system

Abstract

Provide a system. [Solution] A means for collecting past reviews and purchase history from reliable reviewers; means for constructing a generative model that learns from the collected reviewer data; A means for generating an avatar that imitates a reviewer using the trained generative model; a means for passing a question from a user to said alter-ego avatar and returning an answer; A system including:

Description

[Technical Field]

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

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

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

[0004] When making purchasing decisions based on product reviews, there is a problem of difficulty in identifying appropriate reviews. While reliability can be improved by narrowing down to reviews that many people have referenced, the lack of information not included in those reviews makes it difficult for users to make purchasing decisions with complete information. In such cases, users may end up regretting their purchase of an inappropriate product. To solve this problem, a system is needed that allows users to obtain additional information based on reliable reviewers in real time. [Means for solving the problem]

[0005] The present invention provides a system that includes a means for collecting past reviews and purchase histories from trusted reviewers, a means for constructing a generative model that learns from the collected reviewer data, a means for generating an avatar that mimics the reviewer using the learned generative model, and a means for passing questions from a user to the avatar and receiving answers from the avatar. This allows users to ask questions to the virtual avatar of a trusted reviewer and make purchasing decisions based on appropriate information. This system provides an effective means for increasing the reliability of reviews and reducing users' post-purchase regret.

[0006] A "trusted reviewer" is one who posts accurate and detailed reviews of products and receives high ratings from many users.

[0007] "Past reviews" are records of product ratings and comments previously posted by the reviewer.

[0008] "Purchase history" refers to a list of products a reviewer has previously purchased and detailed information about those purchases.

[0009] "Means of collection" refers to the methods and technologies used to obtain a reviewer's past reviews and purchasing history from databases or online platforms.

[0010] A "generative model" is a statistical or machine learning algorithm that learns from collected data and makes predictions and judgments.

[0011] The "means of construction" refers to the methods and techniques used to collect and shape data and train generative models.

[0012] An "avatar" is a virtual agent that mimics the ratings and commenting style of a specific reviewer based on a trained generative model.

[0013] "Question from user" is text that the user can use to ask their avatar about information they want to know about a product.

[0014] "Means for generating an answer" refers to the method or technology by which the avatar creates an appropriate response to a question from the user.

[0015] "Means of returning to the user" refers to the method or technique used to communicate the generated answer to the user. [Brief explanation of the drawings]

[0016] [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

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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."

[0037] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0038] Server Operation

[0039] First, the server collects the past reviews and purchase history of trusted reviewers. This is achieved by pulling data from databases or online platforms. It retrieves the reviewer's data using a specific reviewer ID.

[0040] Next, the server builds a generative model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative model created here can then mimic the reviewer's opinions and evaluations.

[0041] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0042] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0043] Device behavior

[0044] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field for users to enter questions about a particular product or service, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0045] User behavior

[0046] Users enter questions about the product or service they are considering purchasing. For example, they can enter specific questions such as, "Is this product child-safe?" After entering their question, the user presses a submit button to send the question to the server.

[0047] Specific examples

[0048] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0049] The avatar generates a response based on past reviewer data: "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user. As a result, the user can make purchasing decisions based on reliable information.

[0050] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0051] The processing flow will be explained below.

[0052] Step 1:

[0053] The server collects past reviews and purchase history of trusted reviewers from databases and online platforms, and retrieves the required data using a specific reviewer ID.

[0054] Step 2:

[0055] The server builds a generative model based on the collected reviewer data, and the model is trained to learn reviewers' evaluation styles and comments and understand their patterns.

[0056] Step 3:

[0057] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to answer questions from users.

[0058] Step 4:

[0059] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0060] Step 5:

[0061] When a user enters a question and presses the send button, the device sends the question to the server, which transmits the data from the device to the server via an Internet connection.

[0062] Step 6:

[0063] The server receives a question from the user and passes the question to the corresponding avatar, identifies the corresponding avatar in the server, and passes the question data to it.

[0064] Step 7:

[0065] The avatar generates answers to questions based on past reviews and purchasing history. The avatar refers to the data it has learned and creates the most appropriate answer to the user's question.

[0066] Step 8:

[0067] The server receives the generated response and sends it back to the device, which also transmits the data back to the device via the Internet connection.

[0068] Step 9:

[0069] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended."

[0070] Step 10:

[0071] Users can make purchasing decisions based on the displayed answers, gaining more reliable information, which can reduce regrets such as "I shouldn't have bought it."

[0072] As described above, the system of the present invention provides users with the information they need through the virtual avatars of trusted reviewers, and supports their purchasing behavior.

[0073] Example 1

[0074] 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."

[0075] Existing online review systems have problems in that it is difficult to easily refer to the opinions of reliable reviewers, and extracting necessary information from a huge number of reviews requires a lot of time and effort. This makes it difficult for users to quickly obtain reliable information regarding purchasing products and services.

[0076] 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.

[0077] In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for preprocessing the collected reviewer data, means for constructing a generative AI model that learns the preprocessed data, means for generating an avatar that imitates the reviewer using the trained generative AI model, and means for passing questions from users to the avatar and returning answers, thereby enabling users to quickly obtain reliable information.

[0078] A "trusted reviewer" is a reviewer who provides a reliable rating based on past reviews and purchasing history.

[0079] "Past reviews" refer to ratings and comments about products or services that a reviewer has previously written.

[0080] "Purchase history" refers to a record of products or services that a reviewer has previously purchased.

[0081] A "generative AI model" is a model generated from training data using artificial intelligence technology.

[0082] "Preprocessing" refers to the process of properly cleansing collected data and performing format conversion, tokenization, etc.

[0083] An "avatar" is a virtual persona created to mimic a reviewer's rating style and opinions.

[0084] "User interface" refers to the screen containing input forms and buttons that a user can operate on a device.

[0085] "Question handing" refers to the process of providing a user's inquiry to an avatar.

[0086] "Returning an answer" refers to the process of providing the user with a response generated from the avatar.

[0087] "Collection methods" refers to the methods and devices used to obtain the required data from databases and online platforms.

[0088] The "means of construction" refers to the methods and processes required to create a generative AI model.

[0089] "Means of generation" refers to the method or process for creating an avatar using a trained generative AI model.

[0090] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0091] Server Operation

[0092] First, the server collects past reviews and purchase history of trusted reviewers. This can be achieved by retrieving data from a database or online platform. For example, the server can use an online store's API to retrieve reviews and purchase history based on a specific reviewer ID. Specifically, the server could use Amazon's API to retrieve the review history of a specific reviewer.

[0093] The server then preprocesses the collected data. This preprocessing includes data cleansing, formatting, and tokenization. For example, it removes HTML tags from text data and splits reviews into individual words. These operations are performed using Python NLP libraries (e.g., NLTK and spaCy).

[0094] The server builds a generative AI model based on the preprocessed data and performs training. Here, a generative AI model such as GPT-3 (registered trademark) is used to learn the reviewer's evaluation style and comment patterns. Frameworks such as TENSORFLOW (registered trademark) and PyTorch are used to train the model, and multiple epochs are used.

[0095] Using a trained generative AI model, the server generates an avatar that mimics the reviewer's ratings and opinions and generates appropriate answers based on user questions.

[0096] Finally, the server receives the user's question and passes it to the avatar. The question is input into the generative AI model, and the output answer is returned to the user via the server. For example, an API endpoint can be set up using a web framework such as Flask to exchange questions and answers.

[0097] Device behavior

[0098] The device provides a user interface (UI) that allows users to enter questions. This UI includes a text field and a submit button for users to enter questions about specific products or services. When a user enters a question and presses the submit button, the question is sent to the server. Specifically, the UI is created using HTML and JavaScript (registered trademark), and the question is sent to the server via asynchronous communication using a JavaScript Ajax request.

[0099] User behavior

[0100] Users enter questions about the product or service they are considering purchasing into the device's UI. For example, they can enter specific questions such as, "Is this product child-safe?" After entering the question, the user presses a send button to send the question to the server.

[0101] Specific examples

[0102] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0103] The avatar uses past reviewer data to generate a response such as, "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user, allowing the user to make purchasing decisions based on reliable information.

[0104] An example prompt could be, "Is this product safe for children? Please answer based on past reviews and purchasing history from trusted reviewers." Based on this, the generative AI model generates an answer in natural language.

[0105] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0106] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0107] Step 1:

[0108] The server collects past reviews and purchase history from a database or online platform (e.g., an online store's API) based on a trusted reviewer ID. The server receives a specific reviewer ID as input and sends an HTTP request to the database. This request retrieves the reviewer's review history and purchase history in JSON format. This retrieved data is temporarily stored in the server's storage.

[0109] Step 2:

[0110] The server preprocesses the collected review and purchase history data. Specifically, it performs data cleansing (removing unnecessary and duplicate data), format conversion, and tokenization. The server takes the ingested JSON data as input and performs data cleansing and tokenization using a Python NLP library (e.g., NLTK or spaCy). The preprocessed data is then output as a dataset for model training.

[0111] Step 3:

[0112] The server builds and trains a generative AI model based on the preprocessed reviewer data. A generative AI model such as GPT-3 is used for the training process. The server uses the preprocessed reviewer data as input and trains the generative AI model using a machine learning framework such as TensorFlow or PyTorch. This training process outputs a generative model that mimics the reviewer's evaluation style and comment patterns.

[0113] Step 4:

[0114] The server uses the trained model to generate an avatar. This enables the server to generate answers to user questions based on their evaluation style and opinions. The server inputs the trained generative AI model and outputs profile data for the avatar. This profile data includes the results of feature extraction for the reviewer.

[0115] Step 5:

[0116] The device provides a UI that allows users to enter questions. Users enter specific questions about the product or service they are considering purchasing. When a question such as "Is this product child-safe?" is entered, the device treats it as input data and sends it to the server using a JavaScript Ajax request. This request transfers the question data to the server.

[0117] Step 6:

[0118] The server passes the question received from the user to the avatar and generates an answer. The server uses the question data received from the device as input and generates an answer based on the trained generative AI model. The generated answer is output as an HTTP response from the server to the device.

[0119] Step 7:

[0120] The terminal displays the answer received from the server to the user. The terminal receives the HTTP response from the server as input, analyzes its contents, and displays it on the UI. The answer is presented in a format that is easy for the user to understand, and the user can make a purchasing decision based on that information.

[0121] (Application example 1)

[0122] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0123] On traditional online shopping sites, it is difficult to obtain reliable information about products, especially for new or lesser-known products. Furthermore, highly anonymous online reviews often lack credibility, making it difficult for users to provide sufficient information for making confident purchasing decisions. This creates the risk of users making incorrect purchasing decisions based on inaccurate information. To address these issues, a system is needed that generates avatars based on the ratings and purchasing history of trusted reviewers, allowing users to easily obtain reliable information.

[0124] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0125] In this invention, the server includes a means for collecting past reviews and purchase histories of reliable reviewers, a means for constructing a generative AI model that learns the collected review data, and a means for generating an avatar that imitates the reviewer using the trained generative AI model, thereby enabling users to obtain reliable information about products on online shopping sites.

[0126] A "trusted reviewer" is someone who has a high rating and credibility based on past reviews and purchasing history.

[0127] "Past ratings and purchasing history" refers to the reviewer's previous ratings of products and services and the history of products they have purchased.

[0128] "Means of collection" refers to the methods used to obtain reliable reviewer ratings and purchasing data using databases or online platforms.

[0129] A "generative AI model" is an artificial intelligence model that learns from collected reviewer data and imitates reviewer evaluation styles and comment patterns.

[0130] An "avatar" is a virtual persona that uses generative AI models to mimic the opinions and ratings of trusted reviewers.

[0131] "User interface" refers to the screen layout and operation method that allows users to easily enter questions and receive answers.

[0132] The "server" is a central computer that collects reviewer data, trains the generative AI model, generates avatars, and handles questions and answers from users.

[0133] A "terminal" is a device used by a user, an electronic device that operates in the form of a smartphone app, etc.

[0134] The system for implementing the present invention operates mainly based on three entities: a server, a terminal, and a user.

[0135] Server Operation

[0136] The server collects the past ratings and purchase history of trusted reviewers. This is achieved by retrieving reviewer data from databases or online platforms. It uses a specific reviewer ID to collect the data of the corresponding reviewer.

[0137] The server then builds a generative AI model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative AI model created here can then mimic the reviewer's opinions and evaluations.

[0138] The server then uses the trained generative AI model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0139] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0140] Device behavior

[0141] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field where users can enter questions about specific products or services, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0142] User behavior

[0143] The user enters a question about a product or service they are considering purchasing. For example, they can enter a specific question such as, "What age group is this toy suitable for?" After entering the question, the user presses a submit button to send the question to the server.

[0144] Specific examples

[0145] For example, consider the case where a user enters the question "What age group is this toy suitable for?" into the device's UI and presses the send button. The device sends this question to the server, which then forwards it to the corresponding reviewer's avatar. Based on past reviewer data, the avatar generates an answer such as "This toy is particularly popular with children aged 3 to 6. It is also very safe." The server receives this answer and sends it back to the device. Finally, the device can display the received answer to the user. This allows the user to make purchasing decisions based on reliable information.

[0146] Prompt Sentence Examples

[0147] Below are some examples of specific prompt sentences.

[0148] text

[0149] Reviewer data: This toy is safe for children aged 3 to 6. It's very popular among parents.

[0150] Question: What age group is this toy suitable for?

[0151] Answer:

[0152] The server uses this to ask questions to the avatar and provides the generated answers to the user. In this way, the system of the present invention provides reliable information via the avatar that mimics the opinions of reliable reviewers, thereby supporting users' purchasing behavior.

[0153] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0154] Step 1: Collect data from trusted reviewers

[0155] The server collects the past ratings and purchase history of trusted reviewers from databases and online platforms. In this process, it uses the reviewer ID as a key to obtain detailed rating data and purchase history data for each reviewer. The input is the reviewer ID and database connection information, and the output is the reviewer's detailed rating data and purchase history.

[0156] Step 2: Training the generative AI model

[0157] The server builds and trains a generative AI model based on the collected reviewer data. In this process, the reviewer's evaluation style and historical purchasing patterns are used as training data to train the AI ​​model (e.g., GPT-2). The input is the reviewer data, and the output is the trained generative AI model.

[0158] Step 3: Creating an avatar

[0159] The server uses a trained generative AI model to generate an avatar that mimics the reviewer. This avatar has the ability to answer questions based on the collected information. The input is the trained generative AI model, and the output is the avatar.

[0160] Step 4: Providing a User Interface

[0161] The terminal provides a user interface (UI) where the user can enter a question. This UI includes a field for the user to enter a question and a submit button. The input is the user's action feedback, and the output is the question entered in the question field.

[0162] Step 5: Sending the question to the server

[0163] The terminal sends the question entered by the user to the server. In this process, the question and the user ID are sent to the server through an HTTP request. The input is the user's question, and the output is the HTTP request sent to the server.

[0164] Step 6: Transfer the question to your avatar

[0165] The server passes the received user question to the corresponding avatar. When the question is passed to the avatar, an appropriate prompt is generated using a generative AI model. The input is the user's question, and the output is the prompt passed to the avatar.

[0166] Step 7: Generate an answer

[0167] The avatar generates answers based on the questions it receives from users. Using a generative AI model, the answers are generated in a way that mimics the rating style of a trusted reviewer. The input is the prompt, and the output is the generated answer.

[0168] Step 8: Sending the answer back to the server

[0169] The avatar then sends the generated answer back to the server, where it is ready to be sent back to the user. The input is the generated answer, and the output is the answer sent to the server.

[0170] Step 9: Providing answers to users

[0171] The server sends the generated answer back to the user's device, where it can be viewed by the user. The device displays the answer received from the server in a user interface. The input is the answer from the server, and the output is the answer displayed on the device's display screen.

[0172] At each step, users can get answers from their avatars based on the opinions of trusted reviewers, allowing them to make purchasing decisions with confidence.

[0173] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0174] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to ask questions to the avatar to obtain reliable information. It also combines an emotion engine that recognizes the user's emotions and provides answers that correspond to the user's emotions. This system operates appropriately mainly through three entities: a server, a terminal, and a user, as well as the emotion engine.

