System

Abstract

A system is provided.SOLUTION: A system comprising: means for collecting a search history, a browsing history, and a word-of-mouth history of a user and storing the histories in a database; means for extracting a feature amount from collected user information and estimating a preference point of the user; means for generating a personalized introductory comment using a template based on the estimated preference point; and means for displaying the personalized introductory comment when the user accesses a product page.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional product descriptions tend to be general in content and are unable to effectively capture each user's specific interests. This makes it difficult to increase their desire to purchase a product. Furthermore, the inability to properly identify the points of interest of users limits the improvement of the user experience. Given these circumstances, there was a need to provide personalized product descriptions that address the different preferences of each user. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by providing a system that includes a means for collecting a user's search history, browsing history, and review history and storing them in a database, a means for extracting features from the collected user information and inferring the user's preferences, a means for generating a personalized introduction using a template based on the inferred preferences, and a means for displaying the personalized introduction when the user accesses a product page.

[0006] This system collects search history by recording the search query and date and time when the user conducts a search, and predicts preferences by analyzing features using a machine learning model, making it possible to generate more accurate personalized introductions.

[0007] "Search History" refers to the search queries you perform on the Site and the date and time you perform them.

[0008] "Browse history" refers to a record of a user's browsing of specific product pages or service pages, and includes the date and time of browsing and information about the pages that were viewed.

[0009] "Review history" refers to the content of ratings and reviews posted by users about products and services, as well as the date and time of posting.

[0010] "Database" refers to an information management system for storing users' search history, browsing history, review history, etc.

[0011] "Features" refer to indicators and patterns that indicate a user's interests and purchasing tendencies, extracted from user behavior data.

[0012] "Preference points" refer to features or elements of a product or service that a particular user is particularly interested in.

[0013] "Template" refers to a boilerplate that provides a basic structure or pattern for generating personalized testimonials.

[0014] A "machine learning model" refers to an artificial intelligence algorithm that learns patterns through data analysis and makes decisions such as predictions and classifications.

[0015] "Personalized testimonials" refers to product testimonials that are individually customized based on each user's preferences.

[0016] "Product Page" means a web page that displays detailed information about a particular product or service. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

[0019] First, the terms used in the following description will be explained.

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

[0026] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] The present invention provides a system for generating personalized testimonials based on a user's search history, browsing history, and review history, and the specific embodiments thereof will be described below.

[0039] 1. Collection of User Information

[0040] A special script is run to collect the user's search, browsing, and review history within the site. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0041] For example, if a user searches for "smartphone," the search query "smartphone" and the date and time of the search are recorded. If the user then views a product page, the product ID and the date and time of the view are also recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are saved in the database.

[0042] 2. Guessing your favorite points

[0043] Features are extracted from the collected user information to infer the user's preferences. This process is performed using a machine learning model and is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0044] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0045] 3. Generate personalized testimonials

[0046] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0047] For example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" will be generated.

[0048] 4. Displaying personalized testimonials

[0049] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the device retrieves the personalized introduction from the server and displays it on the product page.

[0050] For example, when a user accesses a product page while logged in, a description will appear at the top of the page stating, "This product provides a detailed explanation of the latest AI technology that interests you."

[0051] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized product introductions, thereby realizing product introductions that are attractive to users.

[0052] The processing flow will be explained below.

[0053] Step 1: Collect user information

[0054] Collection of search history

[0055] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0056] Collection of browsing history

[0057] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0058] Collecting review history

[0059] When a user posts a review, the device records the content and date and time of the post. This information is sent to the server and stored in a database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are stored.

[0060] Step 2: Preprocessing user information

[0061] Data Formatting

[0062] The server formats the collected search history, browsing history, and review history into a single data format. This preprocessing ensures that the data is stored in a consistent format.

[0063] Text Preprocessing

[0064] The server tokenizes and normalizes the text data, especially the reviews, removing unnecessary whitespace and special characters, which makes it easier for machine learning models to analyze later.

[0065] Step 3: Feature extraction

[0066] Identifying frequent keywords

[0067] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses words like "latest" and "technology," these frequently occurring keywords are extracted as features.

[0068] Identifying Categories

[0069] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0070] Step 4: Guess your preference points

[0071] Applying machine learning models

[0072] The server inputs the extracted features into a machine learning model to predict the user's preferences. Because the model learns from past data, it can predict the user's preferences with high accuracy.

[0073] Recording points

[0074] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details."

[0075] Step 5: Generate a personalized testimonial

[0076] Using templates

[0077] The server then inserts the inferred preferences into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0078] Customizable wording

[0079] The server adds specific words and phrases based on the user's preferences, such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field."

[0080] Step 6: Displaying a personalized testimonial

[0081] User Identification

[0082] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0083] Delivery of testimonials

[0084] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology in your area of ​​interest" is displayed.

[0085] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0086] Example 1

[0087] 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."

[0088] In today's internet usage environment, it is becoming increasingly difficult for users to efficiently find information that is relevant to them from the vast amount of information available. In particular, when searching for or browsing products, it is difficult to find the product that best suits one's interests from the vast amount of related information available, making it difficult to make purchasing decisions. This reduces user convenience and increases the risk of lost opportunities for sellers.

[0089] 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.

[0090] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's points of interest, means for generating a personalized introductory text using a template based on the inferred points of interest, means for tokenizing and normalizing text as preprocessing of the collected data when inferring the user's points of interest, and means for inserting information for each user into specific variable parts of the introductory text generated using the template. This improves user convenience and increases purchasing motivation by providing information that is likely to interest the user.

[0091] "User search history" refers to data that records the keywords or queries that a user enters into a search form on a website and executes the search, as well as the date and time that the search was performed.

[0092] "User browsing history" refers to data that records the IDs of product pages and content that a user has viewed on a website, as well as the date and time of such viewing.

[0093] "Review history" refers to data that records the ratings and comments posted by users on products and services on a website, as well as the date and time of those posts.

[0094] A "database" is a digital information repository that centrally stores information such as a user's search history, browsing history, and review history, and manages it so that it can be searched and analyzed later.

[0095] "Features" are specific elements or attributes that indicate a user's interests and concerns, extracted from collected user information, and serve as indicators for inferring the user's preferences based on these.

[0096] "Means for predicting preferences" refers to algorithms or machine learning models that analyze collected features and predict what products or information a user will be interested in.

[0097] A "template" is a template for inserting personalized content based on individual user information from a generic text format.

[0098] A "personalized description" is a product description generated using a template and customized based on the user's preferences.

[0099] "Tokenization" is a preprocessing technique that breaks down text data into its smallest units, such as words and phrases.

[0100] "Normalization" is a preprocessing technique that arranges text data into a unified format based on certain rules.

[0101] The present invention relates to a system for generating personalized testimonials based on a user's search history, browsing history, and word-of-mouth history. Specific embodiments of the system are described below.

[0102] 1. Collection of User Information

[0103] When a user searches, browses, or reviews on the site, their history is collected. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0104] For example, if a user searches for "smartphone," the search query "smartphone" and the search date and time are recorded. Next, when the user views a product page, the product ID and the date and time of the view are recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are also saved in the database.

[0105] 2. Guessing your favorite points

[0106] Features are extracted from the collected user information to infer the user's preferences. This process is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0107] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0108] 3. Generate personalized testimonials

[0109] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0110] As a specific example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" is generated.

[0111] 4. Displaying personalized testimonials

[0112] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the terminal retrieves the personalized introduction from the server and displays it on the product page.

[0113] As a specific example, when a user accesses a product page while logged in, an introductory text appears at the top of the page: "This product provides a detailed explanation of the latest AI technology that interests you."

[0114] Example prompts to input to a generative AI model:

[0115] "Generate product descriptions that this user might be interested in based on their search, browsing, and review history."

[0116] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized introductions, thereby realizing product introductions that are attractive to users.

[0117] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0118] Step 1: The user enters keywords into the search form on the site and performs a search. The device records the entered search query and the date and time the search was performed, and sends this information to the server.

[0119] Input: The search query entered by the user and the date and time of the search.

[0120] Output: Search query and search date and time data.

[0121] Specific operation: A search keyword is entered and the content is sent to the server.

[0122] Step 2: The terminal records the ID of the product page viewed by the user and the date and time of the view, and sends this information to the server.

[0123] Input: The ID of the product page the user viewed and the date and time of the view.

[0124] Output: Product page ID and viewed date and time data.

[0125] Specific operation: When a user views a product page, the page ID and the date and time of the view are sent to the server.

[0126] Step 3: When a user posts a product review, the device records the content and posting date and time, and sends that information to the server.

[0127] Input: User-submitted review content and posting date and time.

[0128] Output: Review content and posting date and time data.

[0129] Specific operation: When a user posts a review, the content and date and time are sent to the server.

[0130] Step 4: The server stores the collected user search history, browsing history, and review history in a database.

[0131] Input: Search history, browsing history, review history.

[0132] Output: User history data stored in a database.

[0133] Specific behavior: The collected history is stored in a centralized database.

[0134] Step 5: The server extracts features from the stored data. During this process, the data is preprocessed by tokenizing and normalizing the text.

[0135] Input: Saved user history data.

[0136] Output: Extracted feature data.

[0137] Specific operation: Tokenize and normalize text data to extract features.

[0138] Step 6: The server uses a machine learning model based on the extracted features to infer the user's preferences.

[0139] Input: Feature data.

[0140] Output: Inferred user preference points.

[0141] Specific behavior: Analyzes data using machine learning models to infer user interests and trends.

[0142] Step 7: The server generates a personalized introduction using a template based on the inferred preferences, with specific variables filled in with user-specific information.

[0143] Input: Inferred preference points, template.

[0144] Output: A personalized introduction.

[0145] Specific operation: Generate an introduction based on a template that matches the user's preferences.

[0146] Step 8: When the user accesses the product page, the terminal receives the personalized introduction from the server and displays it on the product page.

[0147] Input: A personalized testimonial.

[0148] Output: The description displayed on the product page.

[0149] Specific operation: When a user accesses a product page, the corresponding description is displayed.

[0150] (Application example 1)

[0151] 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."

[0152] Many online shopping sites today provide users with a uniform introduction, which lacks information tailored to each user's individual preferences. This prevents them from effectively supporting users' purchasing decisions, resulting in insufficient sales promotion effectiveness. Furthermore, when users select products, the lack of personalized information that reflects their past behavioral history also contributes to a poor user experience.

[0153] 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.

[0154] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's preferences, means for generating a personalized introduction using a template and a generative AI model based on the inferred preferences, and means for displaying the personalized introduction when the user accesses a product page. This makes it possible to provide individually optimized product introductions that utilize the user's past behavioral history, thereby improving the user experience and further increasing the effectiveness of sales promotion.

[0155] "Search History" means a record of keyword searches performed by a User on an Online Platform.

[0156] "Browsing history" is a record of the information a user views on an online platform.

[0157] "Review history" means a record of reviews and comments posted by a user on an online platform.

[0158] "Database" means a management system for storing collected user information in a structured manner.

[0159] "Features" are indicators that indicate user behavior and interests extracted from collected data.

[0160] "Preference points" are specific areas of interest or fields of interest that are inferred from a user's past behavioral history.

[0161] A "template" is a framework of text that serves as the basis for creating a personalized introduction.

[0162] A "generative AI model" is an artificial intelligence model that uses user information as input to generate personalized sentences and answers.

[0163] A "testimonial" is a product description or recommendation written based on the user's interests and concerns.

[0164] "Product Page" means a page on the Online Platform that displays detailed information about a particular Product.

[0165] To implement this invention, the system requires a server, a user terminal, and an internet connection environment. The specific operation and configuration of the system will be described below.

[0166] First, the server collects the user's search history, browsing history, and review history and stores this information in a database. For example, this can be done by collecting data via API using the Python requests library. A script running on the server monitors user behavior in real time and stores it in the database accordingly.

[0167] The server then uses a machine learning model to infer user preferences from the collected data. This model is built using machine learning libraries such as scikit-learn and TensorFlow, and natural language processing libraries such as NLTK and spaCy may be used to preprocess the text data.

[0168] Based on the inferred preference points, the server generates a personalized introduction using a template and a generative AI model. The generative AI model can use OpenAI's GPT or other large-scale language models. The template contains generic sentences, and the AI ​​model embeds appropriate variables to generate optimized sentences.

[0169] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. The user's device displays this description on the product page, allowing the user to obtain information that suits their preferences. This process is also performed by a front-end script implemented in JavaScript or similar.

[0170] For example, if a user frequently searches for and browses "smartphones" and "technical books," the system generates and displays a description such as "A product that provides detailed explanations of the latest smartphone technology."

[0171] Prompt Sentence Examples

[0172] An example prompt is:

[0173] “If a user is interested in technical books or smartphone-related products, generate a description that reflects that interest:

[0174] Product name: "Explaining the latest smartphone technology"

[0175] Highlights: "This book provides an in-depth look at cutting-edge smartphone technology. A must-have for engineers and technologists."

[0176] Give reasons why users would love this product."

[0177] This invention makes it possible to introduce individually optimized products by utilizing the user's past behavioral history, which is expected to improve the user experience and increase the effectiveness of sales promotion.

[0178] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0179] Step 1:

[0180] The server collects the user's search history, browsing history, and review history. Specifically, it monitors search queries sent from the device, pages viewed, and review content, and obtains this data through an API. The input is user operation data, and the output is stored in a database in a structured data format (e.g., JSON).

[0181] Step 2:

[0182] The server extracts features from user information stored in a database. Specifically, it tokenizes and normalizes the collected text data using a natural language processing library (e.g., NLTK, spaCy). The input is structured user data, and the output is saved as features (e.g., frequently used words, categories).

[0183] Step 3:

[0184] The server uses a machine learning model to predict the user's preferences based on the extracted features. Specifically, the features are input to a pre-trained model (e.g., scikit-learn, TensorFlow) and the preference points are output as a prediction result. The input is the feature data, and the output is a list of preference points.

[0185] Step 4:

[0186] The server generates a personalized introduction using a template and a generative AI model based on the inferred preference points. Specifically, it sets variables in the template and inputs a prompt to the generative AI model (e.g., OpenAI's GPT) to generate an optimized introduction. The inputs are the preference points, the template, and the prompt from the AI ​​model, and the output is a personalized introduction.

[0187] Step 5:

[0188] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. Specifically, it monitors the user's current access status, identifies the relevant product page, and inserts the generated description. The input is the access history and the generated description, and the output is the updated product page.

[0189] Step 6:

[0190] The terminal displays the personalized description received from the server on the product page. Specifically, it uses a front-end script such as JavaScript to insert the description into an HTML element. The input is the description data from the server, and the output is the product page displayed to the user.

[0191] Through the above steps, individually optimized product introductions are realized by utilizing the user's past behavioral history.

[0192] 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.

[0193] The present invention provides a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0194] 1. Collection of User Information

[0195] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0196] 2. Collecting Emotional Data

[0197] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine. Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to the server and stores them in a database. This allows the user's emotional state to be recorded in real time.

[0198] For example, when a user searches for "smartphone" and then views the product page for "Smartphone A," the device detects the emotion of joy from the user's facial expression and records it. Also, when a user posts a review saying that "Smartphone A" is "very good," positive emotions are detected from the text and this is also recorded on the server.

[0199] 3. Preprocessing user information and emotion data

[0200] To preprocess the collected search history, browsing history, and review history, as well as sentiment data, data cleaning, text tokenization, and normalization are performed. This preprocessing ensures that the data is stored in a consistent format, making later analysis easier.

[0201] 4. Feature extraction and preference estimation

[0202] The server identifies frequently occurring keywords and phrases from the formatted data, and further identifies categories of particular interest to the user based on the product category information viewed. These features and emotional data are then input into a machine learning model to predict the user's preferences. The emotional data also takes into account the user's emotional response to specific items.

