system

Abstract

Provide a system. A means for collecting a user's purchase history; A means of collecting user interest genres, a means of collecting answers to user questions; a means of integrating the collected data; a means for preprocessing the integrated data; means for training a machine learning model using the preprocessed data; A means for predicting suitable rewards for users using a trained machine learning model; and means for transmitting the predicted gift to the user terminal; A means of displaying predicted gifts on the user's device A system including:

Description

[Technical Field]

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

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

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

[0004] In recent years, the spread of e-commerce has led to an increasing demand for product recommendation systems that can match user preferences. However, conventional systems have had difficulty effectively integrating and analyzing a wide variety of data, including not only users' purchase history and genre interests, but also their responses to questions. As a result, there have been problems with insufficient personalized recommendations for users, or with recommendations that do not meet user expectations. The purpose of this invention is to solve these problems and provide a system that can recommend highly accurate personalized thank-you gifts to users. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means. The system includes a means for collecting a user's purchase history, a means for collecting a user's genres of interest, and a means for collecting answers to the user's questions, and a means for integrating these data. The system also provides a means for preprocessing the integrated data, filling in missing values, and normalizing the data. The system includes a means for training a machine learning model using the preprocessed data and predicting a gift suitable for the user using the trained model. The system also includes a means for sending the predicted gift to the user's terminal and displaying the predicted gift on the user's terminal. This makes it possible to integrate and analyze a variety of user data and recommend gifts with greater accuracy.

[0006] "User purchase history" refers to data that indicates information about products that a user has purchased in the past.

[0007] "User interest genre" is data that refers to information about specific categories or fields in which a user is interested.

[0008] "Answers to user questions" refers to data that refers to answers provided to questions asked by users.

[0009] "Data integration" refers to the process of combining data from different data sources into a single data set.

[0010] "Data preprocessing" refers to operations such as complementing, normalizing, and cleaning data in preparation for analysis or machine learning.

[0011] "Missing value imputation" refers to the process of filling in missing values ​​in a dataset in an appropriate way.

[0012] "Data normalization" refers to the process of aligning data units and scales to make them uniform.

[0013] A "machine learning model" refers to an algorithm or program that learns patterns and rules from data and performs tasks such as prediction and classification.

[0014] "Trained model" refers to a machine learning model that has been optimized to perform a specific task using training data.

[0015] "Gift prediction" refers to the process of using a machine learning model to predict and select a gift that is suitable for the user.

[0016] "User terminal" refers to devices such as computers, smartphones, and tablets operated by general users.

[0017] "Displaying recommended items" refers to displaying predicted gifts on the screen of the user's device. [Brief explanation of the drawings]

[0018] [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

[0019] 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.

[0020] First, the terms used in the following description will be explained.

[0021] 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).

[0022] 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.

[0023] 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.

[0024] 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.

[0025] 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."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 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.

[0029] 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).

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 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.

[0036] 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.

[0037] 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.

[0038] 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."

[0039] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0040] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0041] After preprocessing is complete, the server trains the machine learning model. It separates the user's data into features and targets, and splits them into training data and test data. It uses the training data to train a model using a machine learning algorithm such as random forest. The trained model is then used to predict rewards for the user.

[0042] The server then uses the trained model to predict suitable gifts for specific users. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0043] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. Specifically, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0044] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0045] The processing flow will be explained below.

[0046] Step 1:

[0047] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0048] Step 2:

[0049] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0050] Step 3:

[0051] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0052] Step 4:

[0053] The server integrates the collected purchase history, genre interests, and answers to questions. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0054] Step 5:

[0055] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "No response." Finally, it normalizes data with different scales and converts it into a format suitable for analysis and machine learning.

[0056] Step 6:

[0057] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0058] Step 7:

[0059] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, it predicts "sports goods" or "tickets to a cooking class."

[0060] Step 8:

[0061] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0062] Example 1

[0063] 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."

[0064] Conventional recommendation systems struggle to collect, integrate, and analyze diverse user data, making it difficult to provide accurate, personalized rewards to individual users. Furthermore, the accuracy of machine learning models was hindered by insufficient data imputation and normalization. This resulted in failures to increase user satisfaction and to provide optimal services to providers.

[0065] 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.

[0066] In this invention, the server includes means for collecting a user's purchase history from a database, means for collecting the user's genre interests from the database, means for collecting answers to the user's questions from the database, means for combining the collected data into an integrated dataset, means for deleting unnecessary columns from the integrated data, means for imputing missing values, means for normalizing the data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for exporting a list of predicted gifts and sending it to a user terminal, and means for displaying the predicted gifts on the user terminal. This enables efficient collection and preprocessing of a variety of user data and highly accurate personalized gift recommendations.

[0067] A "database" is a system that structures and stores information so that it can be easily retrieved and manipulated.

[0068] "User purchase history" refers to data about products and services purchased by a user in the past.

[0069] "User interest genres" is information about the fields and categories in which a user is interested.

[0070] An "answer to a question" is answer data selected or entered by a user in response to a specific question.

[0071] An "integrated dataset" is one in which data collected from multiple different data sources is combined into one dataset.

[0072] An "unnecessary column" is a data item that is not essential for analysis or processing.

[0073] A "missing value" is a location where data that should be present is missing.

[0074] "Imputation" is the operation of filling in missing data with appropriate values.

[0075] "Normalization" is a data conversion operation to standardize data with different scales or units.

[0076] A "machine learning model" is an algorithm that uses data to learn and make predictions and classifications.

[0077] "Training" is the process of feeding data to a machine learning model to improve its prediction and classification capabilities.

[0078] A "reward" is a gift or benefit given to a user when they take a specific action or fulfill a certain condition.

[0079] "Prediction" is the act of estimating and judging future events and outcomes based on past data.

[0080] "Export" is an operation to output data to the outside in a specific format.

[0081] A "user terminal" is a device such as a computer or smartphone operated by a user.

[0082] "Display" means visually showing information on the screen of a user's terminal.

[0083] To implement this invention, the server must first collect user data. Specifically, the server extracts the user's purchase history, genres of interest, and answers to questions from a database. The database used is a structured database such as MySQL (registered trademark) or PostgreSQL.

[0084] The server combines the collected data into a single unified dataset, which can be done using a data management tool or library (e.g., the Pandas library in Python).

[0085] The server then preprocesses the integrated data. Specifically, it removes unnecessary columns, imputes missing values, and normalizes the data. For example, in the case of purchase history, unnecessary columns such as "purchase date" are removed, unanswered questions are imputed as "unanswered," and data is normalized to unify the units and scale of each item.

[0086] The server uses the preprocessed dataset to train a machine learning model. Algorithms used for this include random forests and support vector machines. The Python scikit-learn library is commonly used for training.

[0087] The server uses the trained model to predict suitable gifts for specific users. By specifying the user ID, the necessary features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0088] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0089] Illustrative prompt examples

[0090] Please recommend a gift based on the data for user ID 12345. The following data is available:

[0091] Purchase History: ['Books', 'Household Goods', 'Gadgets']

[0092] Interests: ['Sports', 'Cooking']

[0093] Answers to questions: { "Q1": "Yes", "Q2": "No", "Q3": "No"}

[0094] This allows for efficient collection and pre-processing of diverse user data, enabling highly accurate personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0095] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0096] Step 1:

[0097] Data collection

[0098] The server collects the user's purchase history, genre interests, and answers to questions from a database.