[0175] Server Operation

[0176] First, the server collects the past reviews and purchase history of trusted reviewers from databases or online platforms, and then retrieves the required data using a specific reviewer ID.

[0177] The server then builds and trains a generative model based on the collected data. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The trained generative model can faithfully mimic the reviewer's opinions and evaluations.

[0178] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and answers questions from users.

[0179] The server then uses an emotion engine to recognize the user's emotions, which analyzes the user's input text and voice to identify their current emotions.

[0180] Device behavior

[0181] The device provides a user interface (UI) that allows the user to enter a question. The UI includes a field for the user to enter a question about a particular product or service and a submit button. When the user enters a question, the device can also capture the user's voice and text sentiment and transmit it to the server.

[0182] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0183] Collaboration between the server and emotion engine

[0184] The server receives the user's question and emotion data, then passes the question to the avatar to generate an answer. At the same time, the emotion engine analyzes the emotion data and identifies the user's emotional state. For example, if the user is feeling anxious, that emotion is communicated to the server.

[0185] The emotional state of the user analyzed by the emotion engine influences the avatar's responses. For example, if the user is feeling anxious, the avatar will adjust its responses to take that into consideration. For example, the answer to the question, "Is this product safe for children?" will be "Yes, it is safe. However, we also have some recommendations for further safety checks." This will add information to alleviate anxiety.

[0186] Specific examples

[0187] For example, suppose a user emotionally inputs the question "Is this product safe for children?" and speaks it in a tone that indicates anxiety. The device captures the user's voice data along with the question and sends it to the server. The server then sends the question to the avatar, while simultaneously analyzing the emotional data with its emotion engine and recognizing the user's anxiety.

[0188] The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." It then receives information from the emotion engine that the user is feeling uneasy and responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0189] The device displays this response to the user, allowing the user to receive reliable information and a sense of security at the same time. In this way, the system of the present invention provides reliable information that reflects the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0190] The processing flow will be explained below.

[0191] Step 1:

[0192] The server collects the past reviews and purchase history of trusted reviewers from databases and online platforms, and then retrieves the required data using a specific reviewer ID.

[0193] Step 2:

[0194] The server builds and trains a generative model based on the collected reviewer data. The model learns and understands reviewer evaluation styles and comment patterns.

[0195] Step 3:

[0196] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to provide answers to user questions.

[0197] Step 4:

[0198] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0199] Step 5:

[0200] When users input questions, the device also captures text and voice data to recognize users' emotions, and has a built-in emotion engine to analyze users' emotions.

[0201] Step 6:

[0202] When a user enters a question and presses the send button, the device sends the question and captured emotion data to the server via an internet connection.

[0203] Step 7:

[0204] The server receives the question and emotion data from the user, and first passes the question to the avatar to generate an answer.

[0205] Step 8:

[0206] The server then uses an emotion engine to analyze the emotion data and identify the user's emotional state, for example, the emotion engine may recognize that the user is feeling anxious.

[0207] Step 9:

[0208] The emotional state of the user analyzed by the emotion engine influences the avatar's response. Specifically, the avatar takes into account the emotional data and generates responses that correspond to the user's emotions.

[0209] Step 10:

[0210] For example, if a user types the question "Is this product child-safe?" with some anxiety, the avatar will generate the answer "This product is child-safe, but we recommend age-appropriate use. Regular maintenance and child supervision are also important as additional safety precautions."

[0211] Step 11:

[0212] The server receives the generated response and sends it back to the device, which also sends the data back to the device via the Internet connection.

[0213] Step 12:

[0214] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended. Regular maintenance and child supervision are also important as additional safety precautions."

[0215] Step 13:

[0216] Users can make purchasing decisions based on the displayed answers, obtaining more reliable information, which reduces the chance of users regretting their purchase.

[0217] As described above, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0218] Example 2

[0219] 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."

[0220] Today's consumers seek reliable information about products and services online, but online reviews and ratings often lack credibility. Furthermore, when users consult online reviews, they face challenges in finding advice that is tailored to their emotions and circumstances. Furthermore, information overload and conflicting reviews make it difficult for users to make appropriate decisions.

[0221] 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 acquiring past evaluation information and transaction history of reliable reviewers, means for constructing a generative model for learning the acquired reviewer data, means for generating a virtual agent that mimics the reviewer using the learned generative model, means for passing questions from a user to the virtual agent and providing the answers, means for identifying an emotional state from the user's input, and means for adjusting the answers of the virtual agent based on the identified emotional state. This allows the user to receive highly reliable information as well as advice that corresponds to their current emotions.

[0222] A "trusted reviewer" is a reviewer who has provided accurate and fair ratings in the past and has received high ratings from many users.

[0223] "Rating Information" refers to the content of the review or feedback provided by the reviewer regarding a product or service.

[0224] "Transaction History" refers to records of products or services that a reviewer has previously purchased or used.

[0225] "Capture methods" refers to software and processes used to collect data from databases and external platforms.

[0226] "Reviewer Data" refers collectively to rating information and transaction history associated with trusted reviewers.

[0227] A "generative model" refers to an artificial intelligence algorithm or mechanism that learns from specific data and generates new data.

[0228] A "virtual agent" is a virtual entity that uses a trained generative model to mimic a specific reviewer and provide answers to their questions.

[0229] "User input" refers to data input, such as text or voice, made by a user to a system.

[0230] "Means for identifying emotional state" refers to the algorithms or engines used to analyze emotions from user input and identify that state.

[0231] "Response tailoring" refers to a process or algorithm for appropriately modifying the responses provided based on the identified emotional state.

[0232] This invention is a system that generates a virtual agent based on the past evaluation information and transaction history of trusted reviewers, and allows users to ask the virtual agent questions to obtain reliable information. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it provides answers that correspond to the user's emotions. This system consists of three entities: a server, a terminal, and a user.

[0233] The server collects the past rating information and transaction history of trusted reviewers from a database or online platform, for example, using SQL Server or a NoSQL database (e.g., MySQL (registered trademark), MongoDB), and retrieves the required data using the reviewer ID.

[0234] Next, the server builds a generative AI model (e.g., GPT-4 (registered trademark)) based on the collected data and trains it. During this training process, a GPU is used to quickly process large amounts of data, and reviewers' evaluation styles and comment patterns are used as training data. This creates a generative AI model that faithfully mimics the reviewers' opinions and evaluations.

[0235] The server then uses the trained generative AI model to generate a virtual agent that mimics the reviewer, which can then use the generative AI model to generate appropriate answers to user questions.

[0236] The device provides a user interface (UI) that allows users to input questions. Specifically, the UI includes a text input field and a submit button. This allows users to easily input and submit questions about specific products or services. The device can also capture and transmit emotional data from the user's voice and text to a server. Voice recognition software (e.g., Google® Speech-to-Text) is used for voice recognition.

[0237] When a user enters a question and presses the send button, the device sends the question and the captured emotion data to the server. The server then passes the received user question to a virtual agent, which generates an answer. In parallel, the server analyzes the emotion data using an emotion engine (e.g., IBM Watson® Tone Analyzer) to identify the user's emotional state. For example, if the user is feeling anxious, the server will recognize that emotion.

[0238] The server adjusts the generated answer based on the user's emotional state analyzed by the emotion engine. For example, if the user is feeling anxious, the server generates an answer that takes their emotion into consideration. In response to the question, "Is this product child-safe?", the virtual agent generates the answer, "This product is child-safe, but we recommend age-appropriate use," and then adds reassuring information such as, "Additional safety measures include regular maintenance and child supervision."

[0239] The device displays the final answer received from the server to the user, allowing the user to obtain reliable information and an appropriate answer that reflects their current emotions.

[0240] As a concrete example, if a user types the question "Is this product child-safe?" in an anxious tone, the following prompt sentence is generated:

[0241] Example prompt: "This product is child-safe, but age-appropriate use is recommended. Additionally, regular maintenance and child supervision are important additional safety precautions."

[0242] In this way, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual agent of a trusted reviewer, thereby supporting the user's decision-making.

[0243] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0244] Step 1:

[0245] The server collects data on trusted reviewers. Database connection information and reviewer IDs are required as input. Using a database system (e.g., MySQL or MongoDB), past review information and transaction history are queried based on the reviewer ID. This query process outputs the reviews and purchase records left by the reviewer.

[0246] Step 2:

[0247] The server trains a generative AI model based on the data collected. The reviewer data collected in step 1 is used as input. The data is input to the generative AI model (e.g., GPT-4) using a GPU, and the training process is carried out. Through this process, the generative AI model learns the reviewer's evaluation style and comment patterns, and outputs the trained model.

[0248] Step 3:

[0249] The server generates a virtual agent using the generative AI model. The generative AI model trained in step 2 is used as input. Based on the trained model, a virtual agent that imitates the reviewer's style is generated and output as an agent that can be used within the system.

[0250] Step 4:

[0251] The device provides a user interface (UI), which includes a text entry field and a submit button. The user enters and submits a question about a product or service. Inputs include the user's text and voice data. The output of this step is the user's question in text format.

[0252] Step 5:

[0253] The user enters a question and presses the submit button. This action causes the user's text and voice data to be stored on the device. The voice data is converted to text using voice recognition software (e.g., Google Speech-to-Text). The input requires the user's question, and the output is converted to text format.

[0254] Step 6:

[0255] The device sends the question and emotion data to the server. The input includes the converted question and speech data stored in step 5. By sending this data to the server, the server receives the question content and emotion data and executes the next step based on them.

[0256] Step 7:

[0257] The server generates an answer to the user's question. The input includes the question text received in step 6. Using the trained generative AI model, it generates an answer based on the question content. The generated answer is based on the virtual agent's mimicking style, and the output is the answer text.

[0258] Step 8:

[0259] The server analyzes the emotion data and reflects it in the answer. The emotion data received in step 6 is included as input. An emotion engine (e.g., IBM Watson Tone Analyzer) is used to analyze the data and identify the emotional state. Based on this emotional state, the answer generated in step 7 is adjusted to output a final answer that takes the user's emotions into consideration.

[0260] Step 9:

[0261] The server sends the final answer to the terminal. As input, the final answer text adjusted in the previous step is required. By sending this data to the terminal, the terminal receives and displays the final answer.

[0262] Step 10:

[0263] The device displays the final answer to the user. The input includes the final answer text received from the server. This is displayed in the user interface, allowing the user to obtain an appropriate answer based on their emotions along with reliable information. The output is the user's ability to view the final answer.

[0264] (Application example 2)

[0265] 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."

[0266] In conventional review systems, users have had difficulty finding reliable information from the vast number of reviews, and there are few ways to obtain appropriate information that matches their own emotions. This invention aims to solve these problems by generating an avatar based on the past reviews and purchase history of reliable reviewers, and providing answers that match the user's emotions.

[0267] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for constructing a generative model that learns the collected review data, means for generating an avatar that imitates the reviewer using the learned generative model, means for adjusting answers according to the user's emotions using an emotion engine that recognizes the user's emotions, and means for passing questions from the user to the avatar and returning the answers. This makes it possible to provide reliable information and obtain appropriate answers according to the user's emotions.

[0268] A "trusted reviewer" is one whose ratings and opinions are accurate and consistent based on past reviews and purchasing history.

[0269] A "review" is an evaluation or comment posted by a user about a specific product or service based on their own experience and opinion.

[0270] "Purchase history" is a record of products and services a user has purchased in the past.

[0271] A "generative model" is an artificial intelligence model that learns from collected data and generates an appropriate output for any input.

[0272] An "avatar" is a virtual entity that mimics the opinions and ratings of trusted reviewers and answers questions from users.

[0273] "User emotion" refers to the user's psychological state or feelings during a question or interaction.

[0274] An "emotion engine" is an algorithm or software that analyzes a user's text or voice data to identify their emotions.

[0275] A "question" is an action in which a user enters a question or request for confirmation regarding a specific product or service.

[0276] An "answer" is information or an opinion provided by an avatar in response to a question from a user.

[0277] The system for implementing this invention is mainly composed of three entities: a server, a terminal, and a user. The function of each entity is explained below.

[0278] Server Operation

[0279] The server operates using the following means:

[0280] 1. Data collection methods:

[0281] The past reviews and purchase history of trusted reviewers are collected from databases and online platforms, and the collected data is used to extract the necessary information using a specific reviewer ID.

[0282] 2. Generative model construction method:

[0283] A generative AI model is constructed based on the collected review data. This generative model uses past reviews and purchase history as training data to learn the reviewer's evaluation style and comment patterns. After training, the generative model is able to faithfully imitate the reviewer's opinions and evaluations.

[0284] 3. Avatar creation method:

[0285] Using a trained generative model, we generate an avatar that mimics the reviewer, and this avatar is used to answer questions from users.

[0286] 4. Emotion recognition means:

[0287] The emotion engine analyzes the user's input text and voice to identify their current emotion.

[0288] This allows the server to generate a virtual avatar of a trusted reviewer and provide answers tailored to the user's emotions.

[0289] Device behavior

[0290] The terminal operates as an interface with the user as follows.

[0291] 1. User interface provision means:

[0292] It provides an interface where users can enter questions. The UI includes a field to enter questions about a specific product or service, a submit button, and the ability to capture user voice and text sentiment.

[0293] 2. Means of data transmission:

[0294] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0295] Examples of emotion-assisted answers

[0296] Specific examples are shown below.

[0297] 1. The user emotionally types the question "Is this product child safe?" and speaks it in a tone that indicates anxiety.

[0298] 2. The device captures the user's voice data along with this question and sends it to the server.

[0299] 3. The server sends a question to the avatar, and at the same time, it analyzes the emotional data using an emotion engine to recognize the user's anxiety.

[0300] 4. The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." Furthermore, based on the emotion engine's information that the user is feeling uneasy, the avatar responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0301] 5. The device will display this answer to the user, providing them with reliable information and a sense of security.

[0302] Prompt Sentence Examples

[0303] Below are some example prompts to provide to the generative AI model:

[0304] User Question: "Is this product safe for children?"

[0305] User sentiment: "Anxiety"

[0306] Avatar response to be generated: "This product is child-safe, however, regular maintenance and child supervision are important additional safety precautions."

[0307] This system allows users to obtain reliable information and receive appropriate answers based on their own emotions.

[0308] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0309] Step 1: Data collection

[0310] The server collects past reviews and purchase history from trusted reviewers. The input is review data and purchase history data obtained using reviewer IDs. The server collects this data from databases and online platforms and stores it in a database. The output is the collected review data and purchase history data.

[0311] Step 2: Building a generative model

[0312] The server builds a generative AI model based on the collected data. Past reviews and purchase history are used as input. The server uses this data as training data and uses a machine learning algorithm to learn the reviewer's evaluation style and comment patterns. The output is a trained generative AI model.

[0313] Step 3: Create an avatar

[0314] The server generates an avatar using a trained generative model. The inputs are the trained generative model and review data. The server uses these data to generate a virtual avatar that closely mimics the reviewer. The output is the avatar.

[0315] Step 4: Submit your question and sentiment data

[0316] A user inputs a question about a product or service into the device's UI. The device captures the user's voice and text emotion data along with the input question and sends it to the server. The input is the user's question and emotion data, and the output is data sent to the server.

[0317] Step 5: Emotion Recognition

[0318] The server analyzes the emotion data sent by the user. The input is the user's emotion data, which is analyzed using the server's emotion engine. The emotion engine analyzes the text and / or voice to determine the user's current emotional state. The output is the analyzed emotional state.

[0319] Step 6: Answer Generation

[0320] The server passes the user's question to the avatar, which generates an answer. The input is the user's question and the analyzed emotional state. The avatar generates an answer based on this. In doing so, it adds adaptive information according to the emotion based on data from the emotion engine. For example, an answer containing additional information that gives a sense of security to an anxious user is generated. The output is the generated answer.

[0321] Step 7: Submit and view your responses

[0322] The server sends the generated answer to the user's device. The device receives the answer and displays it to the user. The input is the generated answer, and the output is the answer displayed to the user. This allows the user to obtain reliable information as well as an appropriate answer based on their emotions.

[0323] 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.