[0203] As a concrete example, if a user searches for information about a large number of technical books and displays a positive emotional response, the server infers that the user is interested in detailed technical information and has positive feelings about it.

[0204] 5. Generate personalized testimonials

[0205] Based on the inferred preferences and emotional data, the server generates a testimonial using a template. The template contains generic sentences, and by substituting specific user information and emotional data for specific variables, a very specific and personalized testimonial is generated.

[0206] For example, for users interested in technical books, an introduction such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[0207] 6. Displaying personalized testimonials

[0208] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0209] For example, when a user accesses a product page while logged in, the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0210] As described above, the system of the present invention can provide a personalized introduction optimized for each user based on both user information and emotional data, and effectively communicate the appeal of a product.

[0211] The processing flow will be explained below.

[0212] Step 1: Collect user information

[0213] Collection of search history

[0214] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0215] Collection of browsing history

[0216] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0217] Collecting review history

[0218] When a user posts a review, the device records the content and date and time of the post. This information is also sent to the server and stored in the database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are saved.

[0219] Step 2: Collecting emotion data

[0220] Facial expression analysis

[0221] While the user is browsing a product page, the device captures the user's facial expressions with a camera and analyzes their emotions using facial recognition technology. The analysis results are sent to a server and stored in a database.

[0222] Audio analysis

[0223] When a user searches for a product name or writes a review by voice, the device collects the voice data and analyzes the emotion using voice recognition technology. The analysis results are sent to a server and stored in a database.

[0224] Text Analysis

[0225] When a user posts a review in text, the device analyzes the text data and extracts emotions using an emotion engine. The analysis results are sent to a server and stored in a database.

[0226] Step 3: Preprocessing user information and sentiment data

[0227] Data Formatting

[0228] The server converts the collected search history, browsing history, review history, and sentiment data into a single data format, allowing the data to be stored in a consistent format.

[0229] Text normalization

[0230] The server tokenizes and normalizes the text data, especially the reviews and sentiment data, removing unnecessary whitespace and special characters. This preprocessing makes it easier for machine learning models to analyze later.

[0231] Step 4: Feature extraction and preference estimation

[0232] Identifying frequent keywords

[0233] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses the words "latest" and "technology," these frequently occurring keywords are extracted as features.

[0234] Identifying Categories

[0235] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0236] Applying machine learning models

[0237] The server inputs the extracted features and emotion data into a machine learning model to predict the user's preferences. The emotion data also takes into account the user's emotional response to a particular item.

[0238] Recording points

[0239] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details and have positive feelings."

[0240] Step 5: Generate a personalized testimonial

[0241] Using templates

[0242] The server then inserts the estimated preferences and emotional data into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0243] Customizable wording

[0244] The server adds specific words and phrases based on the user's preferences. For example, it might generate a description like, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field. Many users have been impressed by its quality and have given it high ratings."

[0245] Step 6: Displaying a personalized testimonial

[0246] User Identification

[0247] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0248] Delivery of testimonials

[0249] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality" is displayed.

[0250] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0251] Example 2

[0252] 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."

[0253] While conventional systems can collect users' search history, browsing history, and review history, they are unable to generate personalized product recommendations that reflect the user's emotional state in real time. This makes it difficult to effectively recommend products that take the user's emotions into account, making it difficult to increase the user's purchase intention.

[0254] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0255] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and saving them in a database, means for collecting emotional data during user operations and sending it to the server to save in the database, means for extracting features from the collected user information and emotional data and inferring the user's favorite points, means for generating a personalized testimonial using a template based on the inferred favorite points and emotional data, and means for displaying the personalized testimonial when the user accesses a product page. This makes it possible to generate and display an optimized testimonial that takes user emotions into consideration in real time.

[0256] "User" means an individual who accesses the system and performs operations such as searching, browsing, and submitting reviews.

[0257] "Search History" means a record of the search queries a user makes within the system, along with the date and time of the queries.

[0258] "Viewing history" is a record of product identification information and the date and time of viewing when a user views a product page within the system.

[0259] "Review history" is a record of the content of reviews posted by users about products within the system and the date and time of posting.

[0260] "Emotional data" refers to emotional information analyzed from the user's facial expressions and voice collected during user operation.

[0261] A "database" is an electronic data structure for storing and managing collected user information and emotional data.

[0262] "Features" are important attributes or information extracted from collected data that are useful for analysis and prediction.

[0263] "Preference points" are specific categories or attributes that users are particularly interested in and express positive feelings about.

[0264] "Template" means a formatted document used to generate a personalized testimonial.

[0265] A "personalized testimonial" is a personalized testimonial that is generated to reflect a user's specific characteristics or emotional state.

[0266] "Server" means a computer system that receives, processes, stores, and analyzes data sent from a user's device.

[0267] "Terminal" means a device used by a User to access and operate the System.

[0268] The present invention is a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0269] Collection of User Information

[0270] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0271] Example: If a user searches for "Smartphone A" and then views the product page for "Smartphone A," the device records the search query, search date and time, product ID, and view date and time, and sends them to the server. The server stores this in a database.

[0272] Collecting Emotional Data

[0273] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine (such as OpenFace or IBM Watson). Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to a server to store in a database. This allows the user's emotional state to be recorded in real time.

[0274] Example: While a user is browsing a product page on "Smartphone A," the device's webcam captures the user's facial expressions, which are then analyzed by the emotion engine. Emotions such as joy and surprise are detected, and the analysis results are sent to the server and stored in a database.

[0275] Preprocessing user information and sentiment data

[0276] The server cleans the search history, browsing history, review history, and sentiment data stored in the database, and tokenizes and normalizes the text data, ensuring that the data is stored in a consistent format for easier later analysis and processing.

[0277] Feature extraction and preference estimation

[0278] The server identifies frequently occurring keywords and phrases from the formatted data and identifies categories that the user is particularly interested in. These features and sentiment data are input into a machine learning model (e.g., Scikit-learn or TensorFlow) to predict the user's preferences.

[0279] Example: If a user searches for a lot of information about technical books and displays positive sentiment, the server infers that the user is interested in technical details and has positive sentiment about them.

[0280] Generate personalized testimonials

[0281] The server generates a testimonial using a template based on the inferred preferences and emotional data. The template contains generic text, and specific, personalized testimonials can be generated by substituting user-specific information and emotional data into specific variable parts.

[0282] Example: For users interested in technical books, a description such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have also been impressed by its quality and have given it high ratings" may be generated.

[0283] Displaying a personalized testimonial

[0284] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0285] Example: When a user logs in and accesses the product page "Technical Book A," the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0286] As a result, the system of the present invention can provide an optimized personalized introduction based on both user information and emotional data, effectively conveying the appeal of the product.

[0287] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0288] Step 1:

[0289] Users take action on the site

[0290] Input: Users search, browse, and post reviews.

[0291] Actions: A user enters keywords into a search form, clicks the search button, browses a product page, and optionally submits a review. These actions take place within the site.

[0292] Output: User operation data (search query, viewed product ID, review content) is recorded.

[0293] Step 2:

[0294] The device records user operation data and sends it to the server.

[0295] Input: User operation data (search query, viewed product ID, review content).

[0296] How it works: The device records the search query and the date and time of the search when the user performs a search. Similarly, when the user views a product page, the device records the product ID and the date and time of the view. When the user posts a review, the device records the content of the review and the date and time of the post.

[0297] Output: The recorded user operation data is sent to the server and stored in a database.

[0298] Step 3:

[0299] The device collects the user's emotional data and sends it to the server.

[0300] Input: User's facial expression data, voice data.

[0301] How it works: The device's webcam and microphone are used to collect the user's facial expressions and voice in real time. The collected data is input into an emotion engine (OpenFace or IBM Watson) to analyze emotions. The analysis results are sent to a server.

[0302] Output: The analyzed emotion data is sent to the server and stored in a database.

[0303] Step 4:

[0304] The server preprocesses user information and emotion data

[0305] Input: Recorded user information (search history, browsing history, review history), analyzed emotional data.

[0306] How it works: The server cleans the raw data stored in the database, removing unnecessary information, then tokenizes and normalizes the text data.

[0307] Output: Preprocessed data in a consistent format.

[0308] Step 5:

[0309] The server extracts features and predicts the user's preferences.

[0310] Input: Preprocessed data.

[0311] How it works: The server extracts frequently occurring keywords and phrases from the data, extracts specific category information, and feeds this data into a machine learning model (Scikit-learn or TensorFlow) to predict user preferences.

[0312] Output: User preference points are inferred.

[0313] Step 6:

[0314] The server generates a personalized introduction

[0315] Input: Inferred preference points, sentiment data.

[0316] How it works: The server generates an introduction based on a template, inserting favorite points and emotional data.

[0317] Output: A personalized testimonial is generated.

[0318] Step 7:

[0319] The device displays a personalized introduction

[0320] Input: User ID, personalized introduction.

[0321] Operation: When a user accesses a product page, the device acquires the user ID and sends it to the server. The server then sends a personalized introduction corresponding to the user ID to the device. The device then displays the introduction on the product page.

[0322] Output: The user will see a personalized introduction.

[0323] (Application example 2)

[0324] 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."

[0325] Conventional online shopping sites recommend products based on a user's search history and browsing history, but do not take into account the user's emotional state. This makes it difficult to generate product reviews that reflect the user's emotions toward a particular product, and the effectiveness of these reviews in increasing the user's purchasing intent is not fully realized. Furthermore, there is a lack of systems that effectively utilize user emotional data, limiting the generation of personalized product reviews. Therefore, the present invention aims to solve these problems and provide a system that provides more effective personalized product reviews that reflect the user's emotional state.

[0326] 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 and storing a user's search history, browsing history, and review history in a database; means for extracting features from the collected user information and emotional data and inferring the user's preferences; and means for generating a personalized introduction using a template based on the inferred preferences and emotional state. This makes it possible to generate a personalized introduction that reflects the user's emotional state. Furthermore, when a user accesses a product page, an appropriate personalized introduction can be displayed, more effectively conveying the product's appeal.

[0327] "Search history" is a record of the search keywords and dates and times that a user searches on the Internet.

[0328] "Browsing history" refers to the IDs of web pages and products that a user has viewed on the Internet, as well as a record of the date and time of the views.

[0329] "Review history" refers to the ratings and reviews posted by users about specific products or services, as well as a record of their content.

[0330] "Emotional data" is information about a user's emotional state analyzed from their facial expressions, voice, and text.

[0331] A "feature" is a specific attribute or variable of the data input into a machine learning model, which represents a user's behavior or emotional state in numerical or categorical terms.

[0332] "Preference points" are elements or categories that a user is particularly interested in regarding a particular product or service, and are inferred from the user's behavioral and emotional data.

[0333] A "template" is a text template used to generate a personalized introduction, in which specific variable parts can be filled in with information for each user.

[0334] A "personalized testimonial" is a product or service testimonial that is individually optimized and generated based on the user's individual data (search history, browsing history, review history, emotional data, etc.).

[0335] A "machine learning model" is an algorithm or system that automatically learns patterns and relationships from data and makes predictions and classifications.

[0336] The present invention provides a system for collecting and analyzing a user's search history, browsing history, review history, and emotion data to generate personalized testimonials. Specific embodiments for implementing the present invention will be described below.

[0337] First, the system collects various data from users. When users search, browse, or post reviews on the site, the device records the search query, the viewed product ID and date and time, the review content, and the posting date and time, and sends this data to the server. This data is then stored in a database.

[0338] Next, emotional data is collected. The device analyzes emotions from facial expressions and voice while the user is using the site, and also extracts emotions from the text when posting reviews. This emotional data is also sent to the server and stored in a database.

[0339] The server pre-processes the stored search history, browsing history, review history, and sentiment data, including data cleaning, text tokenization, and normalization.

[0340] The server then uses a machine learning model to extract features from the preprocessed data and infer the user's preferences. In particular, it uses emotional data to consider the user's emotional reaction to a particular item. The machine learning model uses, for example, the Python library scikit-learn and the natural language processing library TextBlob.

[0341] Based on the inferred preferences and emotional data, the server uses a template to generate a personalized introductory text. This template contains generic text, and by substituting user-specific information and emotional data for specific variable parts, an introductory text optimized for each user is generated.

[0342] Finally, when a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies the appropriate personalized introduction based on the user ID and sends it to the device, allowing the user to view an introduction that reflects their preferences and emotions.

[0343] As a specific example, if a user searches for a technical book called "AI Technical Book" and then views product page "AI Technical Book B," the device will detect positive emotions from the user's facial expression data and record them. Furthermore, if the user posts a review stating "Very Good," positive emotions will also be detected from the text. Based on this data, a personalized introduction will be generated, such as, "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[0344] Example prompt sentence:

[0345] User Data:

[0346] Search history: ['AI technical books', 'machine learning']

[0347] View history: [{'item_id': '002', 'date': '2023-10-03'}, {'item_id': '003', 'date': '2023-10-04'}]

[0348] Review: [{'item_id': '002', 'review': 'Very Good', 'date': '2023-10-03'}]

[0349] Emotion data: [{'item_id': '002', 'emotion': 'joy', 'date': '2023-10-03'}]

[0350] Based on this data, generate a testimonial using the following template:

[0351] Template: "This product is specially designed for users interested in {}. {}\nMany users have been impressed with its quality."

[0352] Generated testimonial:

[0353] "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[0354] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0355] Step 1:

[0356] When a user searches, browses, or posts a review on the site, the device collects the search query, the viewed product ID, and the review content, and records the search date and time, the browse date and time, and the review posting date and time. This data is sent to the server and stored in a database. The input is the user's actions, and the output is the history data stored in the database.

[0357] Step 2:

[0358] The device analyzes the user's facial expressions and voice in real time and extracts emotional data using an emotion engine. When a review is posted, emotions are also analyzed from the text data, and this emotional data is sent to the server and stored in a database. The input is the user's facial expressions, voice, and text data, and the output is the extracted emotional data.

[0359] Step 3:

[0360] The server preprocesses the collected search history, browsing history, review history, and sentiment data. This preprocessing involves data cleaning, text tokenization, normalization, and formatting into a consistent format. The input is the raw data stored in the database, and the output is the preprocessed data.

[0361] Step 4:

[0362] The server extracts features from the preprocessed data and uses a machine learning model to infer the user's preference points. The libraries used are Python's scikit-learn and TextBlob, which also take into account the user's emotional response to specific items. The input is the preprocessed data, and the output is the inferred preference points.

[0363] Step 5:

[0364] The server generates a personalized introduction using a pre-prepared template based on the inferred preferences and emotion data. The template has variable sections that allow specific user information to be inserted. The input is the preferences and template, and the output is a personalized introduction.

[0365] Step 6:

[0366] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. The input is the user ID, and the output is the appropriate personalized introduction.

[0367] Step 7:

[0368] The terminal displays the personalized introduction received from the server on the product page. The input is the introduction received from the server, and the output is the introduction displayed on the product page. This allows the user to see an introduction that reflects their own preferences and emotions.

[0369] 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.

[0370] 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.

[0371] 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.

[0372] [Second embodiment]

[0373] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0374] 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.

[0375] 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).

[0376] 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.

[0377] 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.

[0378] 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).

[0379] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0380] 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.

[0381] 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.

[0382] 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.

[0383] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0384] 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."

[0385] The present invention provides a system for generating personalized testimonials based on a user's search history, browsing history, and review history, and the specific embodiments thereof will be described below.

[0386] 1. Collection of User Information

[0387] A special script is run to collect the user's search, browsing, and review history within the site. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0388] For example, if a user searches for "smartphone," the search query "smartphone" and the date and time of the search are recorded. If the user then views a product page, the product ID and the date and time of the view are also recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are saved in the database.