[0099] Input: User ID

[0100] Output: Dataset of purchase history, interest genres, and answers to questions

[0101] What happens: The server executes the SQL query to extract the required information from the database, and returns the results in the form of a DataFrame or a list.

[0102] Step 2:

[0103] Data integration

[0104] The server combines the collected data into a consolidated data set.

[0105] Input: Dataset of purchase history, genre interests, and answers to questions

[0106] Output: A consolidated dataset

[0107] Specific operation: The server uses Python's Pandas library to merge purchase history, interest genres, and answers to questions to generate a single data frame.

[0108] Step 3:

[0109] Data Preprocessing

[0110] The server removes unnecessary columns from the integrated data, imputes missing values, and normalizes the data.

[0111] Input: Integrated dataset

[0112] Output: Preprocessed dataset

[0113] Specific operation: The server uses Pandas to delete the "Purchase Date" column, complete unanswered items as "Unanswered," and normalize the data to unify the units and scale of each item.

[0114] Step 4:

[0115] Training a machine learning model

[0116] The server uses the preprocessed dataset to train a machine learning model.

[0117] Input: Preprocessed dataset

[0118] Output: A trained machine learning model

[0119] Specific operation: The server separates the data into training data and test data, and trains the model using the random forest algorithm in the scikit-learn library.

[0120] Step 5:

[0121] Gift predictions

[0122] The server uses the trained model to predict suitable gifts for a particular user.

[0123] Input: User ID, preprocessed dataset

[0124] Output: A list of recommended items

[0125] Specific operation: Based on the user ID, the server extracts the necessary features, inputs them into the model, and predicts the recommended items. For example, it might recommend "sports goods" or "cooking class tickets" to user ID 12345.

[0126] Step 6:

[0127] Submit a recommendation

[0128] The server sends a list of predicted gifts to the user's terminal.

[0129] Input: List of recommended items

[0130] Output: Data sent to the terminal

[0131] Specific operation: The server converts the list of recommended items into JSON format and sends it to the user's device.

[0132] Step 7:

[0133] Displaying recommended items

[0134] The terminal processes the received data and displays it to the user.

[0135] Input: Data sent from the server

[0136] Output: Recommended items displayed on the user's screen

[0137] Specific operation: The device parses the received JSON data and displays "Recommended gifts: sports merchandise, cooking class tickets" for user ID 12345 on the screen.

[0138] In this way, user data can be efficiently collected and preprocessed, and personalized gift recommendations can be made with high accuracy. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0139] (Application example 1)

[0140] 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."

[0141] In recent years, the use of online shopping sites has increased, leading to a growing demand for personalized recommendation systems for users. However, conventional recommendation systems have been unable to effectively utilize diverse user data, resulting in low recommendation accuracy and insufficient improvement in user satisfaction. Furthermore, there is a lack of a means to effectively display recommended items on the user's device, which reduces user convenience.

[0142] 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.

[0143] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user's terminal, means for displaying the predicted gift on the user's terminal, means for outputting a list of the predicted gift in JSON format, and means for providing a user interface for displaying the predicted gift on the user's terminal. This enables the integration and analysis of various user data to enable more accurate and personalized gift recommendations. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0144] "User" means an individual or corporation that uses the online shopping site.

[0145] "Purchase history" refers to a record of past purchases of products and services made by a user on an online shopping site.

[0146] "Interest genre" is information about the categories of products and services in which a user is interested.

[0147] "Answers to questions" are data on answers given by users to specific questions.

[0148] "Data integration" is the process of combining multiple types of collected data into a single data set.

[0149] "Data preprocessing" is the process of removing unnecessary columns from the integrated data, filling in missing values, and normalizing the data.

[0150] A "machine learning model" is an algorithm that is trained with data to perform a specific task.

[0151] "Gift prediction" refers to using a machine learning model to predict the best gift for a specific user.

[0152] A "user terminal" is a device operated by a user (such as a smartphone, tablet, or PC).

[0153] "Export predicted gift list in JSON format" means exporting the list of items predicted by the machine learning model in JSON (JavaScript Object Notation) format.

[0154] A "user interface" is a screen or input means through which a user interacts with a digital system.

[0155] To implement this invention, a server first collects user data. This data includes the user's purchase history, genre interests, and answers to questions. This data is collected from user devices such as smartphones, tablets, and PCs and stored on the server. Examples of hardware used include servers, smartphones, and PCs. Examples of software used include Python, Pandas, Firebase, and JSON.

[0156] The server integrates the collected data into a single dataset. At this time, it deletes unnecessary columns (such as purchase date) and imputes missing values ​​as "unanswered." It also standardizes the units and scale of the data and normalizes it. This data preprocessing is performed using Python and Pandas.

[0157] The preprocessed data is used to train a machine learning model on the server. Here, an algorithm such as random forest is used. The data is divided into training data and test data, and the model is trained using the training data. The software used includes Python's Scikit-learn.

[0158] Next, the server uses the trained machine learning model to predict the appropriate gift for the user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. The prediction results are exported in JSON format and sent to the user's device.

[0159] The user receives the predicted gift list and displays it through a user interface, such as a smartphone app or web browser, using React Native and HTML / CSS / JavaScript.

[0160] To give a specific example, a prompt for user ID 12345 might be, "Please recommend gift items based on user ID 12345's past purchase history and genre interests." Processing based on this prompt might result in, for example, "Recommended gifts: sports merchandise, cooking class tickets" being displayed on the user's device.

[0161] This makes it possible to integrate and analyze a variety of user data to recommend more accurate and personalized thank-you gifts. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0162] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0163] Step 1:

[0164] The server collects the user's purchase history, genre interests, and answers to questions from the user's device. The collected data is stored on the server. The input is raw data sent from the user's device, and the output is integrated data stored on the server. Specifically, the smartphone app sends data to the Firebase database, and the server receives the data.

[0165] Step 2:

[0166] The data collected by the server is integrated. At this time, unnecessary columns (such as purchase date) are deleted and missing values ​​are imputed as "unanswered." Furthermore, the data units and scales are unified and normalized. The input is raw data stored on the server, and the output is preprocessed data. Specifically, the data frame is manipulated using Python's Pandas library to fill in missing values.

[0167] Step 3:

[0168] The server uses the preprocessed data to train a machine learning model. It splits the data into training and test data and trains the model using an algorithm such as random forest. The input is the preprocessed data, and the output is the trained machine learning model. Specifically, it uses Scikit-learn functions to split the data and train the model.

[0169] Step 4:

[0170] The server uses a trained machine learning model to predict suitable gifts for the user. By specifying the user ID, features are extracted and input into the model to predict recommended items. The input is the user ID and feature data, and the output is a list of recommended gifts. Specifically, the feature data is input into the trained model to obtain the predicted results.

[0171] Step 5:

[0172] The server exports the predicted gift list in JSON format and sends it to the user's device. The input is the list of recommended gifts, and the output is JSON data. Specifically, the data is converted to JSON format using the Python json library and sent via the API.

[0173] Step 6:

[0174] The JSON data received by the user's device is parsed and displayed on the user interface. The input is the JSON data sent from the server, and the output is a list of gifts that is displayed in a form that the user can see. Specifically, the data is parsed using a framework such as React Native and displayed on the screen.