[0324] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0325] 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.

[0326] [Second embodiment]

[0327] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0328] 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.

[0329] 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).

[0330] 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.

[0331] 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.

[0332] 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).

[0333] 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. 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.

[0334] 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.

[0335] 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.

[0336] 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.

[0337] In the smart glasses 214, 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.

[0338] 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."

[0339] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0340] Server Operation

[0341] First, the server collects the past reviews and purchase history of trusted reviewers. This is achieved by pulling data from databases or online platforms. It retrieves the reviewer's data using a specific reviewer ID.

[0342] Next, the server builds a generative model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative model created here can then mimic the reviewer's opinions and evaluations.

[0343] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0344] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0345] Device behavior

[0346] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field for users to enter questions about a particular product or service, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0347] User behavior

[0348] Users enter questions about the product or service they are considering purchasing. For example, they can enter specific questions such as, "Is this product child-safe?" After entering their question, the user presses a submit button to send the question to the server.

[0349] Specific examples

[0350] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0351] The avatar generates a response based on past reviewer data: "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user. As a result, the user can make purchasing decisions based on reliable information.

[0352] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0353] The processing flow will be explained below.

[0354] Step 1:

[0355] The server collects past reviews and purchase history of trusted reviewers from databases and online platforms, and retrieves the required data using a specific reviewer ID.

[0356] Step 2:

[0357] The server builds a generative model based on the collected reviewer data, and the model is trained to learn reviewers' evaluation styles and comments and understand their patterns.

[0358] Step 3:

[0359] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to answer questions from users.

[0360] Step 4:

[0361] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0362] Step 5:

[0363] When a user enters a question and presses the send button, the device sends the question to the server, which transmits the data from the device to the server via an Internet connection.

[0364] Step 6:

[0365] The server receives a question from the user and passes the question to the corresponding avatar, identifies the corresponding avatar in the server, and passes the question data to it.

[0366] Step 7:

[0367] The avatar generates answers to questions based on past reviews and purchasing history. The avatar refers to the data it has learned and creates the most appropriate answer to the user's question.

[0368] Step 8:

[0369] The server receives the generated response and sends it back to the device, which also transmits the data back to the device via the Internet connection.

[0370] Step 9:

[0371] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended."

[0372] Step 10:

[0373] Users can make purchasing decisions based on the displayed answers, gaining more reliable information, which can reduce regrets such as "I shouldn't have bought it."

[0374] As described above, the system of the present invention provides users with the information they need through the virtual avatars of trusted reviewers, and supports their purchasing behavior.

[0375] Example 1

[0376] 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."

[0377] Existing online review systems have problems in that it is difficult to easily refer to the opinions of reliable reviewers, and extracting necessary information from a huge number of reviews requires a lot of time and effort. This makes it difficult for users to quickly obtain reliable information regarding purchasing products and services.

[0378] 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.

[0379] In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for preprocessing the collected reviewer data, means for constructing a generative AI model that learns the preprocessed data, means for generating an avatar that imitates the reviewer using the trained generative AI model, and means for passing questions from users to the avatar and returning answers, thereby enabling users to quickly obtain reliable information.

[0380] A "trusted reviewer" is a reviewer who provides a reliable rating based on past reviews and purchasing history.

[0381] "Past reviews" refer to ratings and comments about products or services that a reviewer has previously written.

[0382] "Purchase history" refers to a record of products or services that a reviewer has previously purchased.

[0383] A "generative AI model" is a model generated from training data using artificial intelligence technology.

[0384] "Preprocessing" refers to the process of properly cleansing collected data and performing format conversion, tokenization, etc.

[0385] An "avatar" is a virtual persona created to mimic a reviewer's rating style and opinions.

[0386] "User interface" refers to the screen containing input forms and buttons that a user can operate on a device.

[0387] "Question handing" refers to the process of providing a user's inquiry to an avatar.

[0388] "Returning an answer" refers to the process of providing the user with a response generated from the avatar.

[0389] "Collection methods" refers to the methods and devices used to obtain the required data from databases and online platforms.

[0390] The "means of construction" refers to the methods and processes required to create a generative AI model.

[0391] "Means of generation" refers to the method or process for creating an avatar using a trained generative AI model.

[0392] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0393] Server Operation

[0394] First, the server collects past reviews and purchase history of trusted reviewers. This can be achieved by retrieving data from a database or online platform. For example, the server can use an online store's API to retrieve reviews and purchase history based on a specific reviewer ID. Specifically, the server could use Amazon's API to retrieve the review history of a specific reviewer.

[0395] The server then preprocesses the collected data. This preprocessing includes data cleansing, formatting, and tokenization. For example, it removes HTML tags from text data and splits reviews into individual words. These operations are performed using Python NLP libraries (e.g., NLTK and spaCy).

[0396] The server builds a generative AI model based on the preprocessed data and trains it. A generative AI model such as GPT-3 is used to learn reviewer evaluation styles and comment patterns. Frameworks such as TensorFlow and PyTorch are used to train the model, and multiple epochs are used.

[0397] Using a trained generative AI model, the server generates an avatar that mimics the reviewer's ratings and opinions and generates appropriate answers based on user questions.

[0398] Finally, the server receives the user's question and passes it to the avatar. The question is input into the generative AI model, and the output answer is returned to the user via the server. For example, an API endpoint can be set up using a web framework such as Flask to exchange questions and answers.

[0399] Device behavior

[0400] The device provides a user interface (UI) that allows users to enter questions. This UI includes a text field and a submit button for users to enter questions about specific products or services. When a user enters a question and presses the submit button, the question is sent to the server. Specifically, the UI is created using HTML and JavaScript, and the question is sent to the server via asynchronous communication using a JavaScript Ajax request.

[0401] User behavior

[0402] Users enter questions about the product or service they are considering purchasing into the device's UI. For example, they can enter specific questions such as, "Is this product child-safe?" After entering the question, the user presses a send button to send the question to the server.

[0403] Specific examples

[0404] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0405] The avatar uses past reviewer data to generate a response such as, "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user, allowing the user to make purchasing decisions based on reliable information.

[0406] An example prompt could be, "Is this product safe for children? Please answer based on past reviews and purchasing history from trusted reviewers." Based on this, the generative AI model generates an answer in natural language.

[0407] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0408] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0409] Step 1:

[0410] The server collects past reviews and purchase history from a database or online platform (e.g., an online store's API) based on a trusted reviewer ID. The server receives a specific reviewer ID as input and sends an HTTP request to the database. This request retrieves the reviewer's review history and purchase history in JSON format. This retrieved data is temporarily stored in the server's storage.

[0411] Step 2:

[0412] The server preprocesses the collected review and purchase history data. Specifically, it performs data cleansing (removing unnecessary and duplicate data), format conversion, and tokenization. The server takes the ingested JSON data as input and performs data cleansing and tokenization using a Python NLP library (e.g., NLTK or spaCy). The preprocessed data is then output as a dataset for model training.

[0413] Step 3:

[0414] The server builds and trains a generative AI model based on the preprocessed reviewer data. A generative AI model such as GPT-3 is used for the training process. The server uses the preprocessed reviewer data as input and trains the generative AI model using a machine learning framework such as TensorFlow or PyTorch. This training process outputs a generative model that mimics the reviewer's evaluation style and comment patterns.

[0415] Step 4:

[0416] The server uses the trained model to generate an avatar. This enables the server to generate answers to user questions based on their evaluation style and opinions. The server inputs the trained generative AI model and outputs profile data for the avatar. This profile data includes the results of feature extraction for the reviewer.

[0417] Step 5:

[0418] The device provides a UI that allows users to enter questions. Users enter specific questions about the product or service they are considering purchasing. When a question such as "Is this product child-safe?" is entered, the device treats it as input data and sends it to the server using a JavaScript Ajax request. This request transfers the question data to the server.

[0419] Step 6:

[0420] The server passes the question received from the user to the avatar and generates an answer. The server uses the question data received from the device as input and generates an answer based on the trained generative AI model. The generated answer is output as an HTTP response from the server to the device.

[0421] Step 7:

[0422] The terminal displays the answer received from the server to the user. The terminal receives the HTTP response from the server as input, analyzes its contents, and displays it on the UI. The answer is presented in a format that is easy for the user to understand, and the user can make a purchasing decision based on that information.

[0423] (Application example 1)

[0424] 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."

[0425] On traditional online shopping sites, it is difficult to obtain reliable information about products, especially for new or lesser-known products. Furthermore, highly anonymous online reviews often lack credibility, making it difficult for users to provide sufficient information for making confident purchasing decisions. This creates the risk of users making incorrect purchasing decisions based on inaccurate information. To address these issues, a system is needed that generates avatars based on the ratings and purchasing history of trusted reviewers, allowing users to easily obtain reliable information.

[0426] 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.

[0427] In this invention, the server includes a means for collecting past reviews and purchase histories of reliable reviewers, a means for constructing a generative AI model that learns the collected review data, and a means for generating an avatar that imitates the reviewer using the trained generative AI model, thereby enabling users to obtain reliable information about products on online shopping sites.

[0428] A "trusted reviewer" is someone who has a high rating and credibility based on past reviews and purchasing history.

[0429] "Past ratings and purchasing history" refers to the reviewer's previous ratings of products and services and the history of products they have purchased.

[0430] "Means of collection" refers to the methods used to obtain reliable reviewer ratings and purchasing data using databases or online platforms.

[0431] A "generative AI model" is an artificial intelligence model that learns from collected reviewer data and imitates reviewer evaluation styles and comment patterns.

[0432] An "avatar" is a virtual persona that uses generative AI models to mimic the opinions and ratings of trusted reviewers.

[0433] "User interface" refers to the screen layout and operation method that allows users to easily enter questions and receive answers.

[0434] The "server" is a central computer that collects reviewer data, trains the generative AI model, generates avatars, and handles questions and answers from users.

[0435] A "terminal" is a device used by a user, an electronic device that operates in the form of a smartphone app, etc.

[0436] The system for implementing the present invention operates mainly based on three entities: a server, a terminal, and a user.

[0437] Server Operation

[0438] The server collects the past ratings and purchase history of trusted reviewers. This is achieved by retrieving reviewer data from databases or online platforms. It uses a specific reviewer ID to collect the data of the corresponding reviewer.

[0439] The server then builds a generative AI model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative AI model created here can then mimic the reviewer's opinions and evaluations.

[0440] The server then uses the trained generative AI model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0441] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0442] Device behavior

[0443] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field where users can enter questions about specific products or services, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0444] User behavior

[0445] The user enters a question about a product or service they are considering purchasing. For example, they can enter a specific question such as, "What age group is this toy suitable for?" After entering the question, the user presses a submit button to send the question to the server.

[0446] Specific examples

[0447] For example, consider the case where a user enters the question "What age group is this toy suitable for?" into the device's UI and presses the send button. The device sends this question to the server, which then forwards it to the corresponding reviewer's avatar. Based on past reviewer data, the avatar generates an answer such as "This toy is particularly popular with children aged 3 to 6. It is also very safe." The server receives this answer and sends it back to the device. Finally, the device can display the received answer to the user. This allows the user to make purchasing decisions based on reliable information.

[0448] Prompt Sentence Examples

[0449] Below are some examples of specific prompt sentences.

[0450] text

[0451] Reviewer data: This toy is safe for children aged 3 to 6. It's very popular among parents.

[0452] Question: What age group is this toy suitable for?

[0453] Answer:

[0454] The server uses this to ask questions to the avatar and provides the generated answers to the user. In this way, the system of the present invention provides reliable information via the avatar that mimics the opinions of reliable reviewers, thereby supporting users' purchasing behavior.

[0455] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0456] Step 1: Collect data from trusted reviewers

[0457] The server collects the past ratings and purchase history of trusted reviewers from databases and online platforms. In this process, it uses the reviewer ID as a key to obtain detailed rating data and purchase history data for each reviewer. The input is the reviewer ID and database connection information, and the output is the reviewer's detailed rating data and purchase history.

[0458] Step 2: Training the generative AI model

[0459] The server builds and trains a generative AI model based on the collected reviewer data. In this process, the reviewer's evaluation style and historical purchasing patterns are used as training data to train the AI ​​model (e.g., GPT-2). The input is the reviewer data, and the output is the trained generative AI model.

[0460] Step 3: Creating an avatar

[0461] The server uses a trained generative AI model to generate an avatar that mimics the reviewer. This avatar has the ability to answer questions based on the collected information. The input is the trained generative AI model, and the output is the avatar.

[0462] Step 4: Providing a User Interface

[0463] The terminal provides a user interface (UI) where the user can enter a question. This UI includes a field for the user to enter a question and a submit button. The input is the user's action feedback, and the output is the question entered in the question field.

[0464] Step 5: Sending the question to the server

[0465] The terminal sends the question entered by the user to the server. In this process, the question and the user ID are sent to the server through an HTTP request. The input is the user's question, and the output is the HTTP request sent to the server.

[0466] Step 6: Transfer the question to your avatar

[0467] The server passes the received user question to the corresponding avatar. When the question is passed to the avatar, an appropriate prompt is generated using a generative AI model. The input is the user's question, and the output is the prompt passed to the avatar.

[0468] Step 7: Generate an answer

[0469] The avatar generates answers based on the questions it receives from users. Using a generative AI model, the answers are generated in a way that mimics the rating style of a trusted reviewer. The input is the prompt, and the output is the generated answer.

[0470] Step 8: Sending the answer back to the server

[0471] The avatar then sends the generated answer back to the server, where it is ready to be sent back to the user. The input is the generated answer, and the output is the answer sent to the server.

[0472] Step 9: Providing answers to users

[0473] The server sends the generated answer back to the user's device, where it can be viewed by the user. The device displays the answer received from the server in a user interface. The input is the answer from the server, and the output is the answer displayed on the device's display screen.

[0474] At each step, users can get answers from their avatars based on the opinions of trusted reviewers, allowing them to make purchasing decisions with confidence.

[0475] 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.

[0476] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to ask questions to the avatar to obtain reliable information. It also combines an emotion engine that recognizes the user's emotions and provides answers that correspond to the user's emotions. This system operates appropriately mainly through three entities: a server, a terminal, and a user, as well as the emotion engine.

[0477] Server Operation

[0478] First, the server collects the past reviews and purchase history of trusted reviewers from databases or online platforms, and then retrieves the required data using a specific reviewer ID.

[0479] The server then builds and trains a generative model based on the collected data. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The trained generative model can faithfully mimic the reviewer's opinions and evaluations.

[0480] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and answers questions from users.

[0481] The server then uses an emotion engine to recognize the user's emotions, which analyzes the user's input text and voice to identify their current emotions.

[0482] Device behavior

[0483] The device provides a user interface (UI) that allows the user to enter a question. The UI includes a field for the user to enter a question about a particular product or service and a submit button. When the user enters a question, the device can also capture the user's voice and text sentiment and transmit it to the server.

[0484] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0485] Collaboration between the server and emotion engine

[0486] The server receives the user's question and emotion data, then passes the question to the avatar to generate an answer. At the same time, the emotion engine analyzes the emotion data and identifies the user's emotional state. For example, if the user is feeling anxious, that emotion is communicated to the server.

[0487] The emotional state of the user analyzed by the emotion engine influences the avatar's responses. For example, if the user is feeling anxious, the avatar will adjust its responses to take that into consideration. For example, the answer to the question, "Is this product safe for children?" will be "Yes, it is safe. However, we also have some recommendations for further safety checks." This will add information to alleviate anxiety.

[0488] Specific examples

[0489] For example, suppose a user emotionally inputs the question "Is this product safe for children?" and speaks it in a tone that indicates anxiety. The device captures the user's voice data along with the question and sends it to the server. The server then sends the question to the avatar, while simultaneously analyzing the emotional data with its emotion engine and recognizing the user's anxiety.

[0490] The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." It then receives information from the emotion engine that the user is feeling uneasy and responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0491] The device displays this response to the user, allowing the user to receive reliable information and a sense of security at the same time. In this way, the system of the present invention provides reliable information that reflects the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0492] The processing flow will be explained below.