[0389] 2. Guessing your favorite points

[0390] Features are extracted from the collected user information to infer the user's preferences. This process is performed using a machine learning model and is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0391] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0392] 3. Generate personalized testimonials

[0393] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0394] For example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" will be generated.

[0395] 4. Displaying personalized testimonials

[0396] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the device retrieves the personalized introduction from the server and displays it on the product page.

[0397] For example, when a user accesses a product page while logged in, a description will appear at the top of the page stating, "This product provides a detailed explanation of the latest AI technology that interests you."

[0398] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized product introductions, thereby realizing product introductions that are attractive to users.

[0399] The processing flow will be explained below.

[0400] Step 1: Collect user information

[0401] Collection of search history

[0402] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0403] Collection of browsing history

[0404] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0405] Collecting review history

[0406] When a user posts a review, the device records the content and date and time of the post. This information is sent to the server and stored in a database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are stored.

[0407] Step 2: Preprocessing user information

[0408] Data Formatting

[0409] The server formats the collected search history, browsing history, and review history into a single data format. This preprocessing ensures that the data is stored in a consistent format.

[0410] Text Preprocessing

[0411] The server tokenizes and normalizes the text data, especially the reviews, removing unnecessary whitespace and special characters, which makes it easier for machine learning models to analyze later.

[0412] Step 3: Feature extraction

[0413] Identifying frequent keywords

[0414] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses words like "latest" and "technology," these frequently occurring keywords are extracted as features.

[0415] Identifying Categories

[0416] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0417] Step 4: Guess your preference points

[0418] Applying machine learning models

[0419] The server inputs the extracted features into a machine learning model to predict the user's preferences. Because the model learns from past data, it can predict the user's preferences with high accuracy.

[0420] Recording points

[0421] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details."

[0422] Step 5: Generate a personalized testimonial

[0423] Using templates

[0424] The server then inserts the inferred preferences into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0425] Customizable wording

[0426] The server adds specific words and phrases based on the user's preferences, such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field."

[0427] Step 6: Displaying a personalized testimonial

[0428] User Identification

[0429] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0430] Delivery of testimonials

[0431] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology in your area of ​​interest" is displayed.

[0432] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0433] Example 1

[0434] 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."

[0435] In today's internet usage environment, it is becoming increasingly difficult for users to efficiently find information that is relevant to them from the vast amount of information available. In particular, when searching for or browsing products, it is difficult to find the product that best suits one's interests from the vast amount of related information available, making it difficult to make purchasing decisions. This reduces user convenience and increases the risk of lost opportunities for sellers.

[0436] 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.

[0437] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's points of interest, means for generating a personalized introductory text using a template based on the inferred points of interest, means for tokenizing and normalizing text as preprocessing of the collected data when inferring the user's points of interest, and means for inserting information for each user into specific variable parts of the introductory text generated using the template. This improves user convenience and increases purchasing motivation by providing information that is likely to interest the user.

[0438] "User search history" refers to data that records the keywords or queries that a user enters into a search form on a website and executes the search, as well as the date and time that the search was performed.

[0439] "User browsing history" refers to data that records the IDs of product pages and content that a user has viewed on a website, as well as the date and time of such viewing.

[0440] "Review history" refers to data that records the ratings and comments posted by users on products and services on a website, as well as the date and time of those posts.

[0441] A "database" is a digital information repository that centrally stores information such as a user's search history, browsing history, and review history, and manages it so that it can be searched and analyzed later.

[0442] "Features" are specific elements or attributes that indicate a user's interests and concerns, extracted from collected user information, and serve as indicators for inferring the user's preferences based on these.

[0443] "Means for predicting preferences" refers to algorithms or machine learning models that analyze collected features and predict what products or information a user will be interested in.

[0444] A "template" is a template for inserting personalized content based on individual user information from a generic text format.

[0445] A "personalized description" is a product description generated using a template and customized based on the user's preferences.

[0446] "Tokenization" is a preprocessing technique that breaks down text data into its smallest units, such as words and phrases.

[0447] "Normalization" is a preprocessing technique that arranges text data into a unified format based on certain rules.

[0448] The present invention relates to a system for generating personalized testimonials based on a user's search history, browsing history, and word-of-mouth history. Specific embodiments of the system are described below.

[0449] 1. Collection of User Information

[0450] When a user searches, browses, or reviews on the site, their history is collected. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0451] For example, if a user searches for "smartphone," the search query "smartphone" and the search date and time are recorded. Next, when the user views a product page, the product ID and the date and time of the view are recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are also saved in the database.

[0452] 2. Guessing your favorite points

[0453] Features are extracted from the collected user information to infer the user's preferences. This process is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0454] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0455] 3. Generate personalized testimonials

[0456] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0457] As a specific example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" is generated.

[0458] 4. Displaying personalized testimonials

[0459] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the terminal retrieves the personalized introduction from the server and displays it on the product page.

[0460] As a specific example, when a user accesses a product page while logged in, an introductory text appears at the top of the page: "This product provides a detailed explanation of the latest AI technology that interests you."

[0461] Example prompts to input to a generative AI model:

[0462] "Generate product descriptions that this user might be interested in based on their search, browsing, and review history."

[0463] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized introductions, thereby realizing product introductions that are attractive to users.

[0464] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0465] Step 1: The user enters keywords into the search form on the site and performs a search. The device records the entered search query and the date and time the search was performed, and sends this information to the server.

[0466] Input: The search query entered by the user and the date and time of the search.

[0467] Output: Search query and search date and time data.

[0468] Specific operation: A search keyword is entered and the content is sent to the server.

[0469] Step 2: The terminal records the ID of the product page viewed by the user and the date and time of the view, and sends this information to the server.

[0470] Input: The ID of the product page the user viewed and the date and time of the view.

[0471] Output: Product page ID and viewed date and time data.

[0472] Specific operation: When a user views a product page, the page ID and the date and time of the view are sent to the server.

[0473] Step 3: When a user posts a product review, the device records the content and posting date and time, and sends that information to the server.

[0474] Input: User-submitted review content and posting date and time.

[0475] Output: Review content and posting date and time data.

[0476] Specific operation: When a user posts a review, the content and date and time are sent to the server.

[0477] Step 4: The server stores the collected user search history, browsing history, and review history in a database.

[0478] Input: Search history, browsing history, review history.

[0479] Output: User history data stored in a database.

[0480] Specific behavior: The collected history is stored in a centralized database.

[0481] Step 5: The server extracts features from the stored data. During this process, the data is preprocessed by tokenizing and normalizing the text.

[0482] Input: Saved user history data.

[0483] Output: Extracted feature data.

[0484] Specific operation: Tokenize and normalize text data to extract features.

[0485] Step 6: The server uses a machine learning model based on the extracted features to infer the user's preferences.

[0486] Input: Feature data.

[0487] Output: Inferred user preference points.

[0488] Specific behavior: Analyzes data using machine learning models to infer user interests and trends.

[0489] Step 7: The server generates a personalized introduction using a template based on the inferred preferences, with specific variables filled in with user-specific information.

[0490] Input: Inferred preference points, template.

[0491] Output: A personalized introduction.

[0492] Specific operation: Generate an introduction based on a template that matches the user's preferences.

[0493] Step 8: When the user accesses the product page, the terminal receives the personalized introduction from the server and displays it on the product page.

[0494] Input: A personalized testimonial.

[0495] Output: The description displayed on the product page.

[0496] Specific operation: When a user accesses a product page, the corresponding description is displayed.

[0497] (Application example 1)

[0498] 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."

[0499] Many online shopping sites today provide users with a uniform introduction, which lacks information tailored to each user's individual preferences. This prevents them from effectively supporting users' purchasing decisions, resulting in insufficient sales promotion effectiveness. Furthermore, when users select products, the lack of personalized information that reflects their past behavioral history also contributes to a poor user experience.

[0500] 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.

[0501] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's preferences, means for generating a personalized introduction using a template and a generative AI model based on the inferred preferences, and means for displaying the personalized introduction when the user accesses a product page. This makes it possible to provide individually optimized product introductions that utilize the user's past behavioral history, thereby improving the user experience and further increasing the effectiveness of sales promotion.

[0502] "Search History" means a record of keyword searches performed by a User on an Online Platform.

[0503] "Browsing history" is a record of the information a user views on an online platform.

[0504] "Review history" means a record of reviews and comments posted by a user on an online platform.

[0505] "Database" means a management system for storing collected user information in a structured manner.

[0506] "Features" are indicators that indicate user behavior and interests extracted from collected data.

[0507] "Preference points" are specific areas of interest or fields of interest that are inferred from a user's past behavioral history.

[0508] A "template" is a framework of text that serves as the basis for creating a personalized introduction.

[0509] A "generative AI model" is an artificial intelligence model that uses user information as input to generate personalized sentences and answers.

[0510] A "testimonial" is a product description or recommendation written based on the user's interests and concerns.

[0511] "Product Page" means a page on the Online Platform that displays detailed information about a particular Product.

[0512] To implement this invention, the system requires a server, a user terminal, and an internet connection environment. The specific operation and configuration of the system will be described below.

[0513] First, the server collects the user's search history, browsing history, and review history and stores this information in a database. For example, this can be done by collecting data via API using the Python requests library. A script running on the server monitors user behavior in real time and stores it in the database accordingly.

[0514] The server then uses a machine learning model to infer user preferences from the collected data. This model is built using machine learning libraries such as scikit-learn and TensorFlow, and natural language processing libraries such as NLTK and spaCy may be used to preprocess the text data.

[0515] Based on the inferred preference points, the server generates a personalized introduction using a template and a generative AI model. The generative AI model can use OpenAI's GPT or other large-scale language models. The template contains generic sentences, and the AI ​​model embeds appropriate variables to generate optimized sentences.

[0516] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. The user's device displays this description on the product page, allowing the user to obtain information that suits their preferences. This process is also performed by a front-end script implemented in JavaScript or similar.

[0517] For example, if a user frequently searches for and browses "smartphones" and "technical books," the system generates and displays a description such as "A product that provides detailed explanations of the latest smartphone technology."

[0518] Prompt Sentence Examples

[0519] An example prompt is:

[0520] “If a user is interested in technical books or smartphone-related products, generate a description that reflects that interest:

[0521] Product name: "Explaining the latest smartphone technology"

[0522] Highlights: "This book provides an in-depth look at cutting-edge smartphone technology. A must-have for engineers and technologists."

[0523] Give reasons why users would love this product."

[0524] This invention makes it possible to introduce individually optimized products by utilizing the user's past behavioral history, which is expected to improve the user experience and increase the effectiveness of sales promotion.

[0525] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0526] Step 1:

[0527] The server collects the user's search history, browsing history, and review history. Specifically, it monitors search queries sent from the device, pages viewed, and review content, and obtains this data through an API. The input is user operation data, and the output is stored in a database in a structured data format (e.g., JSON).

[0528] Step 2:

[0529] The server extracts features from user information stored in a database. Specifically, it tokenizes and normalizes the collected text data using a natural language processing library (e.g., NLTK, spaCy). The input is structured user data, and the output is saved as features (e.g., frequently used words, categories).

[0530] Step 3:

[0531] The server uses a machine learning model to predict the user's preferences based on the extracted features. Specifically, the features are input to a pre-trained model (e.g., scikit-learn, TensorFlow) and the preference points are output as a prediction result. The input is the feature data, and the output is a list of preference points.

[0532] Step 4:

[0533] The server generates a personalized introduction using a template and a generative AI model based on the inferred preference points. Specifically, it sets variables in the template and inputs a prompt to the generative AI model (e.g., OpenAI's GPT) to generate an optimized introduction. The inputs are the preference points, the template, and the prompt from the AI ​​model, and the output is a personalized introduction.

[0534] Step 5:

[0535] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. Specifically, it monitors the user's current access status, identifies the relevant product page, and inserts the generated description. The input is the access history and the generated description, and the output is the updated product page.

[0536] Step 6:

[0537] The terminal displays the personalized description received from the server on the product page. Specifically, it uses a front-end script such as JavaScript to insert the description into an HTML element. The input is the description data from the server, and the output is the product page displayed to the user.

[0538] Through the above steps, individually optimized product introductions are realized by utilizing the user's past behavioral history.

[0539] 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.

[0540] The present invention provides a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0541] 1. Collection of User Information

[0542] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0543] 2. Collecting Emotional Data

[0544] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine. Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to the server and stores them in a database. This allows the user's emotional state to be recorded in real time.

[0545] For example, when a user searches for "smartphone" and then views the product page for "Smartphone A," the device detects the emotion of joy from the user's facial expression and records it. Also, when a user posts a review saying that "Smartphone A" is "very good," positive emotions are detected from the text and this is also recorded on the server.

[0546] 3. Preprocessing user information and emotion data

[0547] To preprocess the collected search history, browsing history, and review history, as well as sentiment data, data cleaning, text tokenization, and normalization are performed. This preprocessing ensures that the data is stored in a consistent format, making later analysis easier.

[0548] 4. Feature extraction and preference estimation

[0549] The server identifies frequently occurring keywords and phrases from the formatted data, and further identifies categories of particular interest to the user based on the product category information viewed. These features and emotional data are then input into a machine learning model to predict the user's preferences. The emotional data also takes into account the user's emotional response to specific items.

[0550] As a concrete example, if a user searches for information about a large number of technical books and displays a positive emotional response, the server infers that the user is interested in detailed technical information and has positive feelings about it.

[0551] 5. Generate personalized testimonials

[0552] Based on the inferred preferences and emotional data, the server generates a testimonial using a template. The template contains generic sentences, and by substituting specific user information and emotional data for specific variables, a very specific and personalized testimonial is generated.

[0553] For example, for users interested in technical books, an introduction such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[0554] 6. Displaying personalized testimonials

[0555] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0556] For example, when a user accesses a product page while logged in, the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0557] As described above, the system of the present invention can provide a personalized introduction optimized for each user based on both user information and emotional data, and effectively communicate the appeal of a product.

[0558] The processing flow will be explained below.

[0559] Step 1: Collect user information

[0560] Collection of search history

[0561] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0562] Collection of browsing history

[0563] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0564] Collecting review history

[0565] When a user posts a review, the device records the content and date and time of the post. This information is also sent to the server and stored in the database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are saved.

[0566] Step 2: Collecting emotion data

[0567] Facial expression analysis

[0568] While the user is browsing a product page, the device captures the user's facial expressions with a camera and analyzes their emotions using facial recognition technology. The analysis results are sent to a server and stored in a database.

[0569] Audio analysis

[0570] When a user searches for a product name or writes a review by voice, the device collects the voice data and analyzes the emotion using voice recognition technology. The analysis results are sent to a server and stored in a database.

[0571] Text Analysis

[0572] When a user posts a review in text, the device analyzes the text data and extracts emotions using an emotion engine. The analysis results are sent to a server and stored in a database.

[0573] Step 3: Preprocessing user information and sentiment data

[0574] Data Formatting

[0575] The server converts the collected search history, browsing history, review history, and sentiment data into a single data format, allowing the data to be stored in a consistent format.

[0576] Text normalization

[0577] The server tokenizes and normalizes the text data, especially the reviews and sentiment data, removing unnecessary whitespace and special characters. This preprocessing makes it easier for machine learning models to analyze later.

[0578] Step 4: Feature extraction and preference estimation

[0579] Identifying frequent keywords

[0580] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses the words "latest" and "technology," these frequently occurring keywords are extracted as features.

[0581] Identifying Categories

[0582] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0583] Applying machine learning models

[0584] The server inputs the extracted features and emotion data into a machine learning model to predict the user's preferences. The emotion data also takes into account the user's emotional response to a particular item.

[0585] Recording points

[0586] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details and have positive feelings."

[0587] Step 5: Generate a personalized testimonial

[0588] Using templates

[0589] The server then inserts the estimated preferences and emotional data into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0590] Customizable wording

[0591] The server adds specific words and phrases based on the user's preferences. For example, it might generate a description like, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field. Many users have been impressed by its quality and have given it high ratings."

[0592] Step 6: Displaying a personalized testimonial

[0593] User Identification

[0594] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0595] Delivery of testimonials

[0596] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality" is displayed.

[0597] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0598] Example 2

[0599] 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."

[0600] While conventional systems can collect users' search history, browsing history, and review history, they are unable to generate personalized product recommendations that reflect the user's emotional state in real time. This makes it difficult to effectively recommend products that take the user's emotions into account, making it difficult to increase the user's purchase intention.

[0601] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0602] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and saving them in a database, means for collecting emotional data during user operations and sending it to the server to save in the database, means for extracting features from the collected user information and emotional data and inferring the user's favorite points, means for generating a personalized testimonial using a template based on the inferred favorite points and emotional data, and means for displaying the personalized testimonial when the user accesses a product page. This makes it possible to generate and display an optimized testimonial that takes user emotions into consideration in real time.

[0603] "User" means an individual who accesses the system and performs operations such as searching, browsing, and submitting reviews.

[0604] "Search History" means a record of the search queries a user makes within the system, along with the date and time of the queries.

[0605] "Viewing history" is a record of product identification information and the date and time of viewing when a user views a product page within the system.

[0606] "Review history" is a record of the content of reviews posted by users about products within the system and the date and time of posting.

[0607] "Emotional data" refers to emotional information analyzed from the user's facial expressions and voice collected during user operation.

[0608] A "database" is an electronic data structure for storing and managing collected user information and emotional data.

[0609] "Features" are important attributes or information extracted from collected data that are useful for analysis and prediction.

[0610] "Preference points" are specific categories or attributes that users are particularly interested in and express positive feelings about.

[0611] "Template" means a formatted document used to generate a personalized testimonial.

[0612] A "personalized testimonial" is a personalized testimonial that is generated to reflect a user's specific characteristics or emotional state.

[0613] "Server" means a computer system that receives, processes, stores, and analyzes data sent from a user's device.

[0614] "Terminal" means a device used by a User to access and operate the System.

[0615] The present invention is a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0616] Collection of User Information

[0617] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0618] Example: If a user searches for "Smartphone A" and then views the product page for "Smartphone A," the device records the search query, search date and time, product ID, and view date and time, and sends them to the server. The server stores this in a database.

[0619] Collecting Emotional Data

[0620] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine (such as OpenFace or IBM Watson). Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to a server to store in a database. This allows the user's emotional state to be recorded in real time.

[0621] Example: While a user is browsing a product page on "Smartphone A," the device's webcam captures the user's facial expressions, which are then analyzed by the emotion engine. Emotions such as joy and surprise are detected, and the analysis results are sent to the server and stored in a database.

[0622] Preprocessing user information and sentiment data

[0623] The server cleans the search history, browsing history, review history, and sentiment data stored in the database, and tokenizes and normalizes the text data, ensuring that the data is stored in a consistent format for easier later analysis and processing.

[0624] Feature extraction and preference estimation

[0625] The server identifies frequently occurring keywords and phrases from the formatted data and identifies categories that the user is particularly interested in. These features and sentiment data are input into a machine learning model (e.g., Scikit-learn or TensorFlow) to predict the user's preferences.

[0626] Example: If a user searches for a lot of information about technical books and displays positive sentiment, the server infers that the user is interested in technical details and has positive sentiment about them.

[0627] Generate personalized testimonials

[0628] The server generates a testimonial using a template based on the inferred preferences and emotional data. The template contains generic text, and specific, personalized testimonials can be generated by substituting user-specific information and emotional data into specific variable parts.

[0629] Example: For users interested in technical books, a description such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have also been impressed by its quality and have given it high ratings" may be generated.

[0630] Displaying a personalized testimonial

[0631] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0632] Example: When a user logs in and accesses the product page "Technical Book A," the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0633] As a result, the system of the present invention can provide an optimized personalized introduction based on both user information and emotional data, effectively conveying the appeal of the product.

[0634] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0635] Step 1:

[0636] Users take action on the site

[0637] Input: Users search, browse, and post reviews.

[0638] Actions: A user enters keywords into a search form, clicks the search button, browses a product page, and optionally submits a review. These actions take place within the site.

[0639] Output: User operation data (search query, viewed product ID, review content) is recorded.

[0640] Step 2:

[0641] The device records user operation data and sends it to the server.

[0642] Input: User operation data (search query, viewed product ID, review content).

[0643] How it works: The device records the search query and the date and time of the search when the user performs a search. Similarly, when the user views a product page, the device records the product ID and the date and time of the view. When the user posts a review, the device records the content of the review and the date and time of the post.

[0644] Output: The recorded user operation data is sent to the server and stored in a database.

[0645] Step 3:

[0646] The device collects the user's emotional data and sends it to the server.

[0647] Input: User's facial expression data, voice data.

[0648] How it works: The device's webcam and microphone are used to collect the user's facial expressions and voice in real time. The collected data is input into an emotion engine (OpenFace or IBM Watson) to analyze emotions. The analysis results are sent to a server.

[0649] Output: The analyzed emotion data is sent to the server and stored in a database.

[0650] Step 4:

[0651] The server preprocesses user information and emotion data

[0652] Input: Recorded user information (search history, browsing history, review history), analyzed emotional data.

[0653] How it works: The server cleans the raw data stored in the database, removing unnecessary information, then tokenizes and normalizes the text data.

[0654] Output: Preprocessed data in a consistent format.

[0655] Step 5:

[0656] The server extracts features and predicts the user's preferences.

[0657] Input: Preprocessed data.

[0658] How it works: The server extracts frequently occurring keywords and phrases from the data, extracts specific category information, and feeds this data into a machine learning model (Scikit-learn or TensorFlow) to predict user preferences.

[0659] Output: User preference points are inferred.

[0660] Step 6:

[0661] The server generates a personalized introduction

[0662] Input: Inferred preference points, sentiment data.

[0663] How it works: The server generates an introduction based on a template, inserting favorite points and emotional data.

[0664] Output: A personalized testimonial is generated.

[0665] Step 7:

[0666] The device displays a personalized introduction

[0667] Input: User ID, personalized introduction.

[0668] Operation: When a user accesses a product page, the device acquires the user ID and sends it to the server. The server then sends a personalized introduction corresponding to the user ID to the device. The device then displays the introduction on the product page.

[0669] Output: The user will see a personalized introduction.

[0670] (Application example 2)

[0671] 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."

[0672] Conventional online shopping sites recommend products based on a user's search history and browsing history, but do not take into account the user's emotional state. This makes it difficult to generate product reviews that reflect the user's emotions toward a particular product, and the effectiveness of these reviews in increasing the user's purchasing intent is not fully realized. Furthermore, there is a lack of systems that effectively utilize user emotional data, limiting the generation of personalized product reviews. Therefore, the present invention aims to solve these problems and provide a system that provides more effective personalized product reviews that reflect the user's emotional state.

[0673] 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 and storing a user's search history, browsing history, and review history in a database; means for extracting features from the collected user information and emotional data and inferring the user's preferences; and means for generating a personalized introduction using a template based on the inferred preferences and emotional state. This makes it possible to generate a personalized introduction that reflects the user's emotional state. Furthermore, when a user accesses a product page, an appropriate personalized introduction can be displayed, more effectively conveying the product's appeal.

[0674] "Search history" is a record of the search keywords and dates and times that a user searches on the Internet.

[0675] "Browsing history" refers to the IDs of web pages and products that a user has viewed on the Internet, as well as a record of the date and time of the views.

[0676] "Review history" refers to the ratings and reviews posted by users about specific products or services, as well as a record of their content.

[0677] "Emotional data" is information about a user's emotional state analyzed from their facial expressions, voice, and text.

[0678] A "feature" is a specific attribute or variable of the data input into a machine learning model, which represents a user's behavior or emotional state in numerical or categorical terms.

[0679] "Preference points" are elements or categories that a user is particularly interested in regarding a particular product or service, and are inferred from the user's behavioral and emotional data.

[0680] A "template" is a text template used to generate a personalized introduction, in which specific variable parts can be filled in with information for each user.

[0681] A "personalized testimonial" is a product or service testimonial that is individually optimized and generated based on the user's individual data (search history, browsing history, review history, emotional data, etc.).

[0682] A "machine learning model" is an algorithm or system that automatically learns patterns and relationships from data and makes predictions and classifications.

[0683] The present invention provides a system for collecting and analyzing a user's search history, browsing history, review history, and emotion data to generate personalized testimonials. Specific embodiments for implementing the present invention will be described below.

[0684] First, the system collects various data from users. When users search, browse, or post reviews on the site, the device records the search query, the viewed product ID and date and time, the review content, and the posting date and time, and sends this data to the server. This data is then stored in a database.

[0685] Next, emotional data is collected. The device analyzes emotions from facial expressions and voice while the user is using the site, and also extracts emotions from the text when posting reviews. This emotional data is also sent to the server and stored in a database.

[0686] The server pre-processes the stored search history, browsing history, review history, and sentiment data, including data cleaning, text tokenization, and normalization.

[0687] The server then uses a machine learning model to extract features from the preprocessed data and infer the user's preferences. In particular, it uses emotional data to consider the user's emotional reaction to a particular item. The machine learning model uses, for example, the Python library scikit-learn and the natural language processing library TextBlob.

[0688] Based on the inferred preferences and emotional data, the server uses a template to generate a personalized introductory text. This template contains generic text, and by substituting user-specific information and emotional data for specific variable parts, an introductory text optimized for each user is generated.

[0689] Finally, when a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies the appropriate personalized introduction based on the user ID and sends it to the device, allowing the user to view an introduction that reflects their preferences and emotions.

[0690] As a specific example, if a user searches for a technical book called "AI Technical Book" and then views product page "AI Technical Book B," the device will detect positive emotions from the user's facial expression data and record them. Furthermore, if the user posts a review stating "Very Good," positive emotions will also be detected from the text. Based on this data, a personalized introduction will be generated, such as, "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[0691] Example prompt sentence:

[0692] User Data:

[0693] Search history: ['AI technical books', 'machine learning']

[0694] View history: [{'item_id': '002', 'date': '2023-10-03'}, {'item_id': '003', 'date': '2023-10-04'}]

[0695] Review: [{'item_id': '002', 'review': 'Very Good', 'date': '2023-10-03'}]

[0696] Emotion data: [{'item_id': '002', 'emotion': 'joy', 'date': '2023-10-03'}]

[0697] Based on this data, generate a testimonial using the following template:

[0698] Template: "This product is specially designed for users interested in {}. {}\nMany users have been impressed with its quality."

[0699] Generated testimonial:

[0700] "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[0701] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0702] Step 1:

[0703] When a user searches, browses, or posts a review on the site, the device collects the search query, the viewed product ID, and the review content, and records the search date and time, the browse date and time, and the review posting date and time. This data is sent to the server and stored in a database. The input is the user's actions, and the output is the history data stored in the database.

[0704] Step 2:

[0705] The device analyzes the user's facial expressions and voice in real time and extracts emotional data using an emotion engine. When a review is posted, emotions are also analyzed from the text data, and this emotional data is sent to the server and stored in a database. The input is the user's facial expressions, voice, and text data, and the output is the extracted emotional data.

[0706] Step 3:

[0707] The server preprocesses the collected search history, browsing history, review history, and sentiment data. This preprocessing involves data cleaning, text tokenization, normalization, and formatting into a consistent format. The input is the raw data stored in the database, and the output is the preprocessed data.

[0708] Step 4:

[0709] The server extracts features from the preprocessed data and uses a machine learning model to infer the user's preference points. The libraries used are Python's scikit-learn and TextBlob, which also take into account the user's emotional response to specific items. The input is the preprocessed data, and the output is the inferred preference points.

[0710] Step 5:

[0711] The server generates a personalized introduction using a pre-prepared template based on the inferred preferences and emotion data. The template has variable sections that allow specific user information to be inserted. The input is the preferences and template, and the output is a personalized introduction.

[0712] Step 6:

[0713] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. The input is the user ID, and the output is the appropriate personalized introduction.

[0714] Step 7:

[0715] The terminal displays the personalized introduction received from the server on the product page. The input is the introduction received from the server, and the output is the introduction displayed on the product page. This allows the user to see an introduction that reflects their own preferences and emotions.

[0716] 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.

[0717] 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.

[0718] 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.

[0719] [Third embodiment]

[0720] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0721] 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.

[0722] 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).

[0723] 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.

[0724] 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.

[0725] 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).

[0726] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0727] 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.

[0728] 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.

[0729] 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.

[0730] 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.

[0731] 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."

[0732] The present invention provides a system for generating personalized testimonials based on a user's search history, browsing history, and review history, and the specific embodiments thereof will be described below.

[0733] 1. Collection of User Information

[0734] A special script is run to collect the user's search, browsing, and review history within the site. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0735] For example, if a user searches for "smartphone," the search query "smartphone" and the date and time of the search are recorded. If the user then views a product page, the product ID and the date and time of the view are also recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are saved in the database.

[0736] 2. Guessing your favorite points

[0737] Features are extracted from the collected user information to infer the user's preferences. This process is performed using a machine learning model and is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0738] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0739] 3. Generate personalized testimonials

[0740] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0741] For example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" will be generated.

[0742] 4. Displaying personalized testimonials

[0743] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the device retrieves the personalized introduction from the server and displays it on the product page.

[0744] For example, when a user accesses a product page while logged in, a description will appear at the top of the page stating, "This product provides a detailed explanation of the latest AI technology that interests you."

[0745] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized product introductions, thereby realizing product introductions that are attractive to users.

[0746] The processing flow will be explained below.

[0747] Step 1: Collect user information

[0748] Collection of search history

[0749] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0750] Collection of browsing history

[0751] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0752] Collecting review history

[0753] When a user posts a review, the device records the content and date and time of the post. This information is sent to the server and stored in a database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are stored.

[0754] Step 2: Preprocessing user information

[0755] Data Formatting

[0756] The server formats the collected search history, browsing history, and review history into a single data format. This preprocessing ensures that the data is stored in a consistent format.

[0757] Text Preprocessing

[0758] The server tokenizes and normalizes the text data, especially the reviews, removing unnecessary whitespace and special characters, which makes it easier for machine learning models to analyze later.

[0759] Step 3: Feature extraction

[0760] Identifying frequent keywords

[0761] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses words like "latest" and "technology," these frequently occurring keywords are extracted as features.

[0762] Identifying Categories

[0763] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0764] Step 4: Guess your preference points

[0765] Applying machine learning models

[0766] The server inputs the extracted features into a machine learning model to predict the user's preferences. Because the model learns from past data, it can predict the user's preferences with high accuracy.

[0767] Recording points

[0768] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details."

[0769] Step 5: Generate a personalized testimonial

[0770] Using templates

[0771] The server then inserts the inferred preferences into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0772] Customizable wording

[0773] The server adds specific words and phrases based on the user's preferences, such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field."

[0774] Step 6: Displaying a personalized testimonial

[0775] User Identification

[0776] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0777] Delivery of testimonials

[0778] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology in your area of ​​interest" is displayed.

[0779] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0780] Example 1

[0781] 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."

[0782] In today's internet usage environment, it is becoming increasingly difficult for users to efficiently find information that is relevant to them from the vast amount of information available. In particular, when searching for or browsing products, it is difficult to find the product that best suits one's interests from the vast amount of related information available, making it difficult to make purchasing decisions. This reduces user convenience and increases the risk of lost opportunities for sellers.

[0783] 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.

[0784] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's points of interest, means for generating a personalized introductory text using a template based on the inferred points of interest, means for tokenizing and normalizing text as preprocessing of the collected data when inferring the user's points of interest, and means for inserting information for each user into specific variable parts of the introductory text generated using the template. This improves user convenience and increases purchasing motivation by providing information that is likely to interest the user.

[0785] "User search history" refers to data that records the keywords or queries that a user enters into a search form on a website and executes the search, as well as the date and time that the search was performed.

[0786] "User browsing history" refers to data that records the IDs of product pages and content that a user has viewed on a website, as well as the date and time of such viewing.

[0787] "Review history" refers to data that records the ratings and comments posted by users on products and services on a website, as well as the date and time of those posts.

[0788] A "database" is a digital information repository that centrally stores information such as a user's search history, browsing history, and review history, and manages it so that it can be searched and analyzed later.

[0789] "Features" are specific elements or attributes that indicate a user's interests and concerns, extracted from collected user information, and serve as indicators for inferring the user's preferences based on these.

[0790] "Means for predicting preferences" refers to algorithms or machine learning models that analyze collected features and predict what products or information a user will be interested in.

[0791] A "template" is a template for inserting personalized content based on individual user information from a generic text format.

[0792] A "personalized description" is a product description generated using a template and customized based on the user's preferences.

[0793] "Tokenization" is a preprocessing technique that breaks down text data into its smallest units, such as words and phrases.

[0794] "Normalization" is a preprocessing technique that arranges text data into a unified format based on certain rules.

[0795] The present invention relates to a system for generating personalized testimonials based on a user's search history, browsing history, and word-of-mouth history. Specific embodiments of the system are described below.

[0796] 1. Collection of User Information

[0797] When a user searches, browses, or reviews on the site, their history is collected. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[0798] For example, if a user searches for "smartphone," the search query "smartphone" and the search date and time are recorded. Next, when the user views a product page, the product ID and the date and time of the view are recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are also saved in the database.

[0799] 2. Guessing your favorite points

[0800] Features are extracted from the collected user information to infer the user's preferences. This process is performed by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[0801] As a concrete example, if a user searches for information about many "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[0802] 3. Generate personalized testimonials

[0803] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[0804] As a specific example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" is generated.

[0805] 4. Displaying personalized testimonials

[0806] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the terminal retrieves the personalized introduction from the server and displays it on the product page.

[0807] As a specific example, when a user accesses a product page while logged in, an introductory text appears at the top of the page: "This product provides a detailed explanation of the latest AI technology that interests you."

[0808] Example prompts to input to a generative AI model:

[0809] "Generate product descriptions that this user might be interested in based on their search, browsing, and review history."

[0810] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized introductions, thereby realizing product introductions that are attractive to users.

[0811] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0812] Step 1: The user enters keywords into the search form on the site and performs a search. The device records the entered search query and the date and time the search was performed, and sends this information to the server.

[0813] Input: The search query entered by the user and the date and time of the search.

[0814] Output: Search query and search date and time data.

[0815] Specific operation: A search keyword is entered and the content is sent to the server.

[0816] Step 2: The terminal records the ID of the product page viewed by the user and the date and time of the view, and sends this information to the server.

[0817] Input: The ID of the product page the user viewed and the date and time of the view.

[0818] Output: Product page ID and viewed date and time data.

[0819] Specific operation: When a user views a product page, the page ID and the date and time of the view are sent to the server.

[0820] Step 3: When a user posts a product review, the device records the content and posting date and time, and sends that information to the server.

[0821] Input: User-submitted review content and posting date and time.

[0822] Output: Review content and posting date and time data.

[0823] Specific operation: When a user posts a review, the content and date and time are sent to the server.

[0824] Step 4: The server stores the collected user search history, browsing history, and review history in a database.

[0825] Input: Search history, browsing history, review history.

[0826] Output: User history data stored in a database.

[0827] Specific behavior: The collected history is stored in a centralized database.

[0828] Step 5: The server extracts features from the stored data. During this process, the data is preprocessed by tokenizing and normalizing the text.

[0829] Input: Saved user history data.

[0830] Output: Extracted feature data.

[0831] Specific operation: Tokenize and normalize text data to extract features.

[0832] Step 6: The server uses a machine learning model based on the extracted features to infer the user's preferences.

[0833] Input: Feature data.

[0834] Output: Inferred user preference points.

[0835] Specific behavior: Analyzes data using machine learning models to infer user interests and trends.

[0836] Step 7: The server generates a personalized introduction using a template based on the inferred preferences, with specific variables filled in with user-specific information.

[0837] Input: Inferred preference points, template.

[0838] Output: A personalized introduction.

[0839] Specific operation: Generate an introduction based on a template that matches the user's preferences.

[0840] Step 8: When the user accesses the product page, the terminal receives the personalized introduction from the server and displays it on the product page.

[0841] Input: A personalized testimonial.

[0842] Output: The description displayed on the product page.

[0843] Specific operation: When a user accesses a product page, the corresponding description is displayed.

[0844] (Application example 1)

[0845] 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."

[0846] Many online shopping sites today provide users with a uniform introduction, which lacks information tailored to each user's individual preferences. This prevents them from effectively supporting users' purchasing decisions, resulting in insufficient sales promotion effectiveness. Furthermore, when users select products, the lack of personalized information that reflects their past behavioral history also contributes to a poor user experience.

[0847] 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.

[0848] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's preferences, means for generating a personalized introduction using a template and a generative AI model based on the inferred preferences, and means for displaying the personalized introduction when the user accesses a product page. This makes it possible to provide individually optimized product introductions that utilize the user's past behavioral history, thereby improving the user experience and further increasing the effectiveness of sales promotion.

[0849] "Search History" means a record of keyword searches performed by a User on an Online Platform.

[0850] "Browsing history" is a record of the information a user views on an online platform.

[0851] "Review history" means a record of reviews and comments posted by a user on an online platform.

[0852] "Database" means a management system for storing collected user information in a structured manner.

[0853] "Features" are indicators that indicate user behavior and interests extracted from collected data.

[0854] "Preference points" are specific areas of interest or fields of interest that are inferred from a user's past behavioral history.

[0855] A "template" is a framework of text that serves as the basis for creating a personalized introduction.

[0856] A "generative AI model" is an artificial intelligence model that uses user information as input to generate personalized sentences and answers.

[0857] A "testimonial" is a product description or recommendation written based on the user's interests and concerns.

[0858] "Product Page" means a page on the Online Platform that displays detailed information about a particular Product.

[0859] To implement this invention, the system requires a server, a user terminal, and an internet connection environment. The specific operation and configuration of the system will be described below.

[0860] First, the server collects the user's search history, browsing history, and review history and stores this information in a database. For example, this can be done by collecting data via API using the Python requests library. A script running on the server monitors user behavior in real time and stores it in the database accordingly.

[0861] The server then uses a machine learning model to infer user preferences from the collected data. This model is built using machine learning libraries such as scikit-learn and TensorFlow, and natural language processing libraries such as NLTK and spaCy may be used to preprocess the text data.

[0862] Based on the inferred preference points, the server generates a personalized introduction using a template and a generative AI model. The generative AI model can use OpenAI's GPT or other large-scale language models. The template contains generic sentences, and the AI ​​model embeds appropriate variables to generate optimized sentences.

[0863] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. The user's device displays this description on the product page, allowing the user to obtain information that suits their preferences. This process is also performed by a front-end script implemented in JavaScript or similar.

[0864] For example, if a user frequently searches for and browses "smartphones" and "technical books," the system generates and displays a description such as "A product that provides detailed explanations of the latest smartphone technology."

[0865] Prompt Sentence Examples

[0866] An example prompt is:

[0867] “If a user is interested in technical books or smartphone-related products, generate a description that reflects that interest:

[0868] Product name: "Explaining the latest smartphone technology"

[0869] Highlights: "This book provides an in-depth look at cutting-edge smartphone technology. A must-have for engineers and technologists."

[0870] Give reasons why users would love this product."

[0871] This invention makes it possible to introduce individually optimized products by utilizing the user's past behavioral history, which is expected to improve the user experience and increase the effectiveness of sales promotion.

[0872] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0873] Step 1:

[0874] The server collects the user's search history, browsing history, and review history. Specifically, it monitors search queries sent from the device, pages viewed, and review content, and obtains this data through an API. The input is user operation data, and the output is stored in a database in a structured data format (e.g., JSON).

[0875] Step 2:

[0876] The server extracts features from user information stored in a database. Specifically, it tokenizes and normalizes the collected text data using a natural language processing library (e.g., NLTK, spaCy). The input is structured user data, and the output is saved as features (e.g., frequently used words, categories).

[0877] Step 3:

[0878] The server uses a machine learning model to predict the user's preferences based on the extracted features. Specifically, the features are input to a pre-trained model (e.g., scikit-learn, TensorFlow) and the preference points are output as a prediction result. The input is the feature data, and the output is a list of preference points.

[0879] Step 4:

[0880] The server generates a personalized introduction using a template and a generative AI model based on the inferred preference points. Specifically, it sets variables in the template and inputs a prompt to the generative AI model (e.g., OpenAI's GPT) to generate an optimized introduction. The inputs are the preference points, the template, and the prompt from the AI ​​model, and the output is a personalized introduction.

[0881] Step 5:

[0882] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. Specifically, it monitors the user's current access status, identifies the relevant product page, and inserts the generated description. The input is the access history and the generated description, and the output is the updated product page.

[0883] Step 6:

[0884] The terminal displays the personalized description received from the server on the product page. Specifically, it uses a front-end script such as JavaScript to insert the description into an HTML element. The input is the description data from the server, and the output is the product page displayed to the user.

[0885] Through the above steps, individually optimized product introductions are realized by utilizing the user's past behavioral history.

[0886] 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.

[0887] The present invention provides a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0888] 1. Collection of User Information

[0889] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0890] 2. Collecting Emotional Data

[0891] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine. Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to the server and stores them in a database. This allows the user's emotional state to be recorded in real time.

[0892] For example, when a user searches for "smartphone" and then views the product page for "Smartphone A," the device detects the emotion of joy from the user's facial expression and records it. Also, when a user posts a review saying that "Smartphone A" is "very good," positive emotions are detected from the text and this is also recorded on the server.

[0893] 3. Preprocessing user information and emotion data

[0894] To preprocess the collected search history, browsing history, and review history, as well as sentiment data, we perform data cleaning, text tokenization, and normalization. This preprocessing ensures that the data is stored in a consistent format, making later analysis easier.

[0895] 4. Feature extraction and preference estimation

[0896] The server identifies frequently occurring keywords and phrases from the formatted data, and further identifies categories of particular interest to the user based on the product category information viewed. These features and emotional data are then input into a machine learning model to predict the user's preferences. The emotional data also takes into account the user's emotional response to specific items.

[0897] As a concrete example, if a user searches for information about a large number of technical books and displays a positive emotional response, the server infers that the user is interested in detailed technical information and has positive feelings about it.

[0898] 5. Generate personalized testimonials

[0899] Based on the inferred preferences and emotional data, the server generates a description using a template. The template contains generic sentences, and by substituting specific user information and emotional data for specific variables, a very specific and personalized description is generated.

[0900] For example, for users interested in technical books, an introduction such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[0901] 6. Displaying personalized testimonials

[0902] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0903] For example, when a user accesses a product page while logged in, the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0904] As described above, the system of the present invention can provide a personalized introduction optimized for each user based on both user information and emotional data, and effectively communicate the appeal of a product.

[0905] The processing flow will be explained below.

[0906] Step 1: Collect user information

[0907] Collection of search history

[0908] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[0909] Collection of browsing history

[0910] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[0911] Collecting review history

[0912] When a user posts a review, the device records the content and date and time of the post. This information is also sent to the server and stored in the database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are saved.

[0913] Step 2: Collecting emotion data

[0914] Facial expression analysis

[0915] While the user is browsing a product page, the device captures the user's facial expressions with a camera and analyzes their emotions using facial recognition technology. The analysis results are sent to a server and stored in a database.

[0916] Audio analysis

[0917] When a user searches for a product name or writes a review by voice, the device collects the voice data and analyzes the emotion using voice recognition technology. The analysis results are sent to a server and stored in a database.

[0918] Text Analysis

[0919] When a user posts a review in text, the device analyzes the text data and extracts emotions using an emotion engine. The analysis results are sent to a server and stored in a database.

[0920] Step 3: Preprocessing user information and sentiment data

[0921] Data Formatting

[0922] The server converts the collected search history, browsing history, review history, and sentiment data into a single data format, allowing the data to be stored in a consistent format.

[0923] Text normalization

[0924] The server tokenizes and normalizes the text data, especially the reviews and sentiment data, removing unnecessary whitespace and special characters. This preprocessing makes it easier for machine learning models to analyze later.

[0925] Step 4: Feature extraction and preference estimation

[0926] Identifying frequent keywords

[0927] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses the words "latest" and "technology," these frequently occurring keywords are extracted as features.

[0928] Identifying Categories

[0929] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[0930] Applying machine learning models

[0931] The server inputs the extracted features and emotion data into a machine learning model to predict the user's preferences. The emotion data also takes into account the user's emotional response to a particular item.

[0932] Recording points

[0933] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details and have positive feelings."

[0934] Step 5: Generate a personalized testimonial

[0935] Using templates

[0936] The server then inserts the estimated preferences and emotional data into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[0937] Customizable wording

[0938] The server adds specific words and phrases based on the user's preferences. For example, it might generate a description like, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field. Many users have been impressed by its quality and have given it high ratings."

[0939] Step 6: Displaying a personalized testimonial

[0940] User Identification

[0941] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[0942] Delivery of testimonials

[0943] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality" is displayed.

[0944] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[0945] Example 2

[0946] 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."

[0947] While conventional systems can collect users' search history, browsing history, and review history, they are unable to generate personalized product recommendations that reflect the user's emotional state in real time. This makes it difficult to effectively recommend products that take the user's emotions into account, making it difficult to increase the user's purchase intention.

[0948] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0949] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and saving them in a database, means for collecting emotional data during user operations and sending it to the server to save in the database, means for extracting features from the collected user information and emotional data and inferring the user's favorite points, means for generating a personalized testimonial using a template based on the inferred favorite points and emotional data, and means for displaying the personalized testimonial when the user accesses a product page. This makes it possible to generate and display an optimized testimonial that takes user emotions into consideration in real time.

[0950] "User" means an individual who accesses the system and performs operations such as searching, browsing, and submitting reviews.

[0951] "Search History" means a record of the search queries a user makes within the system, along with the date and time of the queries.

[0952] "Viewing history" is a record of product identification information and the date and time of viewing when a user views a product page within the system.

[0953] "Review history" is a record of the content of reviews posted by users about products within the system and the date and time of posting.

[0954] "Emotional data" refers to emotional information analyzed from the user's facial expressions and voice collected during user operation.

[0955] A "database" is an electronic data structure for storing and managing collected user information and emotional data.

[0956] "Features" are important attributes or information extracted from collected data that are useful for analysis and prediction.

[0957] "Preference points" are specific categories or attributes that users are particularly interested in and express positive feelings about.

[0958] "Template" means a formatted document used to generate a personalized testimonial.

[0959] A "personalized testimonial" is a personalized testimonial that is generated to reflect a user's specific characteristics or emotional state.

[0960] "Server" means a computer system that receives, processes, stores, and analyzes data sent from a user's device.

[0961] "Terminal" means a device used by a User to access and operate the System.

[0962] The present invention is a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[0963] Collection of User Information

[0964] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[0965] Example: If a user searches for "Smartphone A" and then views the product page for "Smartphone A," the device records the search query, search date and time, product ID, and view date and time, and sends them to the server. The server stores this in a database.

[0966] Collecting Emotional Data

[0967] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine (such as OpenFace or IBM Watson). Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to a server to store in a database. This allows the user's emotional state to be recorded in real time.

[0968] Example: While a user is browsing a product page on "Smartphone A," the device's webcam captures the user's facial expressions, which are then analyzed by the emotion engine. Emotions such as joy and surprise are detected, and the analysis results are sent to the server and stored in a database.

[0969] Preprocessing user information and sentiment data

[0970] The server cleans the search history, browsing history, review history, and sentiment data stored in the database, and tokenizes and normalizes the text data, ensuring that the data is stored in a consistent format for easier later analysis and processing.

[0971] Feature extraction and preference estimation

[0972] The server identifies frequently occurring keywords and phrases from the formatted data and identifies categories that the user is particularly interested in. These features and sentiment data are input into a machine learning model (e.g., Scikit-learn or TensorFlow) to predict the user's preferences.

[0973] Example: If a user searches for a lot of information about technical books and displays positive sentiment, the server infers that the user is interested in technical details and has positive sentiment about them.

[0974] Generate personalized testimonials

[0975] The server generates a testimonial using a template based on the inferred preferences and emotional data. The template contains generic text, and specific, personalized testimonials can be generated by substituting user-specific information and emotional data into specific variable parts.

[0976] Example: For users interested in technical books, a description such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[0977] Displaying a personalized testimonial

[0978] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[0979] Example: When a user logs in and accesses the product page "Technical Book A," the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[0980] As a result, the system of the present invention can provide an optimized personalized introduction based on both user information and emotional data, effectively conveying the appeal of the product.

[0981] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0982] Step 1:

[0983] A user takes an action on the site

[0984] Input: Users search, browse, and post reviews.

[0985] Actions: A user enters keywords into a search form, clicks the search button, browses a product page, and optionally submits a review. These actions take place within the site.

[0986] Output: User operation data (search query, viewed product ID, review content) is recorded.

[0987] Step 2:

[0988] The device records user operation data and sends it to the server.

[0989] Input: User operation data (search query, viewed product ID, review content).

[0990] How it works: The device records the search query and the date and time of the search when the user performs a search. Similarly, when the user views a product page, the device records the product ID and the date and time of the view. When the user posts a review, the device records the content of the review and the date and time of the post.

[0991] Output: The recorded user operation data is sent to the server and stored in a database.

[0992] Step 3:

[0993] The device collects the user's emotional data and sends it to the server.

[0994] Input: User's facial expression data, voice data.

[0995] How it works: The device's webcam and microphone are used to collect the user's facial expressions and voice in real time. The collected data is input into an emotion engine (OpenFace or IBM Watson) to analyze emotions. The analysis results are sent to a server.

[0996] Output: The analyzed emotion data is sent to the server and stored in a database.

[0997] Step 4:

[0998] The server preprocesses user information and emotion data

[0999] Input: Recorded user information (search history, browsing history, review history), analyzed emotional data.

[1000] How it works: The server cleans the raw data stored in the database, removing unnecessary information, then tokenizes and normalizes the text data.

[1001] Output: Preprocessed data in a consistent format.

[1002] Step 5:

[1003] The server extracts features and predicts the user's preferences.

[1004] Input: Preprocessed data.

[1005] How it works: The server extracts frequently occurring keywords and phrases from the data, extracts specific category information, and feeds this data into a machine learning model (Scikit-learn or TensorFlow) to predict user preferences.

[1006] Output: User preference points are inferred.

[1007] Step 6:

[1008] The server generates a personalized introduction

[1009] Input: Inferred preference points, sentiment data.

[1010] How it works: The server generates an introduction based on a template, inserting favorite points and emotional data.

[1011] Output: A personalized testimonial is generated.

[1012] Step 7:

[1013] The device displays a personalized introduction

[1014] Input: User ID, personalized introduction.

[1015] Operation: When a user accesses a product page, the device acquires the user ID and sends it to the server. The server then sends a personalized introduction corresponding to the user ID to the device. The device then displays the introduction on the product page.

[1016] Output: The user will see a personalized introduction.

[1017] (Application example 2)

[1018] 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."

[1019] Conventional online shopping sites recommend products based on a user's search history and browsing history, but do not take into account the user's emotional state. This makes it difficult to generate product reviews that reflect the user's emotions toward a particular product, and the effectiveness of these reviews in increasing the user's purchasing intent is not fully realized. Furthermore, there is a lack of systems that effectively utilize user emotional data, limiting the generation of personalized product reviews. Therefore, the present invention aims to solve these problems and provide a system that provides more effective personalized product reviews that reflect the user's emotional state.

[1020] 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 and storing a user's search history, browsing history, and review history in a database; means for extracting features from the collected user information and emotional data and inferring the user's preferences; and means for generating a personalized introduction using a template based on the inferred preferences and emotional state. This makes it possible to generate a personalized introduction that reflects the user's emotional state. Furthermore, when a user accesses a product page, an appropriate personalized introduction can be displayed, more effectively conveying the product's appeal.

[1021] "Search history" is a record of the search keywords and dates and times that a user searches on the Internet.

[1022] "Browsing history" refers to the IDs of web pages and products that a user has viewed on the Internet, as well as a record of the date and time of the views.

[1023] "Review history" refers to the ratings and reviews posted by users about specific products or services, as well as a record of their content.

[1024] "Emotional data" is information about a user's emotional state analyzed from their facial expressions, voice, and text.

[1025] A "feature" is a specific attribute or variable of the data input into a machine learning model, which represents a user's behavior or emotional state in numerical or categorical terms.

[1026] "Preference points" are elements or categories that a user is particularly interested in regarding a particular product or service, and are inferred from the user's behavioral and emotional data.

[1027] A "template" is a text template used to generate a personalized introduction, in which specific variable parts can be filled in with information for each user.

[1028] A "personalized testimonial" is a product or service testimonial that is individually optimized and generated based on the user's individual data (search history, browsing history, review history, emotional data, etc.).

[1029] A "machine learning model" is an algorithm or system that automatically learns patterns and relationships from data and makes predictions and classifications.

[1030] The present invention provides a system for collecting and analyzing a user's search history, browsing history, review history, and emotion data to generate personalized testimonials. Specific embodiments for implementing the present invention will be described below.

[1031] First, the system collects various data from users. When users search, browse, or post reviews on the site, the device records the search query, the viewed product ID and date and time, the review content, and the posting date and time, and sends this data to the server. This data is then stored in a database.

[1032] Next, emotional data is collected. The device analyzes emotions from facial expressions and voice while the user is using the site, and also extracts emotions from the text when posting reviews. This emotional data is also sent to the server and stored in a database.

[1033] The server pre-processes the stored search history, browsing history, review history, and sentiment data, including data cleaning, text tokenization, and normalization.

[1034] The server then uses a machine learning model to extract features from the preprocessed data and infer the user's preferences. In particular, it uses emotional data to consider the user's emotional reaction to a particular item. The machine learning model uses, for example, the Python library scikit-learn and the natural language processing library TextBlob.

[1035] Based on the inferred preferences and emotional data, the server uses a template to generate a personalized introductory text. This template contains generic text, and by substituting user-specific information and emotional data for specific variable parts, an introductory text optimized for each user is generated.

[1036] Finally, when a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies the appropriate personalized introduction based on the user ID and sends it to the device, allowing the user to view an introduction that reflects their preferences and emotions.

[1037] As a specific example, if a user searches for a technical book called "AI Technical Book" and then views product page "AI Technical Book B," the device will detect positive emotions from the user's facial expression data and record them. Furthermore, if the user posts a review stating "Very Good," positive emotions will also be detected from the text. Based on this data, a personalized introduction will be generated, such as, "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[1038] Example prompt sentence:

[1039] User Data:

[1040] Search history: ['AI technical books', 'machine learning']

[1041] View history: [{'item_id': '002', 'date': '2023-10-03'}, {'item_id': '003', 'date': '2023-10-04'}]

[1042] Review: [{'item_id': '002', 'review': 'Very Good', 'date': '2023-10-03'}]

[1043] Emotion data: [{'item_id': '002', 'emotion': 'joy', 'date': '2023-10-03'}]

[1044] Based on this data, generate a testimonial using the following template:

[1045] Template: "This product is specially designed for users interested in {}. {}\nMany users have been impressed with its quality."

[1046] Generated testimonial:

[1047] "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[1048] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1049] Step 1:

[1050] When a user searches, browses, or posts a review on the site, the device collects the search query, the viewed product ID, and the review content, and records the search date and time, the browse date and time, and the review posting date and time. This data is sent to the server and stored in a database. The input is the user's actions, and the output is the history data stored in the database.

[1051] Step 2:

[1052] The device analyzes the user's facial expressions and voice in real time and extracts emotional data using an emotion engine. When a review is posted, emotions are also analyzed from the text data, and this emotional data is sent to the server and stored in a database. The input is the user's facial expressions, voice, and text data, and the output is the extracted emotional data.

[1053] Step 3:

[1054] The server preprocesses the collected search history, browsing history, review history, and sentiment data. This preprocessing involves data cleaning, text tokenization, normalization, and formatting into a consistent format. The input is the raw data stored in the database, and the output is the preprocessed data.

[1055] Step 4:

[1056] The server extracts features from the preprocessed data and uses a machine learning model to infer the user's preference points. The libraries used are Python's scikit-learn and TextBlob, which also take into account the user's emotional response to specific items. The input is the preprocessed data, and the output is the inferred preference points.

[1057] Step 5:

[1058] The server generates a personalized introduction using a pre-prepared template based on the inferred preferences and emotion data. The template has variable sections that allow specific user information to be inserted. The input is the preferences and template, and the output is a personalized introduction.

[1059] Step 6:

[1060] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. The input is the user ID, and the output is the appropriate personalized introduction.

[1061] Step 7:

[1062] The terminal displays the personalized introduction received from the server on the product page. The input is the introduction received from the server, and the output is the introduction displayed on the product page. This allows the user to see an introduction that reflects their own preferences and emotions.

[1063] 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.

[1064] 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.

[1065] 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.

[1066] [Fourth embodiment]

[1067] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1068] 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.

[1069] 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).

[1070] 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.

[1071] 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.

[1072] 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).

[1073] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1074] 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.

[1075] 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.

[1076] 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.

[1077] 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.

[1078] 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.

[1079] 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."

[1080] The present invention provides a system for generating personalized testimonials based on a user's search history, browsing history, and review history, and the specific embodiments thereof will be described below.

[1081] 1. Collection of User Information

[1082] A special script is run to collect the user's search, browsing, and review history within the site. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[1083] For example, if a user searches for "smartphone," the search query "smartphone" and the date and time of the search are recorded. If the user then views a product page, the product ID and the date and time of the view are also recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are saved in the database.

[1084] 2. Guessing your favorite points

[1085] Features are extracted from the collected user information to infer the user's preferences. This process is performed using a machine learning model and is handled by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[1086] As a concrete example, if a user searches for information about a large number of "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[1087] 3. Generate personalized testimonials

[1088] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[1089] For example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" will be generated.

[1090] 4. Displaying personalized testimonials

[1091] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the device retrieves the personalized introduction from the server and displays it on the product page.

[1092] For example, when a user accesses a product page while logged in, a description will appear at the top of the page stating, "This product provides a detailed explanation of the latest AI technology that interests you."

[1093] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized product introductions, thereby realizing product introductions that are attractive to users.

[1094] The processing flow will be explained below.

[1095] Step 1: Collect user information

[1096] Collection of search history

[1097] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[1098] Collection of browsing history

[1099] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[1100] Collecting review history

[1101] When a user posts a review, the device records the content and date and time of the post. This information is sent to the server and stored in a database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are stored.

[1102] Step 2: Preprocessing user information

[1103] Data Formatting

[1104] The server formats the collected search history, browsing history, and review history into a single data format. This preprocessing ensures that the data is stored in a consistent format.

[1105] Text Preprocessing

[1106] The server tokenizes and normalizes the text data, especially the reviews, removing unnecessary whitespace and special characters, which makes it easier for machine learning models to analyze later.

[1107] Step 3: Feature extraction

[1108] Identifying frequent keywords

[1109] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses words like "latest" and "technology," these frequently occurring keywords are extracted as features.

[1110] Identifying Categories

[1111] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[1112] Step 4: Guess your preference points

[1113] Applying machine learning models

[1114] The server inputs the extracted features into a machine learning model to predict the user's preferences. Because the model learns from past data, it can predict the user's preferences with high accuracy.

[1115] Recording points

[1116] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details."

[1117] Step 5: Generate a personalized testimonial

[1118] Using templates

[1119] The server then inserts the inferred preferences into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[1120] Customizable wording

[1121] The server adds specific words and phrases based on the user's preferences, such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field."

[1122] Step 6: Displaying a personalized testimonial

[1123] User Identification

[1124] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[1125] Delivery of testimonials

[1126] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology in your area of ​​interest" is displayed.

[1127] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[1128] Example 1

[1129] 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."

[1130] In today's internet usage environment, it is becoming increasingly difficult for users to efficiently find information that is relevant to them from the vast amount of information available. In particular, when searching for or browsing products, it is difficult to find the product that best suits one's interests from the vast amount of related information available, making it difficult to make purchasing decisions. This reduces user convenience and increases the risk of lost opportunities for sellers.

[1131] 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.

[1132] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's points of interest, means for generating a personalized introductory text using a template based on the inferred points of interest, means for tokenizing and normalizing text as preprocessing of the collected data when inferring the user's points of interest, and means for inserting information for each user into specific variable parts of the introductory text generated using the template. This improves user convenience and increases purchasing motivation by providing information that is likely to interest the user.

[1133] "User search history" refers to data that records the keywords or queries that a user enters into a search form on a website and executes the search, as well as the date and time that the search was performed.

[1134] "User browsing history" refers to data that records the IDs of product pages and content that a user has viewed on a website, as well as the date and time of such viewing.

[1135] "Review history" refers to data that records the ratings and comments posted by users on products and services on a website, as well as the date and time of those posts.

[1136] A "database" is a digital information repository that centrally stores information such as a user's search history, browsing history, and review history, and manages it so that it can be searched and analyzed later.

[1137] "Features" are specific elements or attributes that indicate a user's interests and concerns, extracted from collected user information, and serve as indicators for inferring the user's preferences based on these.

[1138] "Means for predicting preferences" refers to algorithms or machine learning models that analyze collected features and predict what products or information a user will be interested in.

[1139] A "template" is a template for inserting personalized content based on individual user information from a generic text format.

[1140] A "personalized description" is a product description generated using a template and customized based on the user's preferences.

[1141] "Tokenization" is a preprocessing technique that breaks down text data into its smallest units, such as words and phrases.

[1142] "Normalization" is a preprocessing technique that arranges text data into a unified format based on certain rules.

[1143] The present invention relates to a system for generating personalized testimonials based on a user's search history, browsing history, and word-of-mouth history. Specific embodiments of the system are described below.

[1144] 1. Collection of User Information

[1145] When a user searches, browses, or reviews on the site, their history is collected. When a user enters keywords into the search form and performs a search, the device records the search query and the date and time of the search. This information is sent to the server and stored in a database.

[1146] For example, if a user searches for "smartphone," the search query "smartphone" and the search date and time are recorded. Next, when the user views a product page, the product ID and the date and time of the view are recorded. Furthermore, if the user posts a review about that product, the content and the date and time of posting are also saved in the database.

[1147] 2. Guessing your favorite points

[1148] Features are extracted from the collected user information to infer the user's preferences. This process is performed by the server. The collected data is preprocessed, for example, by tokenizing and normalizing the text. Next, frequently occurring keywords and specific categories are extracted as features.

[1149] As a concrete example, if a user searches for information about many "technical books" and posts related reviews, the server infers that the user is interested in detailed technical information.

[1150] 3. Generate personalized testimonials

[1151] Based on the inferred preferences, a description is generated using a template. This process is also handled by the server. The template contains generic text, and specific user information is inserted into specific variable parts.

[1152] As a specific example, for users interested in technical books, an introduction such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields" is generated.

[1153] 4. Displaying personalized testimonials

[1154] When a user accesses a specific product page, a personalized introduction is displayed. When a user accesses a page, the terminal retrieves the personalized introduction from the server and displays it on the product page.

[1155] As a specific example, when a user accesses a product page while logged in, an introductory text appears at the top of the page: "This product provides a detailed explanation of the latest AI technology that interests you."

[1156] Example prompts to input to a generative AI model:

[1157] "Generate product descriptions that this user might be interested in based on their search, browsing, and review history."

[1158] As described above, the system of the present invention performs a series of processes from collecting user information to generating and displaying personalized introductions, thereby realizing product introductions that are attractive to users.

[1159] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1160] Step 1: The user enters keywords into the search form on the site and performs a search. The device records the entered search query and the date and time the search was performed, and sends this information to the server.

[1161] Input: The search query entered by the user and the date and time of the search.

[1162] Output: Search query and search date and time data.

[1163] Specific operation: A search keyword is entered and the content is sent to the server.

[1164] Step 2: The terminal records the ID of the product page viewed by the user and the date and time of the view, and sends this information to the server.

[1165] Input: The ID of the product page the user viewed and the date and time of the view.

[1166] Output: Product page ID and viewed date and time data.

[1167] Specific operation: When a user views a product page, the page ID and the date and time of the view are sent to the server.

[1168] Step 3: When a user posts a product review, the device records the content and posting date and time, and sends that information to the server.

[1169] Input: User-submitted review content and posting date and time.

[1170] Output: Review content and posting date and time data.

[1171] Specific operation: When a user posts a review, the content and date and time are sent to the server.

[1172] Step 4: The server stores the collected user search history, browsing history, and review history in a database.

[1173] Input: Search history, browsing history, review history.

[1174] Output: User history data stored in a database.

[1175] Specific behavior: The collected history is stored in a centralized database.

[1176] Step 5: The server extracts features from the stored data. During this process, the data is preprocessed by tokenizing and normalizing the text.

[1177] Input: Saved user history data.

[1178] Output: Extracted feature data.

[1179] Specific operation: Tokenize and normalize text data to extract features.

[1180] Step 6: The server uses a machine learning model based on the extracted features to infer the user's preferences.

[1181] Input: Feature data.

[1182] Output: Inferred user preference points.

[1183] Specific behavior: Analyzes data using machine learning models to infer user interests and trends.

[1184] Step 7: The server generates a personalized introduction using a template based on the inferred preferences, with specific variables filled in with user-specific information.

[1185] Input: Inferred preference points, template.

[1186] Output: A personalized introduction.

[1187] Specific operation: Generate an introduction based on a template that matches the user's preferences.

[1188] Step 8: When the user accesses the product page, the terminal receives the personalized introduction from the server and displays it on the product page.

[1189] Input: A personalized testimonial.

[1190] Output: The description displayed on the product page.

[1191] Specific operation: When a user accesses a product page, the corresponding description is displayed.

[1192] (Application example 1)

[1193] 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."

[1194] Many online shopping sites today provide users with a uniform introduction, which lacks information tailored to each user's individual preferences. This prevents them from effectively supporting users' purchasing decisions, resulting in insufficient sales promotion effectiveness. Furthermore, when users select products, the lack of personalized information that reflects their past behavioral history also contributes to a poor user experience.

[1195] 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.

[1196] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and storing them in a database, means for extracting features from the collected user information and inferring the user's preferences, means for generating a personalized introduction using a template and a generative AI model based on the inferred preferences, and means for displaying the personalized introduction when the user accesses a product page. This makes it possible to provide individually optimized product introductions that utilize the user's past behavioral history, thereby improving the user experience and further increasing the effectiveness of sales promotion.

[1197] "Search History" means a record of keyword searches performed by a User on an Online Platform.

[1198] "Browsing history" is a record of the information a user views on an online platform.

[1199] "Review history" means a record of reviews and comments posted by a user on an online platform.

[1200] "Database" means a management system for storing collected user information in a structured manner.

[1201] "Features" are indicators that indicate user behavior and interests extracted from collected data.

[1202] "Preference points" are specific areas of interest or fields of interest that are inferred from a user's past behavioral history.

[1203] A "template" is a framework of text that serves as the basis for creating a personalized introduction.

[1204] A "generative AI model" is an artificial intelligence model that uses user information as input to generate personalized sentences and answers.

[1205] A "testimonial" is a product description or recommendation written based on the user's interests and concerns.

[1206] "Product Page" means a page on the Online Platform that displays detailed information about a particular Product.

[1207] To implement this invention, the system requires a server, a user terminal, and an internet connection environment. The specific operation and configuration of the system will be described below.

[1208] First, the server collects the user's search history, browsing history, and review history and stores this information in a database. For example, this can be done by collecting data via API using the Python requests library. A script running on the server monitors user behavior in real time and stores it in the database accordingly.

[1209] The server then uses a machine learning model to infer user preferences from the collected data. This model is built using machine learning libraries such as scikit-learn and TensorFlow, and natural language processing libraries such as NLTK and spaCy may be used to preprocess the text data.

[1210] Based on the inferred preference points, the server generates a personalized introduction using a template and a generative AI model. The generative AI model can use OpenAI's GPT or other large-scale language models. The template contains generic sentences, and the AI ​​model embeds appropriate variables to generate optimized sentences.

[1211] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. The user's device displays this description on the product page, allowing the user to obtain information that suits their preferences. This process is also performed by a front-end script implemented in JavaScript or similar.

[1212] For example, if a user frequently searches for and browses "smartphones" and "technical books," the system generates and displays a description such as "A product that provides detailed explanations of the latest smartphone technology."

[1213] Prompt Sentence Examples

[1214] An example prompt is:

[1215] “If a user is interested in technical books or smartphone-related products, generate a description that reflects that interest:

[1216] Product name: "Explaining the latest smartphone technology"

[1217] Highlights: "This book provides an in-depth look at cutting-edge smartphone technology. A must-have for engineers and technologists."

[1218] Give reasons why users would love this product."

[1219] This invention makes it possible to introduce individually optimized products by utilizing the user's past behavioral history, which is expected to improve the user experience and increase the effectiveness of sales promotion.

[1220] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1221] Step 1:

[1222] The server collects the user's search history, browsing history, and review history. Specifically, it monitors search queries sent from the device, pages viewed, and review content, and obtains this data through an API. The input is user operation data, and the output is stored in a database in a structured data format (e.g., JSON).

[1223] Step 2:

[1224] The server extracts features from user information stored in a database. Specifically, it tokenizes and normalizes the collected text data using a natural language processing library (e.g., NLTK, spaCy). The input is structured user data, and the output is saved as features (e.g., frequently used words, categories).

[1225] Step 3:

[1226] The server uses a machine learning model to predict the user's preferences based on the extracted features. Specifically, the features are input to a pre-trained model (e.g., scikit-learn, TensorFlow) and the preference points are output as a prediction result. The input is the feature data, and the output is a list of preference points.

[1227] Step 4:

[1228] The server generates a personalized introduction using a template and a generative AI model based on the inferred preference points. Specifically, it sets variables in the template and inputs a prompt to the generative AI model (e.g., OpenAI's GPT) to generate an optimized introduction. The inputs are the preference points, the template, and the prompt from the AI ​​model, and the output is a personalized introduction.

[1229] Step 5:

[1230] When a user accesses a product page, the server generates a personalized description in real time and sends it to the user's device. Specifically, it monitors the user's current access status, identifies the relevant product page, and inserts the generated description. The input is the access history and the generated description, and the output is the updated product page.

[1231] Step 6:

[1232] The terminal displays the personalized description received from the server on the product page. Specifically, it uses a front-end script such as JavaScript to insert the description into an HTML element. The input is the description data from the server, and the output is the product page displayed to the user.

[1233] Through the above steps, individually optimized product introductions are realized by utilizing the user's past behavioral history.

[1234] 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.

[1235] The present invention provides a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[1236] 1. Collection of User Information

[1237] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[1238] 2. Collecting Emotional Data

[1239] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine. Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to the server and stores them in a database. This allows the user's emotional state to be recorded in real time.

[1240] For example, when a user searches for "smartphone" and then views the product page for "Smartphone A," the device detects the emotion of joy from the user's facial expression and records it. Also, when a user posts a review saying that "Smartphone A" is "very good," positive emotions are detected from the text and this is also recorded on the server.

[1241] 3. Preprocessing user information and emotion data

[1242] To preprocess the collected search history, browsing history, and review history, as well as sentiment data, we perform data cleaning, text tokenization, and normalization. This preprocessing ensures that the data is stored in a consistent format, making later analysis easier.

[1243] 4. Feature extraction and preference estimation

[1244] The server identifies frequently occurring keywords and phrases from the formatted data, and further identifies categories of particular interest to the user based on the product category information viewed. These features and emotional data are then input into a machine learning model to predict the user's preferences. The emotional data also takes into account the user's emotional response to specific items.

[1245] As a concrete example, if a user searches for information about a large number of technical books and displays a positive emotional response, the server infers that the user is interested in detailed technical information and has positive feelings about it.

[1246] 5. Generate personalized testimonials

[1247] Based on the inferred preferences and emotional data, the server generates a description using a template. The template contains generic sentences, and by substituting specific user information and emotional data for specific variables, a very specific and personalized description is generated.

[1248] For example, for users interested in technical books, an introduction such as, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[1249] 6. Displaying personalized testimonials

[1250] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[1251] For example, when a user accesses a product page while logged in, the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[1252] As described above, the system of the present invention can provide a personalized introduction optimized for each user based on both user information and emotional data, and effectively communicate the appeal of a product.

[1253] The processing flow will be explained below.

[1254] Step 1: Collect user information

[1255] Collection of search history

[1256] When a user searches on a website, the device records the search query and the date and time of the search. The recorded information is sent to the server and stored in a database. For example, if a user searches for "smartphone," the query and date and time are stored.

[1257] Collection of browsing history

[1258] When a user views a product page, the device records the product ID and the date and time of the view. This is also sent to the server and stored in the database. For example, if a user views a page for "Smartphone A," the product ID and date and time are recorded.

[1259] Collecting review history

[1260] When a user posts a review, the device records the content and date and time of the post. This information is also sent to the server and stored in the database. For example, if a user rates "Smartphone A" as "Very Good," the content and date and time of the review are saved.

[1261] Step 2: Collecting emotion data

[1262] Facial expression analysis

[1263] While the user is browsing a product page, the device captures the user's facial expressions with a camera and analyzes their emotions using facial recognition technology. The analysis results are sent to a server and stored in a database.

[1264] Audio analysis

[1265] When a user searches for a product name or writes a review by voice, the device collects the voice data and analyzes the emotion using voice recognition technology. The analysis results are sent to a server and stored in a database.

[1266] Text Analysis

[1267] When a user posts a review in text, the device analyzes the text data and extracts emotions using an emotion engine. The analysis results are sent to a server and stored in a database.

[1268] Step 3: Preprocessing user information and sentiment data

[1269] Data Formatting

[1270] The server converts the collected search history, browsing history, review history, and sentiment data into a single data format, allowing the data to be stored in a consistent format.

[1271] Text normalization

[1272] The server tokenizes and normalizes the text data, especially the reviews and sentiment data, removing unnecessary whitespace and special characters. This preprocessing makes it easier for machine learning models to analyze later.

[1273] Step 4: Feature extraction and preference estimation

[1274] Identifying frequent keywords

[1275] The server identifies frequently occurring keywords and phrases from the formatted data. For example, if a user frequently uses the words "latest" and "technology," these frequently occurring keywords are extracted as features.

[1276] Identifying Categories

[1277] The server identifies categories that the user is particularly interested in based on the category information of the viewed products. For example, categories such as technical books and electronic devices are extracted.

[1278] Applying machine learning models

[1279] The server inputs the extracted features and emotion data into a machine learning model to predict the user's preferences. The emotion data also takes into account the user's emotional response to a particular item.

[1280] Recording points

[1281] The server records the inferred preference points for each user and stores them in a database, such as "I'm interested in technical details and have positive feelings."

[1282] Step 5: Generate a personalized testimonial

[1283] Using templates

[1284] The server then inserts the estimated preferences and emotional data into a prepared introductory text template, and generates a personalized introductory text by filling in specific content into the variable parts of the template.

[1285] Customizable wording

[1286] The server adds specific words and phrases based on the user's preferences. For example, it might generate a description like, "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in your field. Many users have been impressed by its quality and have given it high ratings."

[1287] Step 6: Displaying a personalized testimonial

[1288] User Identification

[1289] When a user accesses a product page, the device acquires the user ID and sends it to the server, which then identifies the appropriate personalized introduction based on the user ID.

[1290] Delivery of testimonials

[1291] The server then sends the identified personalized introduction to the terminal and displays it on the product page the user is viewing. For example, when a user accesses a product page, an introduction such as "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality" is displayed.

[1292] Through the above processing steps, the system can provide personalized introductions optimized for individual users.

[1293] Example 2

[1294] 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."

[1295] While conventional systems can collect users' search history, browsing history, and review history, they are unable to generate personalized product recommendations that reflect the user's emotional state in real time. This makes it difficult to effectively recommend products that take the user's emotions into account, making it difficult to increase the user's purchase intention.

[1296] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1297] In this invention, the server includes means for collecting a user's search history, browsing history, and review history and saving them in a database, means for collecting emotional data during user operations and sending it to the server to save in the database, means for extracting features from the collected user information and emotional data and inferring the user's favorite points, means for generating a personalized testimonial using a template based on the inferred favorite points and emotional data, and means for displaying the personalized testimonial when the user accesses a product page. This makes it possible to generate and display an optimized testimonial that takes user emotions into consideration in real time.

[1298] "User" means an individual who accesses the system and performs operations such as searching, browsing, and submitting reviews.

[1299] "Search History" means a record of the search queries a user makes within the system, along with the date and time of the queries.

[1300] "Viewing history" is a record of product identification information and the date and time of viewing when a user views a product page within the system.

[1301] "Review history" is a record of the content of reviews posted by users about products within the system and the date and time of posting.

[1302] "Emotional data" refers to emotional information analyzed from the user's facial expressions and voice collected during user operation.

[1303] A "database" is an electronic data structure for storing and managing collected user information and emotional data.

[1304] "Features" are important attributes or information extracted from collected data that are useful for analysis and prediction.

[1305] "Preference points" are specific categories or attributes that users are particularly interested in and express positive feelings about.

[1306] "Template" means a formatted document used to generate a personalized testimonial.

[1307] A "personalized testimonial" is a personalized testimonial that is generated to reflect a user's specific characteristics or emotional state.

[1308] "Server" means a computer system that receives, processes, stores, and analyzes data sent from a user's device.

[1309] "Terminal" means a device used by a User to access and operate the System.

[1310] The present invention is a system that collects users' search history, browsing history, and review history, as well as their emotional data, and generates personalized testimonials based on this information. This system generates testimonials that reflect the user's emotional state, thereby increasing their purchase intentions. Specific embodiments of the system are described below.

[1311] Collection of User Information

[1312] Special scripts are executed when users search, browse, or post reviews on the site. When a user enters keywords into the search form and performs a search, the device records the search query and the search date and time, sends them to the server, and stores them in a database. Similarly, when a user views a product page, the product ID and the date and time of the view are recorded, and when a user posts a review, the content of the review and the date and time of posting are recorded, sent to the server, and stored in a database.

[1313] Example: If a user searches for "Smartphone A" and then views the product page for "Smartphone A," the device records the search query, search date and time, product ID, and view date and time, and sends them to the server. The server stores this in a database.

[1314] Collecting Emotional Data

[1315] While the user is using the site, the device analyzes the user's facial expressions and voice using an emotion engine (such as OpenFace or IBM Watson). Furthermore, when the user posts a review, the device analyzes the emotion from the text data and sends the results to a server to store in a database. This allows the user's emotional state to be recorded in real time.

[1316] Example: While a user is browsing a product page on "Smartphone A," the device's webcam captures the user's facial expressions, which are then analyzed by the emotion engine. Emotions such as joy and surprise are detected, and the analysis results are sent to the server and stored in a database.

[1317] Preprocessing user information and sentiment data

[1318] The server cleans the search history, browsing history, review history, and sentiment data stored in the database, and tokenizes and normalizes the text data, ensuring that the data is stored in a consistent format for easier later analysis and processing.

[1319] Feature extraction and preference estimation

[1320] The server identifies frequently occurring keywords and phrases from the formatted data and identifies categories that the user is particularly interested in. These features and sentiment data are input into a machine learning model (e.g., Scikit-learn or TensorFlow) to predict the user's preferences.

[1321] Example: If a user searches for a lot of information about technical books and displays positive sentiment, the server infers that the user is interested in technical details and has positive sentiment about them.

[1322] Generate personalized testimonials

[1323] The server generates a testimonial using a template based on the inferred preferences and emotional data. The template contains generic text, and specific, personalized testimonials can be generated by substituting user-specific information and emotional data into specific variable parts.

[1324] Example: For users interested in technical books, a description such as "This book provides a detailed explanation of the latest AI technology and is ideal for improving your skills in specialized fields. Many users have been impressed by its quality and have given it high ratings" may be generated.

[1325] Displaying a personalized testimonial

[1326] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. When the introduction is displayed on the product page the user views, the user can see an introduction that reflects their preferences and emotions.

[1327] Example: When a user logs in and accesses the product page "Technical Book A," the following description is displayed: "This product provides a detailed explanation of the latest AI technology, which is an area of ​​interest to you, and many users have been impressed by its quality."

[1328] As a result, the system of the present invention can provide an optimized personalized introduction based on both user information and emotional data, effectively conveying the appeal of the product.

[1329] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1330] Step 1:

[1331] A user takes an action on the site

[1332] Input: Users search, browse, and post reviews.

[1333] Actions: A user enters keywords into a search form, clicks the search button, browses a product page, and optionally submits a review. These actions take place within the site.

[1334] Output: User operation data (search query, viewed product ID, review content) is recorded.

[1335] Step 2:

[1336] The device records user operation data and sends it to the server.

[1337] Input: User operation data (search query, viewed product ID, review content).

[1338] How it works: The device records the search query and the date and time of the search when the user performs a search. Similarly, when the user views a product page, the device records the product ID and the date and time of the view. When the user posts a review, the device records the content of the review and the date and time of the post.

[1339] Output: The recorded user operation data is sent to the server and stored in a database.

[1340] Step 3:

[1341] The device collects the user's emotional data and sends it to the server.

[1342] Input: User's facial expression data, voice data.

[1343] How it works: The device's webcam and microphone are used to collect the user's facial expressions and voice in real time. The collected data is input into an emotion engine (OpenFace or IBM Watson) to analyze emotions. The analysis results are sent to a server.

[1344] Output: The analyzed emotion data is sent to the server and stored in a database.

[1345] Step 4:

[1346] The server preprocesses user information and emotion data

[1347] Input: Recorded user information (search history, browsing history, review history), analyzed sentiment data.

[1348] How it works: The server cleans the raw data stored in the database, removing unnecessary information, then tokenizes and normalizes the text data.

[1349] Output: Preprocessed data in a consistent format.

[1350] Step 5:

[1351] The server extracts features and predicts the user's preferences.

[1352] Input: Preprocessed data.

[1353] How it works: The server extracts frequently occurring keywords and phrases from the data, extracts specific category information, and feeds this data into a machine learning model (Scikit-learn or TensorFlow) to predict user preferences.

[1354] Output: User preference points are inferred.

[1355] Step 6:

[1356] The server generates a personalized introduction

[1357] Input: Inferred preference points, sentiment data.

[1358] How it works: The server generates an introduction based on a template, inserting favorite points and emotional data.

[1359] Output: A personalized testimonial is generated.

[1360] Step 7:

[1361] The device displays a personalized introduction

[1362] Input: User ID, personalized introduction.

[1363] Operation: When a user accesses a product page, the device acquires the user ID and sends it to the server. The server then sends a personalized introduction corresponding to the user ID to the device. The device then displays the introduction on the product page.

[1364] Output: The user will see a personalized introduction.

[1365] (Application example 2)

[1366] 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."

[1367] Conventional online shopping sites recommend products based on a user's search history and browsing history, but do not take into account the user's emotional state. This makes it difficult to generate product reviews that reflect the user's emotions toward a particular product, and the effectiveness of these reviews in increasing the user's purchasing intent is not fully realized. Furthermore, there is a lack of systems that effectively utilize user emotional data, limiting the generation of personalized product reviews. Therefore, the present invention aims to solve these problems and provide a system that provides more effective personalized product reviews that reflect the user's emotional state.

[1368] 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 and storing a user's search history, browsing history, and review history in a database; means for extracting features from the collected user information and emotional data and inferring the user's preferences; and means for generating a personalized introduction using a template based on the inferred preferences and emotional state. This makes it possible to generate a personalized introduction that reflects the user's emotional state. Furthermore, when a user accesses a product page, an appropriate personalized introduction can be displayed, more effectively conveying the product's appeal.

[1369] "Search history" is a record of the search keywords and dates and times that a user searches on the Internet.

[1370] "Browsing history" refers to the IDs of web pages and products that a user has viewed on the Internet, as well as a record of the date and time of the views.

[1371] "Review history" refers to the ratings and reviews posted by users about specific products or services, as well as a record of their content.

[1372] "Emotional data" is information about a user's emotional state analyzed from their facial expressions, voice, and text.

[1373] A "feature" is a specific attribute or variable of the data input into a machine learning model, which represents a user's behavior or emotional state in numerical or categorical terms.

[1374] "Preference points" are elements or categories that a user is particularly interested in regarding a particular product or service, and are inferred from the user's behavioral and emotional data.

[1375] A "template" is a text template used to generate a personalized introduction, in which specific variable parts can be filled in with information for each user.

[1376] A "personalized testimonial" is a product or service testimonial that is individually optimized and generated based on the user's individual data (search history, browsing history, review history, emotional data, etc.).

[1377] A "machine learning model" is an algorithm or system that automatically learns patterns and relationships from data and makes predictions and classifications.

[1378] The present invention provides a system for collecting and analyzing a user's search history, browsing history, review history, and emotion data to generate personalized testimonials. Specific embodiments for implementing the present invention will be described below.

[1379] First, the system collects various data from users. When users search, browse, or post reviews on the site, the device records the search query, the viewed product ID and date and time, the review content, and the posting date and time, and sends this data to the server. This data is then stored in a database.

[1380] Next, emotional data is collected. The device analyzes emotions from facial expressions and voice while the user is using the site, and also extracts emotions from the text when posting reviews. This emotional data is also sent to the server and stored in a database.

[1381] The server pre-processes the stored search history, browsing history, review history, and sentiment data, including data cleaning, text tokenization, and normalization.

[1382] The server then uses a machine learning model to extract features from the preprocessed data and infer the user's preferences. In particular, it uses emotional data to consider the user's emotional reaction to a particular item. The machine learning model uses, for example, the Python library scikit-learn and the natural language processing library TextBlob.

[1383] Based on the inferred preferences and emotional data, the server uses a template to generate a personalized introductory text. This template contains generic text, and by substituting user-specific information and emotional data for specific variable parts, an introductory text optimized for each user is generated.

[1384] Finally, when a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies the appropriate personalized introduction based on the user ID and sends it to the device, allowing the user to view an introduction that reflects their preferences and emotions.

[1385] As a specific example, if a user searches for a technical book called "AI Technical Book" and then views product page "AI Technical Book B," the device will detect positive emotions from the user's facial expression data and record them. Furthermore, if the user posts a review stating "Very Good," positive emotions will also be detected from the text. Based on this data, a personalized introduction will be generated, such as, "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[1386] Example prompt sentence:

[1387] User Data:

[1388] Search history: ['AI technical books', 'machine learning']

[1389] View history: [{'item_id': '002', 'date': '2023-10-03'}, {'item_id': '003', 'date': '2023-10-04'}]

[1390] Review: [{'item_id': '002', 'review': 'Very Good', 'date': '2023-10-03'}]

[1391] Emotion data: [{'item_id': '002', 'emotion': 'joy', 'date': '2023-10-03'}]

[1392] Based on this data, generate a testimonial using the following template:

[1393] Template: "This product is specially designed for users interested in {}. {}\nMany users have been impressed with its quality."

[1394] Generated testimonial:

[1395] "This product is specially designed for users interested in AI technical books and machine learning. It has received positive reviews. Many users are impressed with its quality."

[1396] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1397] Step 1:

[1398] When a user searches, browses, or posts a review on the site, the device collects the search query, the viewed product ID, and the review content, and records the search date and time, the browse date and time, and the review posting date and time. This data is sent to the server and stored in a database. The input is the user's actions, and the output is the history data stored in the database.

[1399] Step 2:

[1400] The device analyzes the user's facial expressions and voice in real time and extracts emotional data using an emotion engine. When a review is posted, emotions are also analyzed from the text data, and this emotional data is sent to the server and stored in a database. The input is the user's facial expressions, voice, and text data, and the output is the extracted emotional data.

[1401] Step 3:

[1402] The server preprocesses the collected search history, browsing history, review history, and sentiment data. This preprocessing involves data cleaning, text tokenization, normalization, and formatting into a consistent format. The input is the raw data stored in the database, and the output is the preprocessed data.

[1403] Step 4:

[1404] The server extracts features from the preprocessed data and uses a machine learning model to infer the user's preference points. The libraries used are Python's scikit-learn and TextBlob, which also take into account the user's emotional response to specific items. The input is the preprocessed data, and the output is the inferred preference points.

[1405] Step 5:

[1406] The server generates a personalized introduction using a pre-prepared template based on the inferred preferences and emotion data. The template has variable sections that allow specific user information to be inserted. The input is the preferences and template, and the output is a personalized introduction.

[1407] Step 6:

[1408] When a user accesses a product page, the device acquires the user ID and sends it to the server. The server identifies an appropriate personalized introduction based on this user ID and sends it to the device. The input is the user ID, and the output is the appropriate personalized introduction.

[1409] Step 7:

[1410] The terminal displays the personalized introduction received from the server on the product page. The input is the introduction received from the server, and the output is the introduction displayed on the product page. This allows the user to see an introduction that reflects their own preferences and emotions.

[1411] 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.

[1412] 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.

[1413] 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.

[1414] 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.

[1415] 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.

[1416] 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.

[1417] 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).

[1418] 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.

[1419] 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."

[1420] 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.

[1421] 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).

[1422] 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.

[1423] 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.

[1424] 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.

[1425] 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.

[1426] 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.

[1427] 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.

[1428] 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.

[1429] 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.

[1430] 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.

[1431] 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.

[1432] The following is further disclosed regarding the above embodiment.

[1433] (Claim 1)

[1434] A means for collecting and storing users' search history, browsing history, and review history in a database;

[1435] A means of extracting features from collected user information and inferring user preferences;

[1436] a means for generating a personalized introduction using a template based on the inferred preference points;

[1437] A way to display a personalized introduction when a user visits a product page,

[1438] A system including:

[1439] (Claim 2)

[1440] 2. The system of claim 1, wherein the search history collection means records the search query and date and time when the user performs a search.

[1441] (Claim 3)

[1442] The system according to claim 1, wherein the means for predicting preference points analyzes the features using a machine learning model.

[1443] "Example 1"

[1444] (Claim 1)

[1445] A means for collecting and storing users' search history, browsing history, and review history in a database;

[1446] A means of extracting features from collected user information and inferring user preferences;

[1447] a means for generating a personalized introduction using a template based on the inferred preference points;

[1448] A way to display a personalized introduction when a user visits a product page,

[1449] Preprocessing of collected data in inferring user preference points involves tokenizing and normalizing text;

[1450] A means to insert user-specific information into specific variable parts of the introduction text generated using a template;

[1451] A system including:

[1452] (Claim 2)

[1453] 2. The system of claim 1, wherein the search history collection means records the search query and date and time when the user performs a search.

[1454] (Claim 3)

[1455] The system according to claim 1, wherein the means for predicting preference points analyzes the features using a machine learning model.

[1456] "Application Example 1"

[1457] (Claim 1)

[1458] A means for collecting and storing users' search history, browsing history, and review history in a database;

[1459] A means of extracting features from collected user information and inferring user preferences;

[1460] A means for generating a personalized introduction based on the inferred preferences using templates and a generative AI model;

[1461] A way to display a personalized introduction when a user visits a product page,

[1462] A system including:

[1463] (Claim 2)

[1464] 2. The system of claim 1, wherein the search history collection means records the search query and date and time when the user performs a search.

[1465] (Claim 3)

[1466] The system according to claim 1, wherein the means for predicting preference points analyzes the features using a machine learning model and generates a prompt sentence.

[1467] "Example 2: Combining Emotion Engines"

[1468] (Claim 1)

[1469] A means for collecting and storing users' search history, browsing history, and review history in a database;

[1470] A means for collecting emotional data during user operations, transmitting it to a server, and storing it in a database;

[1471] A means for extracting features from collected user information and emotion data and inferring user preferences;

[1472] a means for generating a personalized introduction using a template based on the inferred preference points and sentiment data;

[1473] A way to display a personalized introduction when a user visits a product page,

[1474] A system including:

[1475] (Claim 2)

[1476] 2. The system of claim 1, wherein the search history collection means records the search query and date and time when the user performs a search.

[1477] (Claim 3)

[1478] The system according to claim 1, wherein the means for predicting preference points analyzes the features using a machine learning model.

[1479] "Application example 2 when combining emotion engines"

[1480] (Claim 1)

[1481] A means for collecting and storing users' search history, browsing history, and review history in a database;

[1482] A means for extracting features from collected user information and emotion data and inferring user preferences;

[1483] means for generating a personalized introduction using a template based on the inferred preference points and emotional state;

[1484] A way to display a personalized introduction when a user visits a product page,

[1485] A system including:

[1486] (Claim 2)

[1487] 2. The system of claim 1, wherein the search history collection means records the search query and date and time when the user performs a search.

[1488] (Claim 3)

[1489] 2. The system of claim 1, wherein the means for predicting preference points analyzes the feature and emotion data using a machine learning model. [Explanation of symbols]

[1490] 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 for collecting and storing users' search history, browsing history, and review history in a database; A means of extracting features from collected user information and inferring user preferences; a means for generating a personalized introduction using a template based on the inferred preference points; A way to display a personalized introduction when a user visits a product page, A system including:

2. 2. The system of claim 1, wherein the search history collecting means records search queries and dates and times when users conduct searches.

3. The system according to claim 1 , wherein the means for predicting the preference points analyzes the feature amount using a machine learning model.

Images