[0175] By going through the above steps, a system is created that can recommend personalized gifts to users.

[0176] 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.

[0177] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0178] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0179] After preprocessing is complete, the server trains the machine learning model by dividing the user's data into features and targets, and splitting them into training data and test data. For example, it uses a random forest algorithm to train the model.

[0180] The server then uses the trained model to predict suitable gifts for a specific user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. For example, it might recommend "sports merchandise" or "tickets to a cooking class" to user ID 12345.

[0181] Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. The server uses the emotion engine to collect the user's emotion data (facial recognition data, text analysis data, etc.) and integrates it with their purchase history, genre interests, and answers to questions. This generates integrated data that also takes emotion data into account.

[0182] The server trains a machine learning model based on the integrated data, including emotional data. This model can predict gifts based on the user's emotions. For example, if a user has recently been feeling stressed, the model will enhance recommendations for relaxation items.

[0183] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format or similar and sent to the device. The device analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed on the screen.

[0184] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their tastes and emotions, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0185] The processing flow will be explained below.

[0186] Step 1:

[0187] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0188] Step 2:

[0189] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0190] Step 3:

[0191] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0192] Step 4:

[0193] The server collects emotion data using an emotion engine that recognizes the user's emotions. Specifically, it analyzes the user's emotions using facial recognition data, text analysis data, etc., and obtains information such as "happiness," "sadness," and "stress." For example, the emotion of user ID 12345 may be analyzed as "I've been feeling stressed recently."

[0194] Step 5:

[0195] The server integrates the collected purchase history, genre interests, answers to questions, and emotional data. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0196] Step 6:

[0197] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "no response." Finally, it normalizes data with different scales and converts them into a format suitable for analysis and machine learning. For example, it quantifies and handles emotional data.

[0198] Step 7:

[0199] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0200] Step 8:

[0201] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports goods," "cooking class tickets," and "relaxation goods" are recommended based on past data and emotional data.

[0202] Step 9:

[0203] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" are displayed on the screen. Based on this information, the user can select products and services that match their preferences and emotions.

[0204] Example 2

[0205] 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."

[0206] Modern e-commerce sites and service providers are required to recommend personalized gifts to users. However, existing methods rely only on data such as a user's purchase history, genre interests, and answers to questions, which limits recommendation accuracy and makes it difficult to respond flexibly to users' emotions and current situations. Furthermore, these methods face challenges such as missing data, inconsistent scale, and low-quality training data for machine learning models.

[0207] 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.

[0208] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting suitable items for the user using the trained machine learning model, means for collecting user emotion data and incorporating it into the integrated data, means for retraining the machine learning model using the emotion data, means for transmitting the predicted items to the user terminal, and means for displaying the predicted items on the user terminal. This makes it possible to integrate multifaceted information about the user and recommend personalized gifts with high accuracy.

[0209] "User purchase history" refers to all information about purchases made by a user in the past, specifically data including the purchase date and time, purchased items, quantity, and price.

[0210] "User interest genres" refers to categories or themes that the user is interested in, and specifically includes information on sports, music, cooking, etc.

[0211] "Answers to user questions" refers to the answers users give to surveys and question forms, and is data that reflects the user's preferences and opinions.

[0212] "Means of integrating collected data" refers to the process of combining data obtained from multiple disparate sources into a single integrated data set.

[0213] "Means for preprocessing integrated data" refers to processes such as filling in missing values ​​and deleting unnecessary data, which are carried out to improve the accuracy and quality of the data.

[0214] "Means of training a machine learning model" refers to the process of using collected data to train a machine learning algorithm and prepare it to make predictions on new data in the future.

[0215] "Means for predicting suitable products for a user" means a process that uses a trained machine learning model to predict the products that a user will like or need based on the characteristics of that user.

[0216] "User emotional data" refers to data that indicates the user's emotional state, and specifically includes emotional information obtained from facial expression recognition data, text analysis results, and the like.

[0217] "Means for collecting and incorporating emotional data into aggregated data" refers to the process of collecting data about user emotions and combining it with other user information into a single aggregated data set.

[0218] "Means for retraining a machine learning model with emotion data" refers to the process of retraining a machine learning model using a new dataset that includes emotion data to improve the accuracy of predictions that take emotion into account.

[0219] "Means of sending to user terminal" refers to the process of transferring data from the server to the user's device (such as a smartphone or PC) via a network.

[0220] "Means for displaying predicted items on a user device" refers to a mechanism for visually displaying predicted gifts or recommended products on a user device.

[0221] To implement this invention, the server first collects user data, such as the user's purchase history, genre interests, and answers to questions, from a database. This process involves retrieving the data using SQL queries and organizing it using the Pandas library.

[0222] Next, the server preprocesses the collected data. Preprocessing includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "Purchase Date" column is deleted from purchase history data, and any items with missing values ​​are filled in with a default value such as "No Answer." Normalization is also performed to unify the scale of the numerical data to the range of 0 to 1. These preprocessing steps are performed using data processing libraries such as Pandas and NumPy.

[0223] The server uses the preprocessed data to train a machine learning model. This process uses the Scikit-learn library. For example, the dataset is divided into features and targets, and then further divided into training data and test data. The model is trained using algorithms such as random forests, and then the model's applicability to new data is evaluated.

[0224] Using the trained model, the server predicts suitable items for the user. Specifically, features are extracted based on the specified user ID and input into the trained model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended.

[0225] The server also collects user emotional data and incorporates it into the integrated data. This process uses facial recognition and text analysis data. The collected emotional data is analyzed using an emotional engine and combined with purchase history, genre interests, and answers to questions. This creates a dataset that takes into account the user's emotions.

[0226] The server retrains the machine learning model using a new dataset that includes emotion data. For example, by calculating "stress levels" from facial recognition data and retraining the model using data that includes this, it becomes possible to make predictions that take user emotions into account.

[0227] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format and sent to the device. The user's device analyzes the received data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0228] This system allows users to easily find products and services that match their tastes and emotions, and also allows providers to improve user satisfaction.

[0229] Example prompt sentence:

[0230] "Generate a list of recommended gifts based on the user's recent interests and sentiment."

[0231] "Please recommend suitable items for user ID 12345, taking into account the sentiment data."

[0232] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0233] Step 1: Collect user data

[0234] The server extracts the user's purchase history, interest genres, and answers to questions from a database. The input is a query based on the user ID, and the output is a dataset containing purchase history data, interest genre data, and question-answer data. Specifically, it uses an SQL query to retrieve the required data and converts it into a Pandas data frame. For example, it uses the query SELECT FROM user_data WHERE user_id = '12345';

[0235] Step 2: Integrate the data

[0236] The server merges the collected data. The input is the multiple datasets collected in step 1, and the output is a single merged dataset. Specifically, it merges data frames using the Pandas library. For example, it performs an operation such as pd.merge(purchase_data, interest_data, on='user_id').

[0237] Step 3: Preprocessing the data

[0238] The server preprocesses the integrated data. The input is the integrated dataset, and the output is the preprocessed dataset. Preprocessing includes removing unnecessary columns, imputing missing values, and normalizing the data. Specifically, it uses Pandas to remove unnecessary columns and impute missing values. For example, it performs operations such as df.drop(columns=['purchase_date']) and df.fillna('Unanswered').

[0239] Step 4: Train the machine learning model

[0240] The server trains a machine learning model using the preprocessed data. The input is the preprocessed dataset, and the output is a trained machine learning model. Specifically, a random forest model is trained using the Scikit-learn library. For example, the data is split using the train_test_split function, and the model is trained using RandomForestClassifier.

[0241] Step 5: Predicting recommended items

[0242] The server uses the trained model to predict suitable items for a user. The input is the features of a specific user, and the output is a list of recommended items. Specifically, the features are input into the model, and the prediction result is obtained using the predict function. For example, "sports goods" and "tickets to a cooking class" are recommended for user ID 12345.

[0243] Step 6: Collect and integrate emotion data

[0244] The server collects user emotion data (e.g., facial recognition data, text analysis data) and incorporates it into integrated data. The input is emotion data, and the output is an integrated dataset including emotion data. Specifically, face recognition is performed using the OpenCV library, and analysis is performed using the emotion engine. For example, face information is acquired using cv2.VideoCapture, and the "stress level" is calculated using the emotion engine.

[0245] Step 7: Retrain the model with emotion data

[0246] The server retrains the machine learning model using a new dataset containing emotion data. The input is the integrated dataset containing emotion data, and the output is the retrained model. Specifically, the server again splits the data into training and test data using the Scikit-learn library, and then retrains the model.

[0247] Step 8: Notification of recommended items

[0248] The server sends the predicted gifts to the user's device. The input is a list of recommended items, and the output is the data sent to the user's device. Specifically, Flask or Django is used to export the data in JSON format. The user's device parses this data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0249] (Application example 2)

[0250] 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."

[0251] Traditional recommendation systems only consider a user's purchase history and genre interests, and are unable to reflect the user's emotions and momentary state of mind, thereby failing to maximize the user experience. Furthermore, due to insufficient data preprocessing and missing value imputation, recommendation accuracy is low. Furthermore, traditional systems also suffer from inefficient data integration and machine learning model training.

[0252] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0253] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for collecting the user's emotional data, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user terminal, and means for displaying the predicted gift on the user terminal. This allows for more personalized recommendations that reflect the user's emotions and state of mind. Furthermore, data accuracy can be improved by filling in missing values ​​and normalizing them, thereby improving the accuracy of recommendations.

[0254] "Means for collecting user purchase history" refers to a method for retrieving and storing information about products and services that users have previously purchased from a database.

[0255] "Means for collecting user interest genres" refers to methods for identifying categories or topics that users are interested in and collecting that information.

[0256] A "means for collecting answers to user questions" is a method for collecting answers to questions posed to users and storing them in a database.

[0257] "Means for collecting user emotional data" refers to a method for analyzing emotions that can be read from a user's facial expressions and text, and collecting this as data.

[0258] "Means for integrating collected data" refers to a method for integrating information collected from different data sources, such as purchase history, genre interests, question answers, and emotional data, into a single dataset.

[0259] "Means for preprocessing the integrated data" refers to methods for organizing the data by removing unnecessary columns from the integrated dataset, imputing missing values, and normalizing the data.

[0260] A "means for training a machine learning model using preprocessed data" is a method for providing preprocessed data as input to a machine learning algorithm to train a predictive model.

[0261] "Means for predicting suitable gifts for users using a trained machine learning model" means a method for using a trained machine learning model to predict the best product or service for a particular user.

[0262] "Means for sending predicted gifts to a user's device" refers to a method for sending information about products or services predicted by a machine learning model to a user's device.

[0263] The "means for displaying predicted gifts on the user's device" is a method for displaying the transmitted list of predicted goods and services on the user's device.

[0264] This invention is a system that collects and integrates a user's purchase history, genre interests, answers to questions, and emotional data to recommend highly accurate personalized gifts. This system uses specific means to execute each step, with the aim of improving the user experience. Each means and its implementation method will be explained in detail below.

[0265] First, the server collects user data, including information about products and services the user has purchased in the past, the user's interests, questions the user has answered, and emotional data obtained through facial recognition and text analysis. This data is stored in a database and integrated as needed.

[0266] Next, the server integrates the collected data. The collected purchase history, genre interests, answers to questions, and sentiment data are compiled into a single integrated dataset. This integration is performed using Python's Pandas library, among other tools. The integrated data is then preprocessed, which involves removing unnecessary columns, filling in missing values, and normalizing the data.

[0267] The server uses the preprocessed data to train a machine learning model. For this training, the scikit-learn library is used, applying the random forest algorithm. The data is split into training and test data, and the accuracy of the model is verified as it progresses.

[0268] Once trained, the machine learning model is used to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model. For example, recommendations can be made for user ID 12345, such as "sports goods," "tickets to a cooking class," and "relaxation goods."

[0269] Furthermore, the server can make recommendations that take into account the user's emotions based on the integrated data, including emotional data. For example, if the emotional data indicates that the user has recently been feeling stressed, the server will strengthen recommendations for relaxation products.

[0270] The predicted gift is sent from the server to the user's device and displayed on the device. The data is exported in JSON format and analyzed on the user's smartphone or other device.

[0271] For example, if a user has frequently purchased fitness equipment in the past six months and has recently experienced increased stress, the system will recommend relaxation products, yoga mats, stretching equipment, etc.

[0272] Examples of prompts to input to a generative AI model include:

[0273] "Please use user ID 12345's past purchase history, genre interests, question responses, and sentiment data to predict personalized product recommendations."

[0274] In this way, the system can integrate and analyze a variety of user data to recommend more accurate and personalized gifts, benefiting both users and gift givers.

[0275] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0276] Step 1:

[0277] The server collects the user's data.

[0278] Specifically, the server retrieves user purchase history, genre interests, answers to questions, and emotional data obtained through facial recognition and text analysis from the database. These data are read from separate tables or files and stored in a unified format.

[0279] Input: Raw data from databases and various APIs

[0280] Output: User data stored in a unified format

[0281] Step 2:

[0282] The server consolidates the collected data.

[0283] The collected purchase history, interest genres, question answers, and sentiment data are merged into a single data frame. This integration is performed using the Python Pandas library.

[0284] Input: User data collected in step 1

[0285] Output: A merged data frame

[0286] Step 3:

[0287] The server preprocesses the consolidated data.

[0288] Preprocessing involves deleting unnecessary columns, imputing missing values, and normalizing the data. Specifically, the "Purchase Date" column is deleted, NULL values ​​are imputed as "No Answer," and the units and scales of each data are unified.

[0289] Input: Merged data frame

[0290] Output: Preprocessed data frame

[0291] Step 4:

[0292] The server uses the preprocessed data to train a machine learning model.

[0293] The preprocessed data is divided into features and targets, and then split into training and test data. A model is trained using the random forest algorithm in scikit-learn.

[0294] Input: Preprocessed data frame

[0295] Output: A trained machine learning model

[0296] Step 5:

[0297] The server uses a trained machine learning model to predict suitable gifts for users.

[0298] By extracting features based on a specific user ID and inputting them into an existing model, the system predicts the recommended gift.

[0299] Input: User ID and preprocessed data frame

[0300] Output: A list of recommended gifts

[0301] Step 6:

[0302] The server sends the predicted gift to the user's device.

[0303] Export the list of recommended items in JSON format or other format and send it to the user's smartphone or other device.

[0304] Input: List of recommended gifts

[0305] Output: Data exported in JSON format

[0306] Step 7:

[0307] The terminal displays the predicted gift received from the server.

[0308] The received JSON data is analyzed and displayed on the screen for the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed.

[0309] Input: JSON data sent from the server

[0310] Output: A list of gifts displayed on the screen

[0311] 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.

[0312] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0313] 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.

[0314] [Second embodiment]

[0315] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0316] 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.

[0317] 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).

[0318] 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.

[0319] 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.

[0320] 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).

[0321] 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.

[0322] 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.

[0323] 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.

[0324] 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.

[0325] 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.

[0326] 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."

[0327] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0328] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0329] After preprocessing is complete, the server trains the machine learning model. It separates the user's data into features and targets, and splits them into training data and test data. It uses the training data to train a model using a machine learning algorithm such as random forest. The trained model is then used to predict rewards for the user.

[0330] The server then uses the trained model to predict suitable gifts for specific users. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0331] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. Specifically, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0332] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0333] The processing flow will be explained below.

[0334] Step 1:

[0335] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0336] Step 2:

[0337] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0338] Step 3:

[0339] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0340] Step 4:

[0341] The server integrates the collected purchase history, genre interests, and answers to questions. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0342] Step 5:

[0343] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "No response." Finally, it normalizes data with different scales and converts it into a format suitable for analysis and machine learning.

[0344] Step 6:

[0345] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0346] Step 7:

[0347] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, it predicts "sports goods" or "tickets to a cooking class."

[0348] Step 8:

[0349] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0350] Example 1

[0351] 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."

[0352] Conventional recommendation systems struggle to collect, integrate, and analyze diverse user data, making it difficult to provide accurate, personalized rewards to individual users. Furthermore, the accuracy of machine learning models was hindered by insufficient data imputation and normalization. This resulted in failures to increase user satisfaction and to provide optimal services to providers.

[0353] 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.

[0354] In this invention, the server includes means for collecting a user's purchase history from a database, means for collecting the user's genre interests from the database, means for collecting answers to the user's questions from the database, means for combining the collected data into an integrated dataset, means for deleting unnecessary columns from the integrated data, means for imputing missing values, means for normalizing the data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for exporting a list of predicted gifts and sending it to a user terminal, and means for displaying the predicted gifts on the user terminal. This enables efficient collection and preprocessing of a variety of user data and highly accurate personalized gift recommendations.

[0355] A "database" is a system that structures and stores information so that it can be easily retrieved and manipulated.

[0356] "User purchase history" refers to data about products and services purchased by a user in the past.

[0357] "User interest genres" is information about the fields and categories in which a user is interested.

[0358] An "answer to a question" is answer data selected or entered by a user in response to a specific question.

[0359] An "integrated dataset" is one in which data collected from multiple different data sources is combined into one dataset.

[0360] An "unnecessary column" is a data item that is not essential for analysis or processing.

[0361] A "missing value" is a location where data that should be present is missing.

[0362] "Imputation" is the operation of filling in missing data with appropriate values.

[0363] "Normalization" is a data conversion operation to standardize data with different scales or units.

[0364] A "machine learning model" is an algorithm that uses data to learn and make predictions and classifications.

[0365] "Training" is the process of feeding data to a machine learning model to improve its prediction and classification capabilities.

[0366] A "reward" is a gift or benefit given to a user when they take a specific action or fulfill a certain condition.

[0367] "Prediction" is the act of estimating and judging future events and outcomes based on past data.

[0368] "Export" is an operation to output data to the outside in a specific format.

[0369] A "user terminal" is a device such as a computer or smartphone operated by a user.

[0370] "Display" means visually showing information on the screen of a user's terminal.

[0371] To implement this invention, the server must first collect user data. Specifically, the server extracts the user's purchase history, genres of interest, and answers to questions from a database. The database used is a structured database such as MySQL or PostgreSQL.

[0372] The server combines the collected data into a single unified dataset, which can be done using a data management tool or library (e.g., the Pandas library in Python).

[0373] The server then preprocesses the integrated data. Specifically, it removes unnecessary columns, imputes missing values, and normalizes the data. For example, in the case of purchase history, unnecessary columns such as "purchase date" are removed, unanswered questions are imputed as "unanswered," and data is normalized to unify the units and scale of each item.

[0374] The server uses the preprocessed dataset to train a machine learning model. Algorithms used for this include random forests and support vector machines. The Python scikit-learn library is commonly used for training.

[0375] The server uses the trained model to predict suitable gifts for specific users. By specifying the user ID, the necessary features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0376] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0377] Illustrative prompt examples

[0378] Please recommend a gift based on the data for user ID 12345. The following data is available:

[0379] Purchase History: ['Books', 'Household Goods', 'Gadgets']

[0380] Interests: ['Sports', 'Cooking']

[0381] Answers to questions: { "Q1": "Yes", "Q2": "No", "Q3": "No"}

[0382] This allows for efficient collection and pre-processing of diverse user data, enabling highly accurate personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0383] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0384] Step 1:

[0385] Data collection

[0386] The server collects the user's purchase history, genre interests, and answers to questions from a database.

[0387] Input: User ID

[0388] Output: Dataset of purchase history, interest genres, and answers to questions

[0389] What happens: The server executes the SQL query to extract the required information from the database, and returns the results in the form of a DataFrame or a list.

[0390] Step 2:

[0391] Data integration

[0392] The server combines the collected data into a consolidated data set.

[0393] Input: Dataset of purchase history, genre interests, and answers to questions

[0394] Output: A consolidated dataset

[0395] Specific operation: The server uses Python's Pandas library to merge purchase history, interest genres, and answers to questions to generate a single data frame.

[0396] Step 3:

[0397] Data Preprocessing

[0398] The server removes unnecessary columns from the integrated data, imputes missing values, and normalizes the data.

[0399] Input: Integrated dataset

[0400] Output: Preprocessed dataset

[0401] Specific operation: The server uses Pandas to delete the "Purchase Date" column, complete unanswered items as "Unanswered," and normalize the data to unify the units and scale of each item.

[0402] Step 4:

[0403] Training a machine learning model

[0404] The server uses the preprocessed dataset to train a machine learning model.

[0405] Input: Preprocessed dataset

[0406] Output: A trained machine learning model

[0407] Specific operation: The server separates the data into training data and test data, and trains the model using the random forest algorithm in the scikit-learn library.

[0408] Step 5:

[0409] Gift predictions

[0410] The server uses the trained model to predict suitable gifts for a particular user.

[0411] Input: User ID, preprocessed dataset

[0412] Output: A list of recommended items

[0413] Specific operation: Based on the user ID, the server extracts the necessary features, inputs them into the model, and predicts the recommended items. For example, it might recommend "sports goods" or "cooking class tickets" to user ID 12345.

[0414] Step 6:

[0415] Submit a recommendation

[0416] The server sends a list of predicted gifts to the user's terminal.

[0417] Input: List of recommended items

[0418] Output: Data sent to the terminal

[0419] Specific operation: The server converts the list of recommended items into JSON format and sends it to the user's device.

[0420] Step 7:

[0421] Displaying recommended items

[0422] The terminal processes the received data and displays it to the user.

[0423] Input: Data sent from the server

[0424] Output: Recommended items displayed on the user's screen

[0425] Specific operation: The device parses the received JSON data and displays "Recommended gifts: sports merchandise, cooking class tickets" for user ID 12345 on the screen.

[0426] In this way, user data can be efficiently collected and preprocessed, and personalized gift recommendations can be made with high accuracy. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0427] (Application example 1)

[0428] 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."

[0429] In recent years, the use of online shopping sites has increased, leading to a growing demand for personalized recommendation systems for users. However, conventional recommendation systems have been unable to effectively utilize diverse user data, resulting in low recommendation accuracy and insufficient improvement in user satisfaction. Furthermore, there is a lack of a means to effectively display recommended items on the user's device, which reduces user convenience.

[0430] 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.

[0431] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user's terminal, means for displaying the predicted gift on the user's terminal, means for outputting a list of the predicted gift in JSON format, and means for providing a user interface for displaying the predicted gift on the user's terminal. This enables the integration and analysis of various user data to enable more accurate and personalized gift recommendations. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0432] "User" means an individual or corporation that uses the online shopping site.

[0433] "Purchase history" refers to a record of past purchases of products and services made by a user on an online shopping site.

[0434] "Interest genre" is information about the categories of products and services in which a user is interested.

[0435] "Answers to questions" are data on answers given by users to specific questions.

[0436] "Data integration" is the process of combining multiple types of collected data into a single data set.

[0437] "Data preprocessing" is the process of removing unnecessary columns from the integrated data, filling in missing values, and normalizing the data.

[0438] A "machine learning model" is an algorithm that is trained with data to perform a specific task.

[0439] "Gift prediction" refers to using a machine learning model to predict the best gift for a specific user.

[0440] A "user terminal" is a device operated by a user (such as a smartphone, tablet, or PC).

[0441] "Export predicted gift list in JSON format" means exporting the list of items predicted by the machine learning model in JSON (JavaScript Object Notation) format.

[0442] A "user interface" is a screen or input means through which a user interacts with a digital system.

[0443] To implement this invention, a server first collects user data. This data includes the user's purchase history, genre interests, and answers to questions. This data is collected from user devices such as smartphones, tablets, and PCs and stored on the server. Examples of hardware used include servers, smartphones, and PCs. Examples of software used include Python, Pandas, Firebase, and JSON.

[0444] The server integrates the collected data into a single dataset. At this time, it deletes unnecessary columns (such as purchase date) and imputes missing values ​​as "unanswered." It also standardizes the units and scale of the data and normalizes it. This data preprocessing is performed using Python and Pandas.

[0445] The preprocessed data is used to train a machine learning model on the server. Here, an algorithm such as random forest is used. The data is divided into training data and test data, and the model is trained using the training data. The software used includes Python's Scikit-learn.

[0446] Next, the server uses the trained machine learning model to predict the appropriate gift for the user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. The prediction results are exported in JSON format and sent to the user's device.

[0447] The user receives the predicted gift list and displays it through a user interface, such as a smartphone app or web browser, using React Native and HTML / CSS / JavaScript.

[0448] To give a specific example, a prompt for user ID 12345 might be, "Please recommend gift items based on user ID 12345's past purchase history and genre interests." Processing based on this prompt might result in, for example, "Recommended gifts: sports merchandise, cooking class tickets" being displayed on the user's device.

[0449] This makes it possible to integrate and analyze a variety of user data to recommend more accurate and personalized thank-you gifts. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0450] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0451] Step 1:

[0452] The server collects the user's purchase history, genre interests, and answers to questions from the user's device. The collected data is stored on the server. The input is raw data sent from the user's device, and the output is integrated data stored on the server. Specifically, the smartphone app sends data to the Firebase database, and the server receives the data.

[0453] Step 2:

[0454] The data collected by the server is integrated. At this time, unnecessary columns (such as purchase date) are deleted and missing values ​​are imputed as "unanswered." Furthermore, the data units and scales are unified and normalized. The input is raw data stored on the server, and the output is preprocessed data. Specifically, the data frame is manipulated using Python's Pandas library to fill in missing values.

[0455] Step 3:

[0456] The server uses the preprocessed data to train a machine learning model. It splits the data into training and test data and trains the model using an algorithm such as random forest. The input is the preprocessed data, and the output is the trained machine learning model. Specifically, it uses Scikit-learn functions to split the data and train the model.

[0457] Step 4:

[0458] The server uses a trained machine learning model to predict suitable gifts for the user. By specifying the user ID, features are extracted and input into the model to predict recommended items. The input is the user ID and feature data, and the output is a list of recommended gifts. Specifically, the feature data is input into the trained model to obtain the predicted results.

[0459] Step 5:

[0460] The server exports the predicted gift list in JSON format and sends it to the user's device. The input is the list of recommended gifts, and the output is JSON data. Specifically, the data is converted to JSON format using the Python json library and sent via the API.

[0461] Step 6:

[0462] The JSON data received by the user's device is parsed and displayed on the user interface. The input is the JSON data sent from the server, and the output is a list of gifts that is displayed in a form that the user can see. Specifically, the data is parsed using a framework such as React Native and displayed on the screen.

[0463] By going through the above steps, a system is created that can recommend personalized gifts to users.

[0464] 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.

[0465] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0466] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0467] After preprocessing is complete, the server trains the machine learning model by dividing the user's data into features and targets, and splitting them into training data and test data. For example, it uses a random forest algorithm to train the model.

[0468] The server then uses the trained model to predict suitable gifts for a specific user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. For example, it might recommend "sports merchandise" or "tickets to a cooking class" to user ID 12345.

[0469] Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. The server uses the emotion engine to collect the user's emotion data (facial recognition data, text analysis data, etc.) and integrates it with their purchase history, genre interests, and answers to questions. This generates integrated data that also takes emotion data into account.

[0470] The server trains a machine learning model based on the integrated data, including emotional data. This model can predict gifts based on the user's emotions. For example, if a user has recently been feeling stressed, the model will enhance recommendations for relaxation items.

[0471] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format or similar and sent to the device. The device analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed on the screen.

[0472] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their tastes and emotions, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0473] The processing flow will be explained below.

[0474] Step 1:

[0475] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0476] Step 2:

[0477] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0478] Step 3:

[0479] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0480] Step 4:

[0481] The server collects emotion data using an emotion engine that recognizes the user's emotions. Specifically, it analyzes the user's emotions using facial recognition data, text analysis data, etc., and obtains information such as "happiness," "sadness," and "stress." For example, the emotion of user ID 12345 may be analyzed as "I've been feeling stressed recently."

[0482] Step 5:

[0483] The server integrates the collected purchase history, genre interests, answers to questions, and emotional data. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0484] Step 6:

[0485] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "no response." Finally, it normalizes data with different scales and converts them into a format suitable for analysis and machine learning. For example, it quantifies and handles emotional data.

[0486] Step 7:

[0487] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0488] Step 8:

[0489] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports goods," "cooking class tickets," and "relaxation goods" are recommended based on past data and emotional data.

[0490] Step 9:

[0491] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" are displayed on the screen. Based on this information, the user can select products and services that match their preferences and emotions.

[0492] Example 2

[0493] 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."

[0494] Modern e-commerce sites and service providers are required to recommend personalized gifts to users. However, existing methods rely only on data such as a user's purchase history, genre interests, and answers to questions, which limits recommendation accuracy and makes it difficult to respond flexibly to users' emotions and current situations. Furthermore, these methods face challenges such as missing data, inconsistent scale, and low-quality training data for machine learning models.

[0495] 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.

[0496] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting suitable items for the user using the trained machine learning model, means for collecting user emotion data and incorporating it into the integrated data, means for retraining the machine learning model using the emotion data, means for transmitting the predicted items to the user terminal, and means for displaying the predicted items on the user terminal. This makes it possible to integrate multifaceted information about the user and recommend personalized gifts with high accuracy.

[0497] "User purchase history" refers to all information about purchases made by a user in the past, specifically data including the purchase date and time, purchased items, quantity, and price.

[0498] "User interest genres" refers to categories or themes that the user is interested in, and specifically includes information on sports, music, cooking, etc.

[0499] "Answers to user questions" refers to the answers users give to surveys and question forms, and is data that reflects the user's preferences and opinions.

[0500] "Means of integrating collected data" refers to the process of combining data obtained from multiple disparate sources into a single integrated data set.

[0501] "Means for preprocessing integrated data" refers to processes such as filling in missing values ​​and deleting unnecessary data, which are carried out to improve the accuracy and quality of the data.

[0502] "Means of training a machine learning model" refers to the process of using collected data to train a machine learning algorithm and prepare it to make predictions on new data in the future.

[0503] "Means for predicting suitable products for a user" means a process that uses a trained machine learning model to predict the products that a user will like or need based on the characteristics of that user.

[0504] "User emotional data" refers to data that indicates the user's emotional state, and specifically includes emotional information obtained from facial expression recognition data, text analysis results, and the like.

[0505] "Means for collecting and incorporating emotional data into aggregated data" refers to the process of collecting data about user emotions and combining it with other user information into a single aggregated data set.

[0506] "Means for retraining a machine learning model with emotion data" refers to the process of retraining a machine learning model using a new dataset that includes emotion data to improve the accuracy of predictions that take emotion into account.

[0507] "Means of sending to user terminal" refers to the process of transferring data from the server to the user's device (such as a smartphone or PC) via a network.

[0508] "Means for displaying predicted items on a user device" refers to a mechanism for visually displaying predicted gifts or recommended products on a user device.

[0509] To implement this invention, the server first collects user data, such as the user's purchase history, genre interests, and answers to questions, from a database. This process involves retrieving the data using SQL queries and organizing it using the Pandas library.

[0510] Next, the server preprocesses the collected data. Preprocessing includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "Purchase Date" column is deleted from purchase history data, and any items with missing values ​​are filled in with a default value such as "No Answer." Normalization is also performed to unify the scale of the numerical data to the range of 0 to 1. These preprocessing steps are performed using data processing libraries such as Pandas and NumPy.

[0511] The server uses the preprocessed data to train a machine learning model. This process uses the Scikit-learn library. For example, the dataset is divided into features and targets, and then further divided into training data and test data. The model is trained using algorithms such as random forests, and then the model's applicability to new data is evaluated.

[0512] Using the trained model, the server predicts suitable items for the user. Specifically, features are extracted based on the specified user ID and input into the trained model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended.

[0513] The server also collects user emotional data and incorporates it into the integrated data. This process uses facial recognition and text analysis data. The collected emotional data is analyzed using an emotional engine and combined with purchase history, genre interests, and answers to questions. This creates a dataset that takes into account the user's emotions.

[0514] The server retrains the machine learning model using a new dataset that includes emotion data. For example, by calculating "stress levels" from facial recognition data and retraining the model using data that includes this, it becomes possible to make predictions that take user emotions into account.

[0515] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format and sent to the device. The user's device analyzes the received data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0516] This system allows users to easily find products and services that match their tastes and emotions, and also allows providers to improve user satisfaction.

[0517] Example prompt sentence:

[0518] "Generate a list of recommended gifts based on the user's recent interests and sentiment."

[0519] "Please recommend suitable items for user ID 12345, taking into account the sentiment data."

[0520] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0521] Step 1: Collect user data

[0522] The server extracts the user's purchase history, interest genres, and answers to questions from a database. The input is a query based on the user ID, and the output is a dataset containing purchase history data, interest genre data, and question-answer data. Specifically, it uses an SQL query to retrieve the required data and converts it into a Pandas data frame. For example, it uses the query SELECT FROM user_data WHERE user_id = '12345';

[0523] Step 2: Integrate the data

[0524] The server merges the collected data. The input is the multiple datasets collected in step 1, and the output is a single merged dataset. Specifically, it merges data frames using the Pandas library. For example, it performs an operation such as pd.merge(purchase_data, interest_data, on='user_id').

[0525] Step 3: Preprocessing the data

[0526] The server preprocesses the integrated data. The input is the integrated dataset, and the output is the preprocessed dataset. Preprocessing includes removing unnecessary columns, imputing missing values, and normalizing the data. Specifically, it uses Pandas to remove unnecessary columns and impute missing values. For example, it performs operations such as df.drop(columns=['purchase_date']) and df.fillna('Unanswered').

[0527] Step 4: Train the machine learning model

[0528] The server trains a machine learning model using the preprocessed data. The input is the preprocessed dataset, and the output is a trained machine learning model. Specifically, a random forest model is trained using the Scikit-learn library. For example, the data is split using the train_test_split function, and the model is trained using RandomForestClassifier.

[0529] Step 5: Predicting recommended items

[0530] The server uses the trained model to predict suitable items for a user. The input is the features of a specific user, and the output is a list of recommended items. Specifically, the features are input into the model, and the prediction result is obtained using the predict function. For example, "sports goods" and "tickets to a cooking class" are recommended for user ID 12345.

[0531] Step 6: Collect and integrate emotion data

[0532] The server collects user emotion data (e.g., facial recognition data, text analysis data) and incorporates it into integrated data. The input is emotion data, and the output is an integrated dataset including emotion data. Specifically, face recognition is performed using the OpenCV library, and analysis is performed using the emotion engine. For example, face information is acquired using cv2.VideoCapture, and the "stress level" is calculated using the emotion engine.

[0533] Step 7: Retrain the model with emotion data

[0534] The server retrains the machine learning model using a new dataset containing emotion data. The input is the integrated dataset containing emotion data, and the output is the retrained model. Specifically, the server again splits the data into training and test data using the Scikit-learn library, and then retrains the model.

[0535] Step 8: Notification of recommended items

[0536] The server sends the predicted gifts to the user's device. The input is a list of recommended items, and the output is the data sent to the user's device. Specifically, Flask or Django is used to export the data in JSON format. The user's device parses this data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0537] (Application example 2)

[0538] 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."

[0539] Traditional recommendation systems only consider a user's purchase history and genre interests, and are unable to reflect the user's emotions and momentary state of mind, thereby failing to maximize the user experience. Furthermore, due to insufficient data preprocessing and missing value imputation, recommendation accuracy is low. Furthermore, traditional systems also suffer from inefficient data integration and machine learning model training.

[0540] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0541] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for collecting the user's emotional data, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user terminal, and means for displaying the predicted gift on the user terminal. This allows for more personalized recommendations that reflect the user's emotions and state of mind. Furthermore, data accuracy can be improved by filling in missing values ​​and normalizing them, thereby improving the accuracy of recommendations.

[0542] "Means for collecting user purchase history" refers to a method for retrieving and storing information about products and services that users have previously purchased from a database.

[0543] "Means for collecting user interest genres" refers to methods for identifying categories or topics that users are interested in and collecting that information.

[0544] A "means for collecting answers to user questions" is a method for collecting answers to questions posed to users and storing them in a database.

[0545] "Means for collecting user emotional data" refers to a method for analyzing emotions that can be read from a user's facial expressions and text, and collecting this as data.

[0546] "Means for integrating collected data" refers to a method for integrating information collected from different data sources, such as purchase history, genre interests, question answers, and emotional data, into a single dataset.

[0547] "Means for preprocessing the integrated data" refers to methods for organizing the data by removing unnecessary columns from the integrated dataset, imputing missing values, and normalizing the data.

[0548] A "means for training a machine learning model using preprocessed data" is a method for providing preprocessed data as input to a machine learning algorithm to train a predictive model.

[0549] "Means for predicting suitable gifts for users using a trained machine learning model" means a method for using a trained machine learning model to predict the best product or service for a particular user.

[0550] "Means for sending predicted gifts to a user's device" refers to a method for sending information about products or services predicted by a machine learning model to a user's device.

[0551] The "means for displaying predicted gifts on the user's device" is a method for displaying the transmitted list of predicted goods and services on the user's device.

[0552] This invention is a system that collects and integrates a user's purchase history, genre interests, answers to questions, and emotional data to recommend highly accurate personalized gifts. This system uses specific means to execute each step, with the aim of improving the user experience. Each means and its implementation method will be explained in detail below.

[0553] First, the server collects user data, including information about products and services the user has purchased in the past, the user's interests, questions the user has answered, and emotional data obtained through facial recognition and text analysis. This data is stored in a database and integrated as needed.

[0554] Next, the server integrates the collected data. The collected purchase history, genre interests, answers to questions, and sentiment data are compiled into a single integrated dataset. This integration is performed using Python's Pandas library, among other tools. The integrated data is then preprocessed, which involves removing unnecessary columns, filling in missing values, and normalizing the data.

[0555] The server uses the preprocessed data to train a machine learning model. For this training, the scikit-learn library is used, applying the random forest algorithm. The data is split into training and test data, and the accuracy of the model is verified as it progresses.

[0556] Once trained, the machine learning model is used to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model. For example, recommendations can be made for user ID 12345, such as "sports goods," "tickets to a cooking class," and "relaxation goods."

[0557] Furthermore, the server can make recommendations that take into account the user's emotions based on the integrated data, including emotional data. For example, if the emotional data indicates that the user has recently been feeling stressed, the server will strengthen recommendations for relaxation products.

[0558] The predicted gift is sent from the server to the user's device and displayed on the device. The data is exported in JSON format and analyzed on the user's smartphone or other device.

[0559] For example, if a user has frequently purchased fitness equipment in the past six months and has recently experienced increased stress, the system will recommend relaxation products, yoga mats, stretching equipment, etc.

[0560] Examples of prompts to input to a generative AI model include:

[0561] "Please use user ID 12345's past purchase history, genre interests, question responses, and sentiment data to predict personalized product recommendations."

[0562] In this way, the system can integrate and analyze a variety of user data to recommend more accurate and personalized gifts, benefiting both users and gift givers.

[0563] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0564] Step 1:

[0565] The server collects the user's data.

[0566] Specifically, the server retrieves user purchase history, genre interests, answers to questions, and emotional data obtained through facial recognition and text analysis from the database. These data are read from separate tables or files and stored in a unified format.

[0567] Input: Raw data from databases and various APIs

[0568] Output: User data stored in a unified format

[0569] Step 2:

[0570] The server consolidates the collected data.

[0571] The collected purchase history, interest genres, question answers, and sentiment data are merged into a single data frame. This integration is performed using the Python Pandas library.

[0572] Input: User data collected in step 1

[0573] Output: A merged data frame

[0574] Step 3:

[0575] The server preprocesses the consolidated data.

[0576] Preprocessing involves deleting unnecessary columns, imputing missing values, and normalizing the data. Specifically, the "Purchase Date" column is deleted, NULL values ​​are imputed as "No Answer," and the units and scales of each data are unified.

[0577] Input: Merged data frame

[0578] Output: Preprocessed data frame

[0579] Step 4:

[0580] The server uses the preprocessed data to train a machine learning model.

[0581] The preprocessed data is divided into features and targets, and then split into training and test data. A model is trained using the random forest algorithm in scikit-learn.

[0582] Input: Preprocessed data frame

[0583] Output: A trained machine learning model

[0584] Step 5:

[0585] The server uses a trained machine learning model to predict suitable gifts for users.

[0586] By extracting features based on a specific user ID and inputting them into an existing model, the system predicts the recommended gift.

[0587] Input: User ID and preprocessed data frame

[0588] Output: A list of recommended gifts

[0589] Step 6:

[0590] The server sends the predicted gift to the user's device.

[0591] Export the list of recommended items in JSON format or other format and send it to the user's smartphone or other device.

[0592] Input: List of recommended gifts

[0593] Output: Data exported in JSON format

[0594] Step 7:

[0595] The terminal displays the predicted gift received from the server.

[0596] The received JSON data is analyzed and displayed on the screen for the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed.

[0597] Input: JSON data sent from the server

[0598] Output: A list of gifts displayed on the screen

[0599] 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.

[0600] 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.

[0601] 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.

[0602] [Third embodiment]

[0603] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0604] 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.

[0605] 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).

[0606] 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.

[0607] 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.

[0608] 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).

[0609] 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.

[0610] 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.

[0611] 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.

[0612] 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.

[0613] 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.

[0614] 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."

[0615] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0616] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0617] After preprocessing is complete, the server trains the machine learning model. It separates the user's data into features and targets, and splits them into training data and test data. It uses the training data to train a model using a machine learning algorithm such as random forest. The trained model is then used to predict rewards for the user.

[0618] The server then uses the trained model to predict suitable gifts for specific users. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0619] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. Specifically, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0620] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0621] The processing flow will be explained below.

[0622] Step 1:

[0623] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0624] Step 2:

[0625] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0626] Step 3:

[0627] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0628] Step 4:

[0629] The server integrates the collected purchase history, genre interests, and answers to questions. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0630] Step 5:

[0631] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "No response." Finally, it normalizes data with different scales and converts it into a format suitable for analysis and machine learning.

[0632] Step 6:

[0633] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0634] Step 7:

[0635] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, it predicts "sports goods" or "tickets to a cooking class."

[0636] Step 8:

[0637] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0638] Example 1

[0639] 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."

[0640] Conventional recommendation systems struggle to collect, integrate, and analyze diverse user data, making it difficult to provide accurate, personalized rewards to individual users. Furthermore, the accuracy of machine learning models was hindered by insufficient data imputation and normalization. This resulted in failures to increase user satisfaction and to provide optimal services to providers.

[0641] 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.

[0642] In this invention, the server includes means for collecting a user's purchase history from a database, means for collecting the user's genre interests from the database, means for collecting answers to the user's questions from the database, means for combining the collected data into an integrated dataset, