such as "What do you think about the design of the new smartwatch?". As output, the set questionnaire items are obtained.

[2102] Step 7: Generate the survey format

[2103] The server automatically generates a questionnaire based on the questionnaire items set by the user. As input, the set questionnaire items are provided to the server. The server creates an electronic format containing multiple choice and open-ended questions. As output, the generated questionnaire is obtained.

[2104] Step 8: Send out the survey

[2105] The server distributes the questionnaire to the generated personas. As input, the server is provided with the questionnaire format and persona data. The server sends the questionnaire to each persona and collects the responses. As output, the persona's questionnaire responses are obtained.

[2106] Step 9: Generate and collect answers

[2107] The server generates the answers of the generated persona using natural language processing technology. As input, the persona data and the questionnaire form are provided to the server. The server uses natural language processing technology to generate realistic answers and stores them in a database. As output, the server obtains the questionnaire answers generated by natural language processing.

[2108] Step 10: Sentiment Analysis

[2109] The server performs sentiment analysis based on the collected survey responses. The persona's survey responses are provided to the server as input. The server uses an emotion engine to identify emotions such as "the persona is satisfied" or "surprised" based on the responses. The analyzed emotional data is obtained as output.

[2110] Step 11: Data collection and analysis

[2111] The server statistically aggregates the survey results and emotional data. The survey responses and emotional data are provided to the server as input. The server statistically analyzes this data and obtains results such as "80% of personas are satisfied with the new features" or "60% of women in their 30s are dissatisfied with the design." The output is aggregated statistical information.

[2112] Step 12: Detect trends and outliers

[2113] The server detects key trends and outliers from the collected data. As input, aggregated statistics are provided to the server. The server uses statistical analysis tools to determine results such as "certain age groups are dissatisfied with the price" or "certain hobbies are responding positively to new features." As output, the detected trends and outliers are provided.

[2114] Step 13: Feedback and display of results

[2115] The server displays the analysis results on the user's device. The analysis results and emotional data are provided to the server as input. The server sends this to the user's device, which then displays the results, such as "Male users in their 40s are very satisfied with the new function." The analysis results are displayed on the user's device as output.

[2116] Step 14: Present improvement suggestions

[2117] The server provides specific improvement suggestions to the product development department based on the analysis results and emotion data. Detailed analysis results and emotion data are provided to the server as input. The server generates specific suggestions, such as "focus on improving battery life" or "improve the finer details of the design," and sends them to the user's device. Specific improvement suggestions are obtained as output.

[2118] (Application example 2)

[2119] 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."

[2120] While conventional persona generation systems can analyze product improvements based on statistical information, they have the problem of being unable to obtain information that reflects actual customer behavior and emotions in real time. This can sometimes make it difficult to accurately evaluate the effects of product improvements, creating a need for more accurate and immediate data collection and analysis.

[2121] 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.

[2122] In this invention, the server includes means for generating various personas using a generative AI model, means for conducting a product-related questionnaire for the generated personas, means for aggregating the questionnaire results and generating statistical information, means for analyzing product improvements based on the generated statistical information, means for displaying the analysis results on a user terminal, means for capturing customer behavior and facial expressions using smart glasses, and means for analyzing customer emotions in real time based on the captured data. This makes it possible to collect data reflecting customer behavior and emotions in real time and instantly identify areas for improvement in products and services.

[2123] A "generative AI model" is an artificial intelligence technology used to automatically generate diverse personas.

[2124] A "persona" is a virtual user profile with specific attributes and behavioral patterns.

[2125] A "questionnaire" is a method for asking questions to subjects and collecting their responses.

[2126] "Statistical information" is information presented in the form of numbers, graphs, etc. obtained by aggregating and analyzing multiple data.

[2127] "Analysis" is the process of evaluating collected data and extracting meaningful information.

[2128] A "user terminal" is a device such as a computer or smartphone that is used to display results and data.

[2129] "Smart glasses" are wearable devices that the wearer uses to display external information and collect data through sensors.

[2130] "Behavioral capture" is a method of observing customer behavior and recording it as digital information.

[2131] "Facial expression capture" is a method of observing a customer's facial expressions and recording them digitally.

[2132] "Emotion analysis" is a technology that identifies and evaluates the emotions expressed by customers based on captured facial expression data.

[2133] "Natural language processing technology" is a technology that enables computers to understand and generate human language.

[2134] This invention is a system that uses smart glasses in brick-and-mortar stores to analyze customer behavior and facial expressions in real time and utilizes generative AI models based on that data. By analyzing customer emotions and behavior, this system can instantly identify product evaluations and in-store improvements.

[2135] Hardware and Software Configuration

[2136] To implement this system, the following hardware and software are used.

[2137] Hardware

[2138] Smart glasses: Wearable devices for capturing customer behavior and facial expressions.

[2139] Server: Collects, analyzes and manages data.

[2140] User terminal: The device that displays and manipulates data (e.g., PC, smartphone).

[2141] software

[2142] Generative AI models: Used to generate diverse personas.

[2143] Sentiment analysis engine: Technology that analyzes customer sentiment in real time based on captured data.

[2144] Natural language processing techniques: Used to generate answers for personas.

[2145] Specific operating procedures for the system

[2146] 1. Data collection: The server captures customer behavior and facial expression data in real time through the smart glasses.

[2147] 2. Data Preprocessing: The server preprocesses the captured data and converts it into a format that can be analyzed by the sentiment analysis engine.

[2148] 3. Sentiment analysis: The server uses a sentiment analysis engine to analyze customer sentiment in real time.

[2149] 4. Persona generation: The server uses a generative AI model to generate a persona with specific attributes and behavioral patterns.

[2150] 5. Conducting a survey: The server conducts a survey about the product to the generated personas and collects responses.

[2151] 6. Response analysis: The server aggregates the collected survey results and generates statistical information.

[2152] 7. Analysis of Improvement Points: The server analyzes the improvement points of the product based on the generated statistical information.

[2153] 8. Displaying results: The server displays the analysis results on the user's terminal in real time.

[2154] Specific examples

[2155] For example, when a customer in a store picks up a new smartwatch, their behavior and facial expression data are captured by the smart glasses. If the emotion analysis engine detects emotions such as "surprise" or "satisfaction," the server can instantly generate information such as "this smartwatch is popular among men in their 20s and 30s" and display it on the user's device.

[2156] Prompt Sentence Examples

[2157] Below are examples of prompts that can be used to conduct a survey on the generated personas.

[2158] Age: 30, Gender: Female, Purchasing history: Unknown, Hobbies: Fashion

[2159] In this way, the system of the present invention makes it possible to analyze customer sentiment and behavior in real time, evaluate products, and identify areas for improvement.

[2160] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2161] Step 1:

[2162] The server captures customer behavior and facial expression data in real time through the smart glasses. When customers browse and pick up products in the store, the video data is collected by the camera in the smart glasses. The input is the video data, and the output is the captured raw data.

[2163] Step 2:

[2164] The server preprocesses the captured video data and converts it into a format that can be analyzed by the emotion analysis engine. Specifically, it converts the video data to grayscale, trims unnecessary parts, and extracts facial features. The input is the captured video data, and the output is preprocessed facial data.

[2165] Step 3:

[2166] The server performs emotion analysis using the preprocessed facial data. It uses an emotion analysis engine to identify emotions such as happiness, surprise, sadness, etc. based on the data. The input is the preprocessed facial data, and the output is the identified emotion data.

[2167] Step 4:

[2168] The server inputs customer data into a generative AI model to generate various personas. For example, attribute data such as a customer's age, gender, purchasing history, and hobbies are input to create a virtual user profile. The input is customer attribute data, and the output is the generated persona.

[2169] Step 5:

[2170] The server then conducts a product-related questionnaire for the generated persona. The questionnaire is designed to collect customer preferences and opinions using natural language processing technology. The input is the persona and the questionnaire items, and the output is the persona's responses to the questionnaire.

[2171] Step 6:

[2172] The server aggregates the collected survey results and generates statistical information. It analyzes the response data of each persona and extracts common trends and outliers. The input is the survey response data, and the output is statistical information.

[2173] Step 7:

[2174] The server analyzes product improvements based on the generated statistical information. It identifies trends and outliers to determine which parts of the product or service should be improved. The input is statistical information, and the output is the identification of areas for improvement.

[2175] Step 8:

[2176] The server displays the analysis results in real time on the user's device. The user's device receives information on product evaluations and areas for improvement and displays it in a format that is intuitively easy for the user to understand. The input is the analysis results, and the output is the data displayed on the user's device.

[2177] 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.

[2178] 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.

[2179] 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.

[2180] 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.

[2181] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2182] 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.

[2183] 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).

[2184] 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.

[2185] 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."

[2186] 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.

[2187] 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).

[2188] 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.

[2189] 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.

[2190] 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.

[2191] 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.

[2192] 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.

[2193] 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.

[2194] 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.

[2195] 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.

[2196] 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.

[2197] 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.

[2198] The following is further disclosed regarding the above embodiment.

[2199] (Claim 1)

[2200] A means of generating diverse personas using generative AI models; and

[2201] A means of conducting product surveys on the generated personas;

[2202] a means for aggregating the surv...

Claims

1. A means of generating diverse personas using generative AI models; and A means of conducting product surveys on the generated personas; a means for aggregating the survey results and generating statistical information; A means of analyzing product improvements based on the generated statistical information; The system includes means for displaying the analysis results on a user terminal.

2. The system of claim 1 further comprising means for a user to design questionnaire items and generate a questionnaire format.

3. The system of claim 1 , further comprising means for generating answers for the generated persona using natural language processing techniques.

Images