[0493] Step 1:

[0494] The server collects the past reviews and purchase history of trusted reviewers from databases and online platforms, and then retrieves the required data using a specific reviewer ID.

[0495] Step 2:

[0496] The server builds and trains a generative model based on the collected reviewer data. The model learns and understands reviewer evaluation styles and comment patterns.

[0497] Step 3:

[0498] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to provide answers to user questions.

[0499] Step 4:

[0500] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0501] Step 5:

[0502] When users input questions, the device also captures text and voice data to recognize users' emotions, and has a built-in emotion engine to analyze users' emotions.

[0503] Step 6:

[0504] When a user enters a question and presses the send button, the device sends the question and captured emotion data to the server via an internet connection.

[0505] Step 7:

[0506] The server receives the question and emotion data from the user, and first passes the question to the avatar to generate an answer.

[0507] Step 8:

[0508] The server then uses an emotion engine to analyze the emotion data and identify the user's emotional state, for example, the emotion engine may recognize that the user is feeling anxious.

[0509] Step 9:

[0510] The emotional state of the user analyzed by the emotion engine influences the avatar's response. Specifically, the avatar takes into account the emotional data and generates responses that correspond to the user's emotions.

[0511] Step 10:

[0512] For example, if a user types the question "Is this product child-safe?" with some anxiety, the avatar will generate the answer "This product is child-safe, but we recommend age-appropriate use. Regular maintenance and child supervision are also important as additional safety precautions."

[0513] Step 11:

[0514] The server receives the generated response and sends it back to the device, which also sends the data back to the device via the Internet connection.

[0515] Step 12:

[0516] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended. Regular maintenance and child supervision are also important as additional safety precautions."

[0517] Step 13:

[0518] Users can make purchasing decisions based on the displayed answers, obtaining more reliable information, which reduces the chance of users regretting their purchase.

[0519] As described above, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0520] Example 2

[0521] 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."

[0522] Today's consumers seek reliable information about products and services online, but online reviews and ratings often lack credibility. Furthermore, when users consult online reviews, they face challenges in finding advice that is tailored to their emotions and circumstances. Furthermore, information overload and conflicting reviews make it difficult for users to make appropriate decisions.

[0523] 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 acquiring past evaluation information and transaction history of reliable reviewers, means for constructing a generative model for learning the acquired reviewer data, means for generating a virtual agent that mimics the reviewer using the learned generative model, means for passing questions from a user to the virtual agent and providing the answers, means for identifying an emotional state from the user's input, and means for adjusting the answers of the virtual agent based on the identified emotional state. This allows the user to receive highly reliable information as well as advice that corresponds to their current emotions.

[0524] A "trusted reviewer" is a reviewer who has provided accurate and fair ratings in the past and has received high ratings from many users.

[0525] "Rating Information" refers to the content of the review or feedback provided by the reviewer regarding a product or service.

[0526] "Transaction History" refers to records of products or services that a reviewer has previously purchased or used.

[0527] "Capture methods" refers to software and processes used to collect data from databases and external platforms.

[0528] "Reviewer Data" refers collectively to rating information and transaction history associated with trusted reviewers.

[0529] A "generative model" refers to an artificial intelligence algorithm or mechanism that learns from specific data and generates new data.

[0530] A "virtual agent" is a virtual entity that uses a trained generative model to mimic a specific reviewer and provide answers to their questions.

[0531] "User input" refers to data input, such as text or voice, made by a user to a system.

[0532] "Means for identifying emotional state" refers to the algorithms or engines used to analyze emotions from user input and identify that state.

[0533] "Response tailoring" refers to a process or algorithm for appropriately modifying the responses provided based on the identified emotional state.

[0534] This invention is a system that generates a virtual agent based on the past evaluation information and transaction history of trusted reviewers, and allows users to ask the virtual agent questions to obtain reliable information. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it provides answers that correspond to the user's emotions. This system consists of three entities: a server, a terminal, and a user.

[0535] The server collects the past rating information and transaction history of trusted reviewers from a database or online platform, for example, using SQL Server or a NoSQL database (e.g., MySQL, MongoDB), and retrieves the required data using the reviewer ID.

[0536] Next, the server builds a generative AI model (e.g., GPT-4) based on the collected data and trains it. During this training process, a GPU is used to quickly process large amounts of data, and reviewers' evaluation styles and comment patterns are used as training data. This creates a generative AI model that faithfully mimics the reviewers' opinions and evaluations.

[0537] The server then uses the trained generative AI model to generate a virtual agent that mimics the reviewer, which can then use the generative AI model to generate appropriate answers to user questions.

[0538] The device provides a user interface (UI) that allows users to input questions. Specifically, the UI includes a text input field and a submit button. This allows users to easily input and submit questions about specific products or services. The device can also capture and send emotional data from the user's voice and text to a server. Voice recognition software (e.g., Google Speech-to-Text) is used for voice recognition.

[0539] When a user enters a question and presses the send button, the device sends the question and the captured emotion data to the server. The server then passes the received question to a virtual agent, which generates an answer. In parallel, the server analyzes the emotion data using an emotion engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. For example, if the user is feeling anxious, the server will recognize that emotion.

[0540] The server adjusts the generated answer based on the user's emotional state analyzed by the emotion engine. For example, if the user is feeling anxious, the server generates an answer that takes their emotion into consideration. In response to the question, "Is this product child-safe?", the virtual agent generates the answer, "This product is child-safe, but we recommend age-appropriate use," and then adds reassuring information such as, "Additional safety measures include regular maintenance and child supervision."

[0541] The device displays the final answer received from the server to the user, allowing the user to obtain reliable information and an appropriate answer that reflects their current emotions.

[0542] As a concrete example, if a user types the question "Is this product child-safe?" in an anxious tone, the following prompt sentence is generated:

[0543] Example prompt: "This product is child-safe, but age-appropriate use is recommended. Additionally, regular maintenance and child supervision are important additional safety precautions."

[0544] In this way, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual agent of a trusted reviewer, thereby supporting the user's decision-making.

[0545] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0546] Step 1:

[0547] The server collects data on trusted reviewers. Database connection information and reviewer IDs are required as input. Using a database system (e.g., MySQL or MongoDB), past review information and transaction history are queried based on the reviewer ID. This query process outputs the reviews and purchase records left by the reviewer.

[0548] Step 2:

[0549] The server trains a generative AI model based on the data collected. The reviewer data collected in step 1 is used as input. The data is input to the generative AI model (e.g., GPT-4) using a GPU, and the training process is carried out. Through this process, the generative AI model learns the reviewer's evaluation style and comment patterns, and outputs the trained model.

[0550] Step 3:

[0551] The server generates a virtual agent using the generative AI model. The generative AI model trained in step 2 is used as input. Based on the trained model, a virtual agent that imitates the reviewer's style is generated and output as an agent that can be used within the system.

[0552] Step 4:

[0553] The device provides a user interface (UI), which includes a text entry field and a submit button. The user enters and submits a question about a product or service. Inputs include the user's text and voice data. The output of this step is the user's question in text format.

[0554] Step 5:

[0555] The user enters a question and presses the submit button. This action causes the user's text and voice data to be stored on the device. The voice data is converted to text using voice recognition software (e.g., Google Speech-to-Text). The input requires the user's question, and the output is converted to text format.

[0556] Step 6:

[0557] The device sends the question and emotion data to the server. The input includes the converted question and speech data stored in step 5. By sending this data to the server, the server receives the question content and emotion data and executes the next step based on them.

[0558] Step 7:

[0559] The server generates an answer to the user's question. The input includes the question text received in step 6. Using the trained generative AI model, it generates an answer based on the question content. The generated answer is based on the virtual agent's mimicking style, and the output is the answer text.

[0560] Step 8:

[0561] The server analyzes the emotion data and reflects it in the answer. The emotion data received in step 6 is included as input. An emotion engine (e.g., IBM Watson Tone Analyzer) is used to analyze the data and identify the emotional state. Based on this emotional state, the answer generated in step 7 is adjusted to output a final answer that takes the user's emotions into consideration.

[0562] Step 9:

[0563] The server sends the final answer to the terminal. As input, the final answer text adjusted in the previous step is required. By sending this data to the terminal, the terminal receives and displays the final answer.

[0564] Step 10:

[0565] The device displays the final answer to the user. The input includes the final answer text received from the server. This is displayed in the user interface, allowing the user to obtain an appropriate answer based on their emotions along with reliable information. The output is the user's ability to view the final answer.

[0566] (Application example 2)

[0567] 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."

[0568] In conventional review systems, users have had difficulty finding reliable information from the vast number of reviews, and there are few ways to obtain appropriate information that matches their own emotions. This invention aims to solve these problems by generating an avatar based on the past reviews and purchase history of reliable reviewers, and providing answers that match the user's emotions.

[0569] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for constructing a generative model that learns the collected review data, means for generating an avatar that imitates the reviewer using the learned generative model, means for adjusting answers according to the user's emotions using an emotion engine that recognizes the user's emotions, and means for passing questions from the user to the avatar and returning the answers. This makes it possible to provide reliable information and obtain appropriate answers according to the user's emotions.

[0570] A "trusted reviewer" is one whose ratings and opinions are accurate and consistent based on past reviews and purchasing history.

[0571] A "review" is an evaluation or comment posted by a user about a specific product or service based on their own experience and opinion.

[0572] "Purchase history" is a record of products and services a user has purchased in the past.

[0573] A "generative model" is an artificial intelligence model that learns from collected data and generates an appropriate output for any input.

[0574] An "avatar" is a virtual entity that mimics the opinions and ratings of trusted reviewers and answers questions from users.

[0575] "User emotion" refers to the user's psychological state or feelings during a question or interaction.

[0576] An "emotion engine" is an algorithm or software that analyzes a user's text or voice data to identify their emotions.

[0577] A "question" is an action in which a user enters a question or request for confirmation regarding a specific product or service.

[0578] An "answer" is information or an opinion provided by an avatar in response to a question from a user.

[0579] The system for implementing this invention is mainly composed of three entities: a server, a terminal, and a user. The function of each entity is explained below.

[0580] Server Operation

[0581] The server operates using the following means:

[0582] 1. Data collection methods:

[0583] The past reviews and purchase history of trusted reviewers are collected from databases and online platforms, and the collected data is used to extract the necessary information using a specific reviewer ID.

[0584] 2. Generative model construction method:

[0585] A generative AI model is constructed based on the collected review data. This generative model uses past reviews and purchase history as training data to learn the reviewer's evaluation style and comment patterns. After training, the generative model is able to faithfully imitate the reviewer's opinions and evaluations.

[0586] 3. Avatar creation method:

[0587] Using a trained generative model, we generate an avatar that mimics the reviewer, and this avatar is used to answer questions from users.

[0588] 4. Emotion recognition means:

[0589] The emotion engine analyzes the user's input text and voice to identify their current emotion.

[0590] This allows the server to generate a virtual avatar of a trusted reviewer and provide answers tailored to the user's emotions.

[0591] Device behavior

[0592] The terminal operates as an interface with the user as follows.

[0593] 1. User interface provision means:

[0594] It provides an interface where users can enter questions. The UI includes a field to enter questions about a specific product or service, a submit button, and the ability to capture user voice and text sentiment.

[0595] 2. Means of data transmission:

[0596] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0597] Examples of emotion-assisted answers

[0598] Specific examples are shown below.

[0599] 1. The user emotionally types the question "Is this product child safe?" and speaks it in a tone that indicates anxiety.

[0600] 2. The device captures the user's voice data along with this question and sends it to the server.

[0601] 3. The server sends a question to the avatar, and at the same time, it analyzes the emotional data using an emotion engine to recognize the user's anxiety.

[0602] 4. The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." Furthermore, based on the emotion engine's information that the user is feeling uneasy, the avatar responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0603] 5. The device will display this answer to the user, providing them with reliable information and a sense of security.

[0604] Prompt Sentence Examples

[0605] Below are some example prompts to provide to the generative AI model:

[0606] User Question: "Is this product safe for children?"

[0607] User sentiment: "Anxiety"

[0608] Avatar response to be generated: "This product is child-safe, however, regular maintenance and child supervision are important additional safety precautions."

[0609] This system allows users to obtain reliable information and receive appropriate answers based on their own emotions.

[0610] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0611] Step 1: Data collection

[0612] The server collects past reviews and purchase history from trusted reviewers. The input is review data and purchase history data obtained using reviewer IDs. The server collects this data from databases and online platforms and stores it in a database. The output is the collected review data and purchase history data.

[0613] Step 2: Building a generative model

[0614] The server builds a generative AI model based on the collected data. Past reviews and purchase history are used as input. The server uses this data as training data and uses a machine learning algorithm to learn the reviewer's evaluation style and comment patterns. The output is a trained generative AI model.

[0615] Step 3: Create an avatar

[0616] The server generates an avatar using a trained generative model. The inputs are the trained generative model and review data. The server uses these data to generate a virtual avatar that closely mimics the reviewer. The output is the avatar.

[0617] Step 4: Submit your question and sentiment data

[0618] A user inputs a question about a product or service into the device's UI. The device captures the user's voice and text emotion data along with the input question and sends it to the server. The input is the user's question and emotion data, and the output is data sent to the server.

[0619] Step 5: Emotion Recognition

[0620] The server analyzes the emotion data sent by the user. The input is the user's emotion data, which is analyzed using the server's emotion engine. The emotion engine analyzes the text and / or voice to determine the user's current emotional state. The output is the analyzed emotional state.

[0621] Step 6: Answer Generation

[0622] The server passes the user's question to the avatar, which generates an answer. The input is the user's question and the analyzed emotional state. The avatar generates an answer based on this. In doing so, it adds adaptive information according to the emotion based on data from the emotion engine. For example, an answer containing additional information that gives a sense of security to an anxious user is generated. The output is the generated answer.

[0623] Step 7: Submit and view your responses

[0624] The server sends the generated answer to the user's device. The device receives the answer and displays it to the user. The input is the generated answer, and the output is the answer displayed to the user. This allows the user to obtain reliable information as well as an appropriate answer based on their emotions.

[0625] 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.

[0626] 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.

[0627] 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.

[0628] [Third embodiment]

[0629] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0630] 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.

[0631] 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).

[0632] 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.

[0633] 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.

[0634] 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).

[0635] 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. 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.

[0636] 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.

[0637] 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.

[0638] 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.

[0639] 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.

[0640] 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."

[0641] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0642] Server Operation

[0643] First, the server collects the past reviews and purchase history of trusted reviewers. This is achieved by pulling data from databases or online platforms. It retrieves the reviewer's data using a specific reviewer ID.

[0644] Next, the server builds a generative model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative model created here can then mimic the reviewer's opinions and evaluations.

[0645] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0646] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0647] Device behavior

[0648] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field for users to enter questions about a particular product or service, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0649] User behavior

[0650] Users enter questions about the product or service they are considering purchasing. For example, they can enter specific questions such as, "Is this product child-safe?" After entering their question, the user presses a submit button to send the question to the server.

[0651] Specific examples

[0652] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0653] The avatar generates a response based on past reviewer data: "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user. As a result, the user can make purchasing decisions based on reliable information.

[0654] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0655] The processing flow will be explained below.

[0656] Step 1:

[0657] The server collects past reviews and purchase history of trusted reviewers from databases and online platforms, and retrieves the required data using a specific reviewer ID.

[0658] Step 2:

[0659] The server builds a generative model based on the collected reviewer data, and the model is trained to learn reviewers' evaluation styles and comments and understand their patterns.

[0660] Step 3:

[0661] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to answer questions from users.

[0662] Step 4:

[0663] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0664] Step 5:

[0665] When a user enters a question and presses the send button, the device sends the question to the server, which transmits the data from the device to the server via an Internet connection.

[0666] Step 6:

[0667] The server receives a question from the user and passes the question to the corresponding avatar, identifies the corresponding avatar in the server, and passes the question data to it.

[0668] Step 7:

[0669] The avatar generates answers to questions based on past reviews and purchasing history. The avatar refers to the data it has learned and creates the most appropriate answer to the user's question.

[0670] Step 8:

[0671] The server receives the generated response and sends it back to the device, which also transmits the data back to the device via the Internet connection.

[0672] Step 9:

[0673] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended."

[0674] Step 10:

[0675] Users can make purchasing decisions based on the displayed answers, gaining more reliable information, which can reduce regrets such as "I shouldn't have bought it."

[0676] As described above, the system of the present invention provides users with the information they need through the virtual avatars of trusted reviewers, and supports their purchasing behavior.

[0677] Example 1

[0678] 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."

[0679] Existing online review systems have problems in that it is difficult to easily refer to the opinions of reliable reviewers, and extracting necessary information from a huge number of reviews requires a lot of time and effort. This makes it difficult for users to quickly obtain reliable information regarding purchasing products and services.

[0680] 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.

[0681] In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for preprocessing the collected reviewer data, means for constructing a generative AI model that learns the preprocessed data, means for generating an avatar that imitates the reviewer using the trained generative AI model, and means for passing questions from users to the avatar and returning answers, thereby enabling users to quickly obtain reliable information.

[0682] A "trusted reviewer" is a reviewer who provides a reliable rating based on past reviews and purchasing history.

[0683] "Past reviews" refer to ratings and comments about products or services that a reviewer has previously written.

[0684] "Purchase history" refers to a record of products or services that a reviewer has previously purchased.

[0685] A "generative AI model" is a model generated from training data using artificial intelligence technology.

[0686] "Preprocessing" refers to the process of properly cleansing collected data and performing format conversion, tokenization, etc.

[0687] An "avatar" is a virtual persona created to mimic a reviewer's rating style and opinions.

[0688] "User interface" refers to the screen containing input forms and buttons that a user can operate on a device.

[0689] "Question handing" refers to the process of providing a user's inquiry to an avatar.

[0690] "Returning an answer" refers to the process of providing the user with a response generated from the avatar.

[0691] "Collection methods" refers to the methods and devices used to obtain the required data from databases and online platforms.

[0692] The "means of construction" refers to the methods and processes required to create a generative AI model.

[0693] "Means of generation" refers to the method or process for creating an avatar using a trained generative AI model.

[0694] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0695] Server Operation

[0696] First, the server collects past reviews and purchase history of trusted reviewers. This can be achieved by retrieving data from a database or online platform. For example, the server can use an online store's API to retrieve reviews and purchase history based on a specific reviewer ID. Specifically, the server could use Amazon's API to retrieve the review history of a specific reviewer.

[0697] The server then preprocesses the collected data. This preprocessing includes data cleansing, formatting, and tokenization. For example, it removes HTML tags from text data and splits reviews into individual words. These operations are performed using Python NLP libraries (e.g., NLTK and spaCy).

[0698] The server builds a generative AI model based on the preprocessed data and trains it. A generative AI model such as GPT-3 is used to learn reviewer evaluation styles and comment patterns. Frameworks such as TensorFlow and PyTorch are used to train the model, and multiple epochs are used.

[0699] Using a trained generative AI model, the server generates an avatar that mimics the reviewer's ratings and opinions and generates appropriate answers based on user questions.

[0700] Finally, the server receives the user's question and passes it to the avatar. The question is input into the generative AI model, and the output answer is returned to the user via the server. For example, an API endpoint can be set up using a web framework such as Flask to exchange questions and answers.

[0701] Device behavior

[0702] The device provides a user interface (UI) that allows users to enter questions. This UI includes a text field and a submit button for users to enter questions about specific products or services. When a user enters a question and presses the submit button, the question is sent to the server. Specifically, the UI is created using HTML and JavaScript, and the question is sent to the server via asynchronous communication using a JavaScript Ajax request.

[0703] User behavior

[0704] Users enter questions about the product or service they are considering purchasing into the device's UI. For example, they can enter specific questions such as, "Is this product child-safe?" After entering the question, the user presses a send button to send the question to the server.

[0705] Specific examples

[0706] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0707] The avatar uses past reviewer data to generate a response such as, "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user, allowing the user to make purchasing decisions based on reliable information.

[0708] An example prompt could be, "Is this product safe for children? Please answer based on past reviews and purchasing history from trusted reviewers." Based on this, the generative AI model generates an answer in natural language.

[0709] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0710] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0711] Step 1:

[0712] The server collects past reviews and purchase history from a database or online platform (e.g., an online store's API) based on a trusted reviewer ID. The server receives a specific reviewer ID as input and sends an HTTP request to the database. This request retrieves the reviewer's review history and purchase history in JSON format. This retrieved data is temporarily stored in the server's storage.

[0713] Step 2:

[0714] The server preprocesses the collected review and purchase history data. Specifically, it performs data cleansing (removing unnecessary and duplicate data), format conversion, and tokenization. The server takes the ingested JSON data as input and performs data cleansing and tokenization using a Python NLP library (e.g., NLTK or spaCy). The preprocessed data is then output as a dataset for model training.

[0715] Step 3:

[0716] The server builds and trains a generative AI model based on the preprocessed reviewer data. A generative AI model such as GPT-3 is used for the training process. The server uses the preprocessed reviewer data as input and trains the generative AI model using a machine learning framework such as TensorFlow or PyTorch. This training process outputs a generative model that mimics the reviewer's evaluation style and comment patterns.

[0717] Step 4:

[0718] The server uses the trained model to generate an avatar. This enables the server to generate answers to user questions based on their evaluation style and opinions. The server inputs the trained generative AI model and outputs profile data for the avatar. This profile data includes the results of feature extraction for the reviewer.

[0719] Step 5:

[0720] The device provides a UI that allows users to enter questions. Users enter specific questions about the product or service they are considering purchasing. When a question such as "Is this product child-safe?" is entered, the device treats it as input data and sends it to the server using a JavaScript Ajax request. This request transfers the question data to the server.

[0721] Step 6:

[0722] The server passes the question received from the user to the avatar and generates an answer. The server uses the question data received from the device as input and generates an answer based on the trained generative AI model. The generated answer is output as an HTTP response from the server to the device.

[0723] Step 7:

[0724] The terminal displays the answer received from the server to the user. The terminal receives the HTTP response from the server as input, analyzes its contents, and displays it on the UI. The answer is presented in a format that is easy for the user to understand, and the user can make a purchasing decision based on that information.

[0725] (Application example 1)

[0726] 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."

[0727] On traditional online shopping sites, it is difficult to obtain reliable information about products, especially for new or lesser-known products. Furthermore, highly anonymous online reviews often lack credibility, making it difficult for users to provide sufficient information for making confident purchasing decisions. This creates the risk of users making incorrect purchasing decisions based on inaccurate information. To address these issues, a system is needed that generates avatars based on the ratings and purchasing history of trusted reviewers, allowing users to easily obtain reliable information.

[0728] 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.

[0729] In this invention, the server includes a means for collecting past reviews and purchase histories of reliable reviewers, a means for constructing a generative AI model that learns the collected review data, and a means for generating an avatar that imitates the reviewer using the trained generative AI model, thereby enabling users to obtain reliable information about products on online shopping sites.

[0730] A "trusted reviewer" is someone who has a high rating and credibility based on past reviews and purchasing history.

[0731] "Past ratings and purchasing history" refers to the reviewer's previous ratings of products and services and the history of products they have purchased.

[0732] "Means of collection" refers to the methods used to obtain reliable reviewer ratings and purchasing data using databases or online platforms.

[0733] A "generative AI model" is an artificial intelligence model that learns from collected reviewer data and imitates reviewer evaluation styles and comment patterns.

[0734] An "avatar" is a virtual persona that uses generative AI models to mimic the opinions and ratings of trusted reviewers.

[0735] "User interface" refers to the screen layout and operation method that allows users to easily enter questions and receive answers.

[0736] The "server" is a central computer that collects reviewer data, trains the generative AI model, generates avatars, and handles questions and answers from users.

[0737] A "terminal" is a device used by a user, an electronic device that operates in the form of a smartphone app, etc.

[0738] The system for implementing the present invention operates mainly based on three entities: a server, a terminal, and a user.

[0739] Server Operation

[0740] The server collects the past ratings and purchase history of trusted reviewers. This is achieved by retrieving reviewer data from databases or online platforms. It uses a specific reviewer ID to collect the data of the corresponding reviewer.

[0741] The server then builds a generative AI model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative AI model created here can then mimic the reviewer's opinions and evaluations.

[0742] The server then uses the trained generative AI model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0743] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0744] Device behavior

[0745] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field where users can enter questions about specific products or services, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0746] User behavior

[0747] The user enters a question about a product or service they are considering purchasing. For example, they can enter a specific question such as, "What age group is this toy suitable for?" After entering the question, the user presses a submit button to send the question to the server.

[0748] Specific examples

[0749] For example, consider the case where a user enters the question "What age group is this toy suitable for?" into the device's UI and presses the send button. The device sends this question to the server, which then forwards it to the corresponding reviewer's avatar. Based on past reviewer data, the avatar generates an answer such as "This toy is particularly popular with children aged 3 to 6. It is also very safe." The server receives this answer and sends it back to the device. Finally, the device can display the received answer to the user. This allows the user to make purchasing decisions based on reliable information.

[0750] Prompt Sentence Examples

[0751] Below are some examples of specific prompt sentences.

[0752] text

[0753] Reviewer data: This toy is safe for children aged 3 to 6. It's very popular among parents.

[0754] Question: What age group is this toy suitable for?

[0755] Answer:

[0756] The server uses this to ask questions to the avatar and provides the generated answers to the user. In this way, the system of the present invention provides reliable information via the avatar that mimics the opinions of reliable reviewers, thereby supporting users' purchasing behavior.

[0757] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0758] Step 1: Collect data from trusted reviewers

[0759] The server collects the past ratings and purchase history of trusted reviewers from databases and online platforms. In this process, it uses the reviewer ID as a key to obtain detailed rating data and purchase history data for each reviewer. The input is the reviewer ID and database connection information, and the output is the reviewer's detailed rating data and purchase history.

[0760] Step 2: Training the generative AI model

[0761] The server builds and trains a generative AI model based on the collected reviewer data. In this process, the reviewer's evaluation style and historical purchasing patterns are used as training data to train the AI ​​model (e.g., GPT-2). The input is the reviewer data, and the output is the trained generative AI model.

[0762] Step 3: Creating an avatar

[0763] The server uses a trained generative AI model to generate an avatar that mimics the reviewer. This avatar has the ability to answer questions based on the collected information. The input is the trained generative AI model, and the output is the avatar.

[0764] Step 4: Providing a User Interface

[0765] The terminal provides a user interface (UI) where the user can enter a question. This UI includes a field for the user to enter a question and a submit button. The input is the user's action feedback, and the output is the question entered in the question field.

[0766] Step 5: Sending the question to the server

[0767] The terminal sends the question entered by the user to the server. In this process, the question and the user ID are sent to the server through an HTTP request. The input is the user's question, and the output is the HTTP request sent to the server.

[0768] Step 6: Transfer the question to your avatar

[0769] The server passes the received user question to the corresponding avatar. When the question is passed to the avatar, an appropriate prompt is generated using a generative AI model. The input is the user's question, and the output is the prompt passed to the avatar.

[0770] Step 7: Generate an answer

[0771] The avatar generates answers based on the questions it receives from users. Using a generative AI model, the answers are generated in a way that mimics the rating style of a trusted reviewer. The input is the prompt, and the output is the generated answer.

[0772] Step 8: Sending the answer back to the server

[0773] The avatar then sends the generated answer back to the server, where it is ready to be sent back to the user. The input is the generated answer, and the output is the answer sent to the server.

[0774] Step 9: Providing answers to users

[0775] The server sends the generated answer back to the user's device, where it can be viewed by the user. The device displays the answer received from the server in a user interface. The input is the answer from the server, and the output is the answer displayed on the device's display screen.

[0776] At each step, users can get answers from their avatars based on the opinions of trusted reviewers, allowing them to make purchasing decisions with confidence.

[0777] 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.

[0778] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to ask questions to the avatar to obtain reliable information. It also combines an emotion engine that recognizes the user's emotions and provides answers that correspond to the user's emotions. This system operates appropriately mainly through three entities: a server, a terminal, and a user, as well as the emotion engine.

[0779] Server Operation

[0780] First, the server collects the past reviews and purchase history of trusted reviewers from databases or online platforms, and then retrieves the required data using a specific reviewer ID.

[0781] The server then builds and trains a generative model based on the collected data. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The trained generative model can faithfully mimic the reviewer's opinions and evaluations.

[0782] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and answers questions from users.

[0783] The server then uses an emotion engine to recognize the user's emotions, which analyzes the user's input text and voice to identify their current emotions.

[0784] Device behavior

[0785] The device provides a user interface (UI) that allows the user to enter a question. The UI includes a field for the user to enter a question about a particular product or service and a submit button. When the user enters a question, the device can also capture the user's voice and text sentiment and transmit it to the server.

[0786] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0787] Collaboration between the server and emotion engine

[0788] The server receives the user's question and emotion data, then passes the question to the avatar to generate an answer. At the same time, the emotion engine analyzes the emotion data and identifies the user's emotional state. For example, if the user is feeling anxious, that emotion is communicated to the server.

[0789] The emotional state of the user analyzed by the emotion engine influences the avatar's responses. For example, if the user is feeling anxious, the avatar will adjust its responses to take that into consideration. For example, the answer to the question, "Is this product safe for children?" will be "Yes, it is safe. However, we also have some recommendations for further safety checks." This will add information to alleviate anxiety.

[0790] Specific examples

[0791] For example, suppose a user emotionally inputs the question "Is this product safe for children?" and speaks it in a tone that indicates anxiety. The device captures the user's voice data along with the question and sends it to the server. The server then sends the question to the avatar, while simultaneously analyzing the emotional data with its emotion engine and recognizing the user's anxiety.

[0792] The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." It then receives information from the emotion engine that the user is feeling uneasy and responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0793] The device displays this response to the user, allowing the user to receive reliable information and a sense of security at the same time. In this way, the system of the present invention provides reliable information that reflects the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0794] The processing flow will be explained below.

[0795] Step 1:

[0796] The server collects the past reviews and purchase history of trusted reviewers from databases and online platforms, and then retrieves the required data using a specific reviewer ID.

[0797] Step 2:

[0798] The server builds and trains a generative model based on the collected reviewer data. The model learns and understands reviewer evaluation styles and comment patterns.

[0799] Step 3:

[0800] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to provide answers to user questions.

[0801] Step 4:

[0802] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0803] Step 5:

[0804] When users input questions, the device also captures text and voice data to recognize users' emotions, and has a built-in emotion engine to analyze users' emotions.

[0805] Step 6:

[0806] When a user enters a question and presses the send button, the device sends the question and captured emotion data to the server via an internet connection.

[0807] Step 7:

[0808] The server receives the question and emotion data from the user, and first passes the question to the avatar to generate an answer.

[0809] Step 8:

[0810] The server then uses an emotion engine to analyze the emotion data and identify the user's emotional state, for example, the emotion engine may recognize that the user is feeling anxious.

[0811] Step 9:

[0812] The emotional state of the user analyzed by the emotion engine influences the avatar's response. Specifically, the avatar takes into account the emotional data and generates responses that correspond to the user's emotions.

[0813] Step 10:

[0814] For example, if a user types the question "Is this product child-safe?" with some anxiety, the avatar will generate the answer "This product is child-safe, but we recommend age-appropriate use. Regular maintenance and child supervision are also important as additional safety precautions."

[0815] Step 11:

[0816] The server receives the generated response and sends it back to the device, which also sends the data back to the device via the Internet connection.

[0817] Step 12:

[0818] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended. Regular maintenance and child supervision are also important as additional safety precautions."

[0819] Step 13:

[0820] Users can make purchasing decisions based on the displayed answers, obtaining more reliable information, which reduces the chance of users regretting their purchase.

[0821] As described above, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[0822] Example 2

[0823] 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."

[0824] Today's consumers seek reliable information about products and services online, but online reviews and ratings often lack credibility. Furthermore, when users consult online reviews, they face challenges in finding advice that is tailored to their emotions and circumstances. Furthermore, information overload and conflicting reviews make it difficult for users to make appropriate decisions.

[0825] 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 acquiring past evaluation information and transaction history of reliable reviewers, means for constructing a generative model for learning the acquired reviewer data, means for generating a virtual agent that mimics the reviewer using the learned generative model, means for passing questions from a user to the virtual agent and providing the answers, means for identifying an emotional state from the user's input, and means for adjusting the answers of the virtual agent based on the identified emotional state. This allows the user to receive highly reliable information as well as advice that corresponds to their current emotions.

[0826] A "trusted reviewer" is a reviewer who has provided accurate and fair ratings in the past and has received high ratings from many users.

[0827] "Rating Information" refers to the content of the review or feedback provided by the reviewer regarding a product or service.

[0828] "Transaction History" refers to records of products or services that a reviewer has previously purchased or used.

[0829] "Capture methods" refers to software and processes used to collect data from databases and external platforms.

[0830] "Reviewer Data" refers collectively to rating information and transaction history associated with trusted reviewers.

[0831] A "generative model" refers to an artificial intelligence algorithm or mechanism that learns from specific data and generates new data.

[0832] A "virtual agent" is a virtual entity that uses a trained generative model to mimic a specific reviewer and provide answers to their questions.

[0833] "User input" refers to data input, such as text or voice, made by a user to a system.

[0834] "Means for identifying emotional state" refers to the algorithms or engines used to analyze emotions from user input and identify that state.

[0835] "Response tailoring" refers to a process or algorithm for appropriately modifying the responses provided based on the identified emotional state.

[0836] This invention is a system that generates a virtual agent based on the past evaluation information and transaction history of trusted reviewers, and allows users to ask the virtual agent questions to obtain reliable information. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it provides answers that correspond to the user's emotions. This system consists of three entities: a server, a terminal, and a user.

[0837] The server collects the past rating information and transaction history of trusted reviewers from a database or online platform, for example, using SQL Server or a NoSQL database (e.g., MySQL, MongoDB), and retrieves the required data using the reviewer ID.

[0838] Next, the server builds a generative AI model (e.g., GPT-4) based on the collected data and trains it. During this training process, a GPU is used to quickly process large amounts of data, and reviewers' evaluation styles and comment patterns are used as training data. This creates a generative AI model that faithfully mimics the reviewers' opinions and evaluations.

[0839] The server then uses the trained generative AI model to generate a virtual agent that mimics the reviewer, which can then use the generative AI model to generate appropriate answers to user questions.

[0840] The device provides a user interface (UI) that allows users to input questions. Specifically, the UI includes a text input field and a submit button. This allows users to easily input and submit questions about specific products or services. The device can also capture and send emotional data from the user's voice and text to a server. Voice recognition software (e.g., Google Speech-to-Text) is used for voice recognition.

[0841] When a user enters a question and presses the send button, the device sends the question and the captured emotion data to the server. The server then passes the received question to a virtual agent, which generates an answer. In parallel, the server analyzes the emotion data using an emotion engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. For example, if the user is feeling anxious, the server will recognize that emotion.

[0842] The server adjusts the generated answer based on the user's emotional state analyzed by the emotion engine. For example, if the user is feeling anxious, the server generates an answer that takes their emotion into consideration. In response to the question, "Is this product child-safe?", the virtual agent generates the answer, "This product is child-safe, but we recommend age-appropriate use," and then adds reassuring information such as, "Additional safety measures include regular maintenance and child supervision."

[0843] The device displays the final answer received from the server to the user, allowing the user to obtain reliable information and an appropriate answer that reflects their current emotions.

[0844] As a concrete example, if a user types the question "Is this product child-safe?" in an anxious tone, the following prompt sentence is generated:

[0845] Example prompt: "This product is child-safe, but age-appropriate use is recommended. Additionally, regular maintenance and child supervision are important additional safety precautions."

[0846] In this way, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual agent of a trusted reviewer, thereby supporting the user's decision-making.

[0847] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0848] Step 1:

[0849] The server collects data on trusted reviewers. Database connection information and reviewer IDs are required as input. Using a database system (e.g., MySQL or MongoDB), past review information and transaction history are queried based on the reviewer ID. This query process outputs the reviews and purchase records left by the reviewer.

[0850] Step 2:

[0851] The server trains a generative AI model based on the data collected. The reviewer data collected in step 1 is used as input. The data is input to the generative AI model (e.g., GPT-4) using a GPU, and the training process is carried out. Through this process, the generative AI model learns the reviewer's evaluation style and comment patterns, and outputs the trained model.

[0852] Step 3:

[0853] The server generates a virtual agent using the generative AI model. The generative AI model trained in step 2 is used as input. Based on the trained model, a virtual agent that imitates the reviewer's style is generated and output as an agent that can be used within the system.

[0854] Step 4:

[0855] The device provides a user interface (UI), which includes a text entry field and a submit button. The user enters and submits a question about a product or service. Inputs include the user's text and voice data. The output of this step is the user's question in text format.

[0856] Step 5:

[0857] The user enters a question and presses the submit button. This action causes the user's text and voice data to be stored on the device. The voice data is converted to text using voice recognition software (e.g., Google Speech-to-Text). The input requires the user's question, and the output is converted to text format.

[0858] Step 6:

[0859] The device sends the question and emotion data to the server. The input includes the converted question and speech data stored in step 5. By sending this data to the server, the server receives the question content and emotion data and executes the next step based on them.

[0860] Step 7:

[0861] The server generates an answer to the user's question. The input includes the question text received in step 6. Using the trained generative AI model, it generates an answer based on the question content. The generated answer is based on the virtual agent's mimicking style, and the output is the answer text.

[0862] Step 8:

[0863] The server analyzes the emotion data and reflects it in the answer. The emotion data received in step 6 is included as input. An emotion engine (e.g., IBM Watson Tone Analyzer) is used to analyze the data and identify the emotional state. Based on this emotional state, the answer generated in step 7 is adjusted to output a final answer that takes the user's emotions into consideration.

[0864] Step 9:

[0865] The server sends the final answer to the terminal. As input, the final answer text adjusted in the previous step is required. By sending this data to the terminal, the terminal receives and displays the final answer.

[0866] Step 10:

[0867] The device displays the final answer to the user. The input includes the final answer text received from the server. This is displayed in the user interface, allowing the user to obtain an appropriate answer based on their emotions along with reliable information. The output is the user's ability to view the final answer.

[0868] (Application example 2)

[0869] 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."

[0870] In conventional review systems, users have had difficulty finding reliable information from the vast number of reviews, and there are few ways to obtain appropriate information that matches their own emotions. This invention aims to solve these problems by generating an avatar based on the past reviews and purchase history of reliable reviewers, and providing answers that match the user's emotions.

[0871] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for constructing a generative model that learns the collected review data, means for generating an avatar that imitates the reviewer using the learned generative model, means for adjusting answers according to the user's emotions using an emotion engine that recognizes the user's emotions, and means for passing questions from the user to the avatar and returning the answers. This makes it possible to provide reliable information and obtain appropriate answers according to the user's emotions.

[0872] A "trusted reviewer" is one whose ratings and opinions are accurate and consistent based on past reviews and purchasing history.

[0873] A "review" is an evaluation or comment posted by a user about a specific product or service based on their own experience and opinion.

[0874] "Purchase history" is a record of products and services a user has purchased in the past.

[0875] A "generative model" is an artificial intelligence model that learns from collected data and generates an appropriate output for any input.

[0876] An "avatar" is a virtual entity that mimics the opinions and ratings of trusted reviewers and answers questions from users.

[0877] "User emotion" refers to the user's psychological state or feelings during a question or interaction.

[0878] An "emotion engine" is an algorithm or software that analyzes a user's text or voice data to identify their emotions.

[0879] A "question" is an action in which a user enters a question or request for confirmation regarding a specific product or service.

[0880] An "answer" is information or an opinion provided by an avatar in response to a question from a user.

[0881] The system for implementing this invention is mainly composed of three entities: a server, a terminal, and a user. The function of each entity is explained below.

[0882] Server Operation

[0883] The server operates using the following means:

[0884] 1. Data collection methods:

[0885] The past reviews and purchase history of trusted reviewers are collected from databases and online platforms, and the collected data is used to extract the necessary information using a specific reviewer ID.

[0886] 2. Generative model construction method:

[0887] A generative AI model is constructed based on the collected review data. This generative model uses past reviews and purchase history as training data to learn the reviewer's evaluation style and comment patterns. After training, the generative model is able to faithfully imitate the reviewer's opinions and evaluations.

[0888] 3. Avatar creation method:

[0889] Using a trained generative model, we generate an avatar that mimics the reviewer, and this avatar is used to answer questions from users.

[0890] 4. Emotion recognition means:

[0891] The emotion engine analyzes the user's input text and voice to identify their current emotion.

[0892] This allows the server to generate a virtual avatar of a trusted reviewer and provide answers tailored to the user's emotions.

[0893] Device behavior

[0894] The terminal operates as an interface with the user as follows.

[0895] 1. User interface provision means:

[0896] It provides an interface where users can enter questions. The UI includes a field to enter questions about a specific product or service, a submit button, and the ability to capture user voice and text sentiment.

[0897] 2. Means of data transmission:

[0898] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[0899] Examples of emotion-assisted answers

[0900] Specific examples are shown below.

[0901] 1. The user emotionally types the question "Is this product child safe?" and speaks it in a tone that indicates anxiety.

[0902] 2. The device captures the user's voice data along with this question and sends it to the server.

[0903] 3. The server sends a question to the avatar, and at the same time, it analyzes the emotional data using an emotion engine to recognize the user's anxiety.

[0904] 4. The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." Furthermore, based on the emotion engine's information that the user is feeling uneasy, the avatar responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[0905] 5. The device will display this answer to the user, providing them with reliable information and a sense of security.

[0906] Prompt Sentence Examples

[0907] Below are some example prompts to provide to the generative AI model:

[0908] User Question: "Is this product safe for children?"

[0909] User sentiment: "Anxiety"

[0910] Avatar response to be generated: "This product is child-safe, however, regular maintenance and child supervision are important additional safety precautions."

[0911] This system allows users to obtain reliable information and receive appropriate answers based on their own emotions.

[0912] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0913] Step 1: Data collection

[0914] The server collects past reviews and purchase history from trusted reviewers. The input is review data and purchase history data obtained using reviewer IDs. The server collects this data from databases and online platforms and stores it in a database. The output is the collected review data and purchase history data.

[0915] Step 2: Building a generative model

[0916] The server builds a generative AI model based on the collected data. Past reviews and purchase history are used as input. The server uses this data as training data and uses a machine learning algorithm to learn the reviewer's evaluation style and comment patterns. The output is a trained generative AI model.

[0917] Step 3: Create an avatar

[0918] The server generates an avatar using a trained generative model. The inputs are the trained generative model and review data. The server uses these data to generate a virtual avatar that closely mimics the reviewer. The output is the avatar.

[0919] Step 4: Submit your question and sentiment data

[0920] A user inputs a question about a product or service into the device's UI. The device captures the user's voice and text emotion data along with the input question and sends it to the server. The input is the user's question and emotion data, and the output is data sent to the server.

[0921] Step 5: Emotion Recognition

[0922] The server analyzes the emotion data sent by the user. The input is the user's emotion data, which is analyzed using the server's emotion engine. The emotion engine analyzes the text and / or voice to determine the user's current emotional state. The output is the analyzed emotional state.

[0923] Step 6: Answer Generation

[0924] The server passes the user's question to the avatar, which generates an answer. The input is the user's question and the analyzed emotional state. The avatar generates an answer based on this. In doing so, it adds adaptive information according to the emotion based on data from the emotion engine. For example, an answer containing additional information that gives a sense of security to an anxious user is generated. The output is the generated answer.

[0925] Step 7: Submit and view your responses

[0926] The server sends the generated answer to the user's device. The device receives the answer and displays it to the user. The input is the generated answer, and the output is the answer displayed to the user. This allows the user to obtain reliable information as well as an appropriate answer based on their emotions.

[0927] 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.

[0928] 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.

[0929] 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.

[0930] [Fourth embodiment]

[0931] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0932] 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.

[0933] 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).

[0934] 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.

[0935] 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.

[0936] 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).

[0937] 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. 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.

[0938] 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.

[0939] 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.

[0940] 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.

[0941] 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.

[0942] 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.

[0943] 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."

[0944] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0945] Server Operation

[0946] First, the server collects the past reviews and purchase history of trusted reviewers. This is achieved by pulling data from databases or online platforms. It retrieves the reviewer's data using a specific reviewer ID.

[0947] Next, the server builds a generative model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative model created here can then mimic the reviewer's opinions and evaluations.

[0948] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[0949] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[0950] Device behavior

[0951] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field for users to enter questions about a particular product or service, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[0952] User behavior

[0953] Users enter questions about the product or service they are considering purchasing. For example, they can enter specific questions such as, "Is this product child-safe?" After entering their question, the user presses a submit button to send the question to the server.

[0954] Specific examples

[0955] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[0956] The avatar generates a response based on past reviewer data: "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user. As a result, the user can make purchasing decisions based on reliable information.

[0957] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[0958] The processing flow will be explained below.

[0959] Step 1:

[0960] The server collects past reviews and purchase history of trusted reviewers from databases and online platforms, and retrieves the required data using a specific reviewer ID.

[0961] Step 2:

[0962] The server builds a generative model based on the collected reviewer data, and the model is trained to learn reviewers' evaluation styles and comments and understand their patterns.

[0963] Step 3:

[0964] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to answer questions from users.

[0965] Step 4:

[0966] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[0967] Step 5:

[0968] When a user enters a question and presses the send button, the device sends the question to the server, which transmits the data from the device to the server via an Internet connection.

[0969] Step 6:

[0970] The server receives a question from the user and passes the question to the corresponding avatar, identifies the corresponding avatar in the server, and passes the question data to it.

[0971] Step 7:

[0972] The avatar generates answers to questions based on past reviews and purchasing history. The avatar refers to the data it has learned and creates the most appropriate answer to the user's question.

[0973] Step 8:

[0974] The server receives the generated response and sends it back to the device, which also transmits the data back to the device via the Internet connection.

[0975] Step 9:

[0976] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended."

[0977] Step 10:

[0978] Users can make purchasing decisions based on the displayed answers, gaining more reliable information, which can reduce regrets such as "I shouldn't have bought it."

[0979] As described above, the system of the present invention provides users with the information they need through the virtual avatars of trusted reviewers, and supports their purchasing behavior.

[0980] Example 1

[0981] 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."

[0982] Existing online review systems have problems in that it is difficult to easily refer to the opinions of reliable reviewers, and extracting necessary information from a huge number of reviews requires a lot of time and effort. This makes it difficult for users to quickly obtain reliable information regarding purchasing products and services.

[0983] 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.

[0984] In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for preprocessing the collected reviewer data, means for constructing a generative AI model that learns the preprocessed data, means for generating an avatar that imitates the reviewer using the trained generative AI model, and means for passing questions from users to the avatar and returning answers, thereby enabling users to quickly obtain reliable information.

[0985] A "trusted reviewer" is a reviewer who provides a reliable rating based on past reviews and purchasing history.

[0986] "Past reviews" refer to ratings and comments about products or services that a reviewer has previously written.

[0987] "Purchase history" refers to a record of products or services that a reviewer has previously purchased.

[0988] A "generative AI model" is a model generated from training data using artificial intelligence technology.

[0989] "Preprocessing" refers to the process of properly cleansing collected data and performing format conversion, tokenization, etc.

[0990] An "avatar" is a virtual persona created to mimic a reviewer's rating style and opinions.

[0991] "User interface" refers to the screen containing input forms and buttons that a user can operate on a device.

[0992] "Question handing" refers to the process of providing a user's inquiry to an avatar.

[0993] "Returning an answer" refers to the process of providing the user with a response generated from the avatar.

[0994] "Collection methods" refers to the methods and devices used to obtain the required data from databases and online platforms.

[0995] The "means of construction" refers to the methods and processes required to create a generative AI model.

[0996] "Means of generation" refers to the method or process for creating an avatar using a trained generative AI model.

[0997] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to obtain reliable information by asking questions to the avatar. This system operates appropriately mainly through three entities: a server, a terminal, and a user.

[0998] Server Operation

[0999] First, the server collects past reviews and purchase history of trusted reviewers. This can be achieved by retrieving data from a database or online platform. For example, the server can use an online store's API to retrieve reviews and purchase history based on a specific reviewer ID. Specifically, the server could use Amazon's API to retrieve the review history of a specific reviewer.

[1000] The server then preprocesses the collected data. This preprocessing includes data cleansing, formatting, and tokenization. For example, it removes HTML tags from text data and splits reviews into individual words. These operations are performed using Python NLP libraries (e.g., NLTK and spaCy).

[1001] The server builds a generative AI model based on the preprocessed data and trains it. A generative AI model such as GPT-3 is used to learn reviewer evaluation styles and comment patterns. Frameworks such as TensorFlow and PyTorch are used to train the model, and multiple epochs are used.

[1002] Using a trained generative AI model, the server generates an avatar that mimics the reviewer's ratings and opinions and generates appropriate answers based on user questions.

[1003] Finally, the server receives the user's question and passes it to the avatar. The question is input into the generative AI model, and the output answer is returned to the user via the server. For example, an API endpoint can be set up using a web framework such as Flask to exchange questions and answers.

[1004] Device behavior

[1005] The device provides a user interface (UI) that allows users to enter questions. This UI includes a text field and a submit button for users to enter questions about specific products or services. When a user enters a question and presses the submit button, the question is sent to the server. Specifically, the UI is created using HTML and JavaScript, and the question is sent to the server via asynchronous communication using a JavaScript Ajax request.

[1006] User behavior

[1007] Users enter questions about the product or service they are considering purchasing into the device's UI. For example, they can enter specific questions such as, "Is this product child-safe?" After entering the question, the user presses a send button to send the question to the server.

[1008] Specific examples

[1009] For example, if a user enters the question "Is this product child-safe?" into the device's UI and presses the submit button, the device sends this question to the server, which then forwards it to the reviewer's avatar.

[1010] The avatar uses past reviewer data to generate a response such as, "This product is safe for children, but age-appropriate use is recommended." The server receives this response and sends it back to the device. Finally, the device displays the received response to the user, allowing the user to make purchasing decisions based on reliable information.

[1011] An example prompt could be, "Is this product safe for children? Please answer based on past reviews and purchasing history from trusted reviewers." Based on this, the generative AI model generates an answer in natural language.

[1012] In this way, the system of the present invention provides reliable information via avatars that mimic the opinions of trusted reviewers, thereby supporting users' purchasing behavior.

[1013] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1014] Step 1:

[1015] The server collects past reviews and purchase history from a database or online platform (e.g., an online store's API) based on a trusted reviewer ID. The server receives a specific reviewer ID as input and sends an HTTP request to the database. This request retrieves the reviewer's review history and purchase history in JSON format. This retrieved data is temporarily stored in the server's storage.

[1016] Step 2:

[1017] The server preprocesses the collected review and purchase history data. Specifically, it performs data cleansing (removing unnecessary and duplicate data), format conversion, and tokenization. The server takes the ingested JSON data as input and performs data cleansing and tokenization using a Python NLP library (e.g., NLTK or spaCy). The preprocessed data is then output as a dataset for model training.

[1018] Step 3:

[1019] The server builds and trains a generative AI model based on the preprocessed reviewer data. A generative AI model such as GPT-3 is used for the training process. The server uses the preprocessed reviewer data as input and trains the generative AI model using a machine learning framework such as TensorFlow or PyTorch. This training process outputs a generative model that mimics the reviewer's evaluation style and comment patterns.

[1020] Step 4:

[1021] The server uses the trained model to generate an avatar. This enables the server to generate answers to user questions based on their evaluation style and opinions. The server inputs the trained generative AI model and outputs profile data for the avatar. This profile data includes the results of feature extraction for the reviewer.

[1022] Step 5:

[1023] The device provides a UI that allows users to enter questions. Users enter specific questions about the product or service they are considering purchasing. When a question such as "Is this product child-safe?" is entered, the device treats it as input data and sends it to the server using a JavaScript Ajax request. This request transfers the question data to the server.

[1024] Step 6:

[1025] The server passes the question received from the user to the avatar and generates an answer. The server uses the question data received from the device as input and generates an answer based on the trained generative AI model. The generated answer is output as an HTTP response from the server to the device.

[1026] Step 7:

[1027] The terminal displays the answer received from the server to the user. The terminal receives the HTTP response from the server as input, analyzes its contents, and displays it on the UI. The answer is presented in a format that is easy for the user to understand, and the user can make a purchasing decision based on that information.

[1028] (Application example 1)

[1029] 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."

[1030] On traditional online shopping sites, it is difficult to obtain reliable information about products, especially for new or lesser-known products. Furthermore, highly anonymous online reviews often lack credibility, making it difficult for users to provide sufficient information for making confident purchasing decisions. This creates the risk of users making incorrect purchasing decisions based on inaccurate information. To address these issues, a system is needed that generates avatars based on the ratings and purchasing history of trusted reviewers, allowing users to easily obtain reliable information.

[1031] 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.

[1032] In this invention, the server includes a means for collecting past reviews and purchase histories of reliable reviewers, a means for constructing a generative AI model that learns the collected review data, and a means for generating an avatar that imitates the reviewer using the trained generative AI model, thereby enabling users to obtain reliable information about products on online shopping sites.

[1033] A "trusted reviewer" is someone who has a high rating and credibility based on past reviews and purchasing history.

[1034] "Past ratings and purchasing history" refers to the reviewer's previous ratings of products and services and the history of products they have purchased.

[1035] "Means of collection" refers to the methods used to obtain reliable reviewer ratings and purchasing data using databases or online platforms.

[1036] A "generative AI model" is an artificial intelligence model that learns from collected reviewer data and imitates reviewer evaluation styles and comment patterns.

[1037] An "avatar" is a virtual persona that uses generative AI models to mimic the opinions and ratings of trusted reviewers.

[1038] "User interface" refers to the screen layout and operation method that allows users to easily enter questions and receive answers.

[1039] The "server" is a central computer that collects reviewer data, trains the generative AI model, generates avatars, and handles questions and answers from users.

[1040] A "terminal" is a device used by a user, an electronic device that operates in the form of a smartphone app, etc.

[1041] The system for implementing the present invention operates mainly based on three entities: a server, a terminal, and a user.

[1042] Server Operation

[1043] The server collects the past ratings and purchase history of trusted reviewers. This is achieved by retrieving reviewer data from databases or online platforms. It uses a specific reviewer ID to collect the data of the corresponding reviewer.

[1044] The server then builds a generative AI model based on the collected data and trains it. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The generative AI model created here can then mimic the reviewer's opinions and evaluations.

[1045] The server then uses the trained generative AI model to generate an avatar that mimics the reviewer's opinions and ratings and provides appropriate answers to user questions.

[1046] Finally, the server receives the user's question and passes it to the avatar, which generates an answer based on the question and returns the result to the server, which then provides the generated answer to the user.

[1047] Device behavior

[1048] The terminal provides a user interface (UI) that allows users to enter questions. This UI has a field where users can enter questions about specific products or services, and a submit button. When a user enters a question and presses the submit button, the question is sent to the server.

[1049] User behavior

[1050] The user enters a question about a product or service they are considering purchasing. For example, they can enter a specific question such as, "What age group is this toy suitable for?" After entering the question, the user presses a submit button to send the question to the server.

[1051] Specific examples

[1052] For example, consider the case where a user enters the question "What age group is this toy suitable for?" into the device's UI and presses the send button. The device sends this question to the server, which then forwards it to the corresponding reviewer's avatar. Based on past reviewer data, the avatar generates an answer such as "This toy is particularly popular with children aged 3 to 6. It is also very safe." The server receives this answer and sends it back to the device. Finally, the device can display the received answer to the user. This allows the user to make purchasing decisions based on reliable information.

[1053] Prompt Sentence Examples

[1054] Below are some examples of specific prompt sentences.

[1055] text

[1056] Reviewer data: This toy is safe for children aged 3 to 6. It's very popular among parents.

[1057] Question: What age group is this toy suitable for?

[1058] Answer:

[1059] The server uses this to ask questions to the avatar and provides the generated answers to the user. In this way, the system of the present invention provides reliable information via the avatar that mimics the opinions of reliable reviewers, thereby supporting users' purchasing behavior.

[1060] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1061] Step 1: Collect data from trusted reviewers

[1062] The server collects the past ratings and purchase history of trusted reviewers from databases and online platforms. In this process, it uses the reviewer ID as a key to obtain detailed rating data and purchase history data for each reviewer. The input is the reviewer ID and database connection information, and the output is the reviewer's detailed rating data and purchase history.

[1063] Step 2: Training the generative AI model

[1064] The server builds and trains a generative AI model based on the collected reviewer data. In this process, the reviewer's evaluation style and historical purchasing patterns are used as training data to train the AI ​​model (e.g., GPT-2). The input is the reviewer data, and the output is the trained generative AI model.

[1065] Step 3: Creating an avatar

[1066] The server uses a trained generative AI model to generate an avatar that mimics the reviewer. This avatar has the ability to answer questions based on the collected information. The input is the trained generative AI model, and the output is the avatar.

[1067] Step 4: Providing a User Interface

[1068] The terminal provides a user interface (UI) where the user can enter a question. This UI includes a field for the user to enter a question and a submit button. The input is the user's action feedback, and the output is the question entered in the question field.

[1069] Step 5: Sending the question to the server

[1070] The terminal sends the question entered by the user to the server. In this process, the question and the user ID are sent to the server through an HTTP request. The input is the user's question, and the output is the HTTP request sent to the server.

[1071] Step 6: Transfer the question to your avatar

[1072] The server passes the received user question to the corresponding avatar. When the question is passed to the avatar, an appropriate prompt is generated using a generative AI model. The input is the user's question, and the output is the prompt passed to the avatar.

[1073] Step 7: Generate an answer

[1074] The avatar generates answers based on the questions it receives from users. Using a generative AI model, the answers are generated in a way that mimics the rating style of a trusted reviewer. The input is the prompt, and the output is the generated answer.

[1075] Step 8: Sending the answer back to the server

[1076] The avatar then sends the generated answer back to the server, where it is ready to be sent back to the user. The input is the generated answer, and the output is the answer sent to the server.

[1077] Step 9: Providing answers to users

[1078] The server sends the generated answer back to the user's device, where it can be viewed by the user. The device displays the answer received from the server in a user interface. The input is the answer from the server, and the output is the answer displayed on the device's display screen.

[1079] At each step, users can get answers from their avatars based on the opinions of trusted reviewers, allowing them to make purchasing decisions with confidence.

[1080] 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.

[1081] This invention is a system that generates an avatar based on the past reviews and purchase history of trusted reviewers, and allows users to ask questions to the avatar to obtain reliable information. It also combines an emotion engine that recognizes the user's emotions and provides answers that correspond to the user's emotions. This system operates appropriately mainly through three entities: a server, a terminal, and a user, as well as the emotion engine.

[1082] Server Operation

[1083] First, the server collects the past reviews and purchase history of trusted reviewers from databases or online platforms, and then retrieves the required data using a specific reviewer ID.

[1084] The server then builds and trains a generative model based on the collected data. During this training process, past reviews and purchase history are used as training data to learn the reviewer's evaluation style and comment patterns. The trained generative model can faithfully mimic the reviewer's opinions and evaluations.

[1085] The server then uses the trained generative model to generate an avatar that mimics the reviewer's opinions and ratings and answers questions from users.

[1086] The server then uses an emotion engine to recognize the user's emotions, which analyzes the user's input text and voice to identify their current emotions.

[1087] Device behavior

[1088] The device provides a user interface (UI) that allows the user to enter a question. The UI includes a field for the user to enter a question about a particular product or service and a submit button. When the user enters a question, the device can also capture the user's voice and text sentiment and transmit it to the server.

[1089] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[1090] Collaboration between the server and emotion engine

[1091] The server receives the user's question and emotion data, then passes the question to the avatar to generate an answer. At the same time, the emotion engine analyzes the emotion data and identifies the user's emotional state. For example, if the user is feeling anxious, that emotion is communicated to the server.

[1092] The emotional state of the user analyzed by the emotion engine influences the avatar's responses. For example, if the user is feeling anxious, the avatar will adjust its responses to take that into consideration. For example, the answer to the question, "Is this product safe for children?" will be "Yes, it is safe. However, we also have some recommendations for further safety checks." This will add information to alleviate anxiety.

[1093] Specific examples

[1094] For example, suppose a user emotionally inputs the question "Is this product safe for children?" and speaks it in a tone that indicates anxiety. The device captures the user's voice data along with the question and sends it to the server. The server then sends the question to the avatar, while simultaneously analyzing the emotional data with its emotion engine and recognizing the user's anxiety.

[1095] The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." It then receives information from the emotion engine that the user is feeling uneasy and responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[1096] The device displays this response to the user, allowing the user to receive reliable information and a sense of security at the same time. In this way, the system of the present invention provides reliable information that reflects the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[1097] The processing flow will be explained below.

[1098] Step 1:

[1099] The server collects the past reviews and purchase history of trusted reviewers from databases and online platforms, and then retrieves the required data using a specific reviewer ID.

[1100] Step 2:

[1101] The server builds and trains a generative model based on the collected reviewer data. The model learns and understands reviewer evaluation styles and comment patterns.

[1102] Step 3:

[1103] The server uses the trained generative model to generate an avatar, which mimics the reviewer's opinions and ratings and is used to provide answers to user questions.

[1104] Step 4:

[1105] The device provides a user interface (UI) that allows users to enter specific questions about the product or service they are considering purchasing.

[1106] Step 5:

[1107] When users input questions, the device also captures text and voice data to recognize users' emotions, and has a built-in emotion engine to analyze users' emotions.

[1108] Step 6:

[1109] When a user enters a question and presses the send button, the device sends the question and captured emotion data to the server via an internet connection.

[1110] Step 7:

[1111] The server receives the question and emotion data from the user, and first passes the question to the avatar to generate an answer.

[1112] Step 8:

[1113] The server then uses an emotion engine to analyze the emotion data and identify the user's emotional state, for example, the emotion engine may recognize that the user is feeling anxious.

[1114] Step 9:

[1115] The emotional state of the user analyzed by the emotion engine influences the avatar's response. Specifically, the avatar takes into account the emotional data and generates responses that correspond to the user's emotions.

[1116] Step 10:

[1117] For example, if a user types the question "Is this product child-safe?" with some anxiety, the avatar will generate the answer "This product is child-safe, but we recommend age-appropriate use. Regular maintenance and child supervision are also important as additional safety precautions."

[1118] Step 11:

[1119] The server receives the generated response and sends it back to the device, which also sends the data back to the device via the Internet connection.

[1120] Step 12:

[1121] The device will then display the answer received from the server to the user, for example, "This product is child-safe, but age-appropriate use is recommended. Regular maintenance and child supervision are also important as additional safety precautions."

[1122] Step 13:

[1123] Users can make purchasing decisions based on the displayed answers, obtaining more reliable information, which reduces the chance of users regretting their purchase.

[1124] As described above, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual avatar of a trusted reviewer, thereby supporting the user's purchasing behavior.

[1125] Example 2

[1126] 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."

[1127] Today's consumers seek reliable information about products and services online, but online reviews and ratings often lack credibility. Furthermore, when users consult online reviews, they face challenges in finding advice that is tailored to their emotions and circumstances. Furthermore, information overload and conflicting reviews make it difficult for users to make appropriate decisions.

[1128] 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 acquiring past evaluation information and transaction history of reliable reviewers, means for constructing a generative model for learning the acquired reviewer data, means for generating a virtual agent that mimics the reviewer using the learned generative model, means for passing questions from a user to the virtual agent and providing the answers, means for identifying an emotional state from the user's input, and means for adjusting the answers of the virtual agent based on the identified emotional state. This allows the user to receive highly reliable information as well as advice that corresponds to their current emotions.

[1129] A "trusted reviewer" is a reviewer who has provided accurate and fair ratings in the past and has received high ratings from many users.

[1130] "Rating Information" refers to the content of the review or feedback provided by the reviewer regarding a product or service.

[1131] "Transaction History" refers to records of products or services that a reviewer has previously purchased or used.

[1132] "Capture methods" refers to software and processes used to collect data from databases and external platforms.

[1133] "Reviewer Data" refers collectively to rating information and transaction history associated with trusted reviewers.

[1134] A "generative model" refers to an artificial intelligence algorithm or mechanism that learns from specific data and generates new data.

[1135] A "virtual agent" is a virtual entity that uses a trained generative model to mimic a specific reviewer and provide answers to their questions.

[1136] "User input" refers to data input, such as text or voice, made by a user to a system.

[1137] "Means for identifying emotional state" refers to the algorithms or engines used to analyze emotions from user input and identify that state.

[1138] "Response tailoring" refers to a process or algorithm for appropriately modifying the responses provided based on the identified emotional state.

[1139] This invention is a system that generates a virtual agent based on the past evaluation information and transaction history of trusted reviewers, and allows users to ask the virtual agent questions to obtain reliable information. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it provides answers that correspond to the user's emotions. This system consists of three entities: a server, a terminal, and a user.

[1140] The server collects the past rating information and transaction history of trusted reviewers from a database or online platform, for example, using SQL Server or a NoSQL database (e.g., MySQL, MongoDB), and retrieves the required data using the reviewer ID.

[1141] Next, the server builds a generative AI model (e.g., GPT-4) based on the collected data and trains it. During this training process, a GPU is used to quickly process large amounts of data, and reviewers' evaluation styles and comment patterns are used as training data. This creates a generative AI model that faithfully mimics the reviewers' opinions and evaluations.

[1142] The server then uses the trained generative AI model to generate a virtual agent that mimics the reviewer, which can then use the generative AI model to generate appropriate answers to user questions.

[1143] The device provides a user interface (UI) that allows users to input questions. Specifically, the UI includes a text input field and a submit button. This allows users to easily input and submit questions about specific products or services. The device can also capture and send emotional data from the user's voice and text to a server. Voice recognition software (e.g., Google Speech-to-Text) is used for voice recognition.

[1144] When a user enters a question and presses the send button, the device sends the question and the captured emotion data to the server. The server then passes the received question to a virtual agent, which generates an answer. In parallel, the server analyzes the emotion data using an emotion engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. For example, if the user is feeling anxious, the server will recognize that emotion.

[1145] The server adjusts the generated answer based on the user's emotional state analyzed by the emotion engine. For example, if the user is feeling anxious, the server generates an answer that takes their emotion into consideration. In response to the question, "Is this product child-safe?", the virtual agent generates the answer, "This product is child-safe, but we recommend age-appropriate use," and then adds reassuring information such as, "Additional safety measures include regular maintenance and child supervision."

[1146] The device displays the final answer received from the server to the user, allowing the user to obtain reliable information and an appropriate answer that reflects their current emotions.

[1147] As a concrete example, if a user types the question "Is this product child-safe?" in an anxious tone, the following prompt sentence is generated:

[1148] Example prompt: "This product is child-safe, but age-appropriate use is recommended. Additionally, regular maintenance and child supervision are important additional safety precautions."

[1149] In this way, the system of the present invention provides reliable information that corresponds to the user's emotions through the virtual agent of a trusted reviewer, thereby supporting the user's decision-making.

[1150] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1151] Step 1:

[1152] The server collects data on trusted reviewers. Database connection information and reviewer IDs are required as input. Using a database system (e.g., MySQL or MongoDB), past review information and transaction history are queried based on the reviewer ID. This query process outputs the reviews and purchase records left by the reviewer.

[1153] Step 2:

[1154] The server trains a generative AI model based on the data collected. The reviewer data collected in step 1 is used as input. The data is input to the generative AI model (e.g., GPT-4) using a GPU, and the training process is carried out. Through this process, the generative AI model learns the reviewer's evaluation style and comment patterns, and outputs the trained model.

[1155] Step 3:

[1156] The server generates a virtual agent using the generative AI model. The generative AI model trained in step 2 is used as input. Based on the trained model, a virtual agent that imitates the reviewer's style is generated and output as an agent that can be used within the system.

[1157] Step 4:

[1158] The device provides a user interface (UI), which includes a text entry field and a submit button. The user enters and submits a question about a product or service. Inputs include the user's text and voice data. The output of this step is the user's question in text format.

[1159] Step 5:

[1160] The user enters a question and presses the submit button. This action causes the user's text and voice data to be stored on the device. The voice data is converted to text using voice recognition software (e.g., Google Speech-to-Text). The input requires the user's question, and the output is converted to text format.

[1161] Step 6:

[1162] The device sends the question and emotion data to the server. The input includes the converted question and speech data stored in step 5. By sending this data to the server, the server receives the question content and emotion data and executes the next step based on them.

[1163] Step 7:

[1164] The server generates an answer to the user's question. The input includes the question text received in step 6. Using the trained generative AI model, it generates an answer based on the question content. The generated answer is based on the virtual agent's mimicking style, and the output is the answer text.

[1165] Step 8:

[1166] The server analyzes the emotion data and reflects it in the answer. The emotion data received in step 6 is included as input. An emotion engine (e.g., IBM Watson Tone Analyzer) is used to analyze the data and identify the emotional state. Based on this emotional state, the answer generated in step 7 is adjusted to output a final answer that takes the user's emotions into consideration.

[1167] Step 9:

[1168] The server sends the final answer to the terminal. As input, the final answer text adjusted in the previous step is required. By sending this data to the terminal, the terminal receives and displays the final answer.

[1169] Step 10:

[1170] The device displays the final answer to the user. The input includes the final answer text received from the server. This is displayed in the user interface, allowing the user to obtain an appropriate answer based on their emotions along with reliable information. The output is the user's ability to view the final answer.

[1171] (Application example 2)

[1172] 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."

[1173] In conventional review systems, users have had difficulty finding reliable information from the vast number of reviews, and there are few ways to obtain appropriate information that matches their own emotions. This invention aims to solve these problems by generating an avatar based on the past reviews and purchase history of reliable reviewers, and providing answers that match the user's emotions.

[1174] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting past reviews and purchase histories of reliable reviewers, means for constructing a generative model that learns the collected review data, means for generating an avatar that imitates the reviewer using the learned generative model, means for adjusting answers according to the user's emotions using an emotion engine that recognizes the user's emotions, and means for passing questions from the user to the avatar and returning the answers. This makes it possible to provide reliable information and obtain appropriate answers according to the user's emotions.

[1175] A "trusted reviewer" is one whose ratings and opinions are accurate and consistent based on past reviews and purchasing history.

[1176] A "review" is an evaluation or comment posted by a user about a specific product or service based on their own experience and opinion.

[1177] "Purchase history" is a record of products and services a user has purchased in the past.

[1178] A "generative model" is an artificial intelligence model that learns from collected data and generates an appropriate output for any input.

[1179] An "avatar" is a virtual entity that mimics the opinions and ratings of trusted reviewers and answers questions from users.

[1180] "User emotion" refers to the user's psychological state or feelings during a question or interaction.

[1181] An "emotion engine" is an algorithm or software that analyzes a user's text or voice data to identify their emotions.

[1182] A "question" is an action in which a user enters a question or request for confirmation regarding a specific product or service.

[1183] An "answer" is information or an opinion provided by an avatar in response to a question from a user.

[1184] The system for implementing this invention is mainly composed of three entities: a server, a terminal, and a user. The function of each entity is explained below.

[1185] Server Operation

[1186] The server operates using the following means:

[1187] 1. Data collection methods:

[1188] The past reviews and purchase history of trusted reviewers are collected from databases and online platforms, and the collected data is used to extract the necessary information using a specific reviewer ID.

[1189] 2. Generative model construction method:

[1190] A generative AI model is constructed based on the collected review data. This generative model uses past reviews and purchase history as training data to learn the reviewer's evaluation style and comment patterns. After training, the generative model is able to faithfully imitate the reviewer's opinions and evaluations.

[1191] 3. Avatar creation method:

[1192] Using a trained generative model, we generate an avatar that mimics the reviewer, and this avatar is used to answer questions from users.

[1193] 4. Emotion recognition means:

[1194] The emotion engine analyzes the user's input text and voice to identify their current emotion.

[1195] This allows the server to generate a virtual avatar of a trusted reviewer and provide answers tailored to the user's emotions.

[1196] Device behavior

[1197] The terminal operates as an interface with the user as follows.

[1198] 1. User interface provision means:

[1199] It provides an interface where users can enter questions. The UI includes a field to enter questions about a specific product or service, a submit button, and the ability to capture user voice and text sentiment.

[1200] 2. Means of data transmission:

[1201] When the user enters a question and presses the send button, the device sends the question and the captured emotion data to the server.

[1202] Examples of emotion-assisted answers

[1203] Specific examples are shown below.

[1204] 1. The user emotionally types the question "Is this product child safe?" and speaks it in a tone that indicates anxiety.

[1205] 2. The device captures the user's voice data along with this question and sends it to the server.

[1206] 3. The server sends a question to the avatar, and at the same time, it analyzes the emotional data using an emotion engine to recognize the user's anxiety.

[1207] 4. The avatar generates a basic answer such as "This product is child-safe, but age-appropriate use is recommended." Furthermore, based on the emotion engine's information that the user is feeling uneasy, the avatar responds with a reassuring answer such as "Additional safety measures include regular maintenance and child supervision."

[1208] 5. The device will display this answer to the user, providing them with reliable information and a sense of security.

[1209] Prompt Sentence Examples

[1210] Below are some example prompts to provide to the generative AI model:

[1211] User Question: "Is this product safe for children?"

[1212] User sentiment: "Anxiety"

[1213] Avatar response to be generated: "This product is child-safe, however, regular maintenance and child supervision are important additional safety precautions."

[1214] This system allows users to obtain reliable information and receive appropriate answers based on their own emotions.

[1215] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1216] Step 1: Data collection

[1217] The server collects past reviews and purchase history from trusted reviewers. The input is review data and purchase history data obtained using reviewer IDs. The server collects this data from databases and online platforms and stores it in a database. The output is the collected review data and purchase history data.

[1218] Step 2: Building a generative model

[1219] The server builds a generative AI model based on the collected data. Past reviews and purchase history are used as input. The server uses this data as training data and uses a machine learning algorithm to learn the reviewer's evaluation style and comment patterns. The output is a trained generative AI model.

[1220] Step 3: Create an avatar

[1221] The server generates an avatar using a trained generative model. The inputs are the trained generative model and review data. The server uses these data to generate a virtual avatar that closely mimics the reviewer. The output is the avatar.

[1222] Step 4: Submit your question and sentiment data

[1223] A user inputs a question about a product or service into the device's UI. The device captures the user's voice and text emotion data along with the input question and sends it to the server. The input is the user's question and emotion data, and the output is data sent to the server.

[1224] Step 5: Emotion Recognition

[1225] The server analyzes the emotion data sent by the user. The input is the user's emotion data, which is analyzed using the server's emotion engine. The emotion engine analyzes the text and / or voice to determine the user's current emotional state. The output is the analyzed emotional state.

[1226] Step 6: Answer Generation

[1227] The server passes the user's question to the avatar, which generates an answer. The input is the user's question and the analyzed emotional state. The avatar generates an answer based on this. In doing so, it adds adaptive information according to the emotion based on data from the emotion engine. For example, an answer containing additional information that gives a sense of security to an anxious user is generated. The output is the generated answer.

[1228] Step 7: Submit and view your responses

[1229] The server sends the generated answer to the user's device. The device receives the answer and displays it to the user. The input is the generated answer, and the output is the answer displayed to the user. This allows the user to obtain reliable information as well as an appropriate answer based on their emotions.

[1230] 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.

[1231] 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.

[1232] 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.

[1233] 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.

[1234] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1235] 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.

[1236] 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).

[1237] 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.

[1238] 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."

[1239] 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.

[1240] 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).

[1241] 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.

[1242] 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.

[1243] 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.

[1244] 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.

[1245] 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.

[1246] 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.

[1247] 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.

[1248] 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.

[1249] 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.

[1250] 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.

[1251] The following is further disclosed regarding the above embodiment.

[1252] (Claim 1)

[1253] A means to collect past reviews and purchase history from trusted reviewers,

[1254] means for constructing a generative model that learns from the collected reviewer data;

[1255] A means for generating an avatar that imitates a reviewer using the trained generative model;

[1256] a means for passing a question from a user to said alter-ego avatar and returning an answer;

[1257] A system including:

[1258] (Claim 2)

[1259] The system of claim 1 , wherein the alter-ego avatar further comprises means for generating an answer based on a question from a user.

[1260] (Claim 3)

[1261] The system of claim 1 , wherein the generative model is a generative AI model.

[1262] "Example 1"

[1263] (Claim 1)

[1264] A means to collect past reviews and purchase history from trusted reviewers,

[1265] means for preprocessing the collected reviewer data;

[1266] means for constructing a generative AI model that learns from the preprocessed data;

[1267] A means for generating an avatar that imitates a reviewer using the trained generative AI model;

[1268] a means for passing a question from a user to said alter-ego avatar and returning an answer;

[1269] A system including:

[1270] (Claim 2)

[1271] The system of claim 1 , wherein the alter-ego avatar further comprises means for generating an answer based on a question from a user.

[1272] (Claim 3)

[1273] 10. The system of claim 1, further comprising means for providing a user interface through which the terminal transmits a question entered by a user to the server.

[1274] "Application Example 1"

[1275] (Claim 1)

[1276] A means of collecting past ratings and purchasing history from trusted reviewers,

[1277] means for constructing a generative AI model that learns from the collected evaluation data;

[1278] A means for generating an avatar that imitates the evaluator using the trained generative AI model;

[1279] means for providing a user interface through which a user can input a question;

[1280] a means for passing a question from a user to said alter-ego avatar and returning an answer thereto;

[1281] A means for operating by three entities, a server and a terminal;

[1282] A system including:

[1283] (Claim 2)

[1284] The system of claim 1 , wherein the alter-ego avatar further comprises means for generating an answer based on a question from a user.

[1285] (Claim 3)

[1286] 2. The system according to claim 1, wherein the alter-ego avatar operates as part of a smartphone app.

[1287] "Example 2: Combining Emotion Engines"

[1288] (Claim 1)

[1289] A means to obtain past evaluation information and transaction history of trusted reviewers, and

[1290] means for constructing a generative model for learning the acquired reviewer data;

[1291] A means for generating a virtual agent that mimics a reviewer using the trained generative model;

[1292] means for passing user questions to said virtual agent and providing answers thereto;

[1293] a means for determining an emotional state from a user's input;

[1294] means for adjusting the responses of the virtual agent based on the identified emotional state;

[1295] A system including:

[1296] (Claim 2)

[1297] The system of claim 1 , wherein the virtual agent further comprises means for generating an answer based on a question posed by a user.

[1298] (Claim 3)

[1299] The system of claim 1 , wherein the generative model is a generative AI model.

[1300] "Application example 2 when combining emotion engines"

[1301] (Claim 1)

[1302] A means to collect past reviews and purchase history from trusted reviewers,

[1303] means for constructing a generative model that learns from the collected review data;

[1304] A means for generating an avatar that imitates a reviewer using the trained generative model;

[1305] a means for passing a question from a user to said alter-ego avatar and returning an answer;

[1306] a means for adjusting responses according to emotions using an emotion engine that recognizes the user's emotions;

[1307] A system including:

[1308] (Claim 2)

[1309] The system of claim 1 , further comprising means for the avatar to generate an answer based on a question from the user and further comprising means for adjusting the answer based on an emotion engine.

[1310] (Claim 3)

[1311] The system of claim 1 , wherein the generative model is a generative AI model. [Explanation of symbols]

[1312] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means to collect past reviews and purchase history from trusted reviewers, means for constructing a generative model that learns from the collected reviewer data; A means for generating an avatar that imitates a reviewer using the trained generative model; a means for passing a question from a user to said alter-ego avatar and returning an answer; A system including:

2. The system of claim 1 , wherein the alter-ego avatar further comprises means for generating an answer based on a question from a user.

3. The system of claim 1 , wherein the generative model is a generative AI model.

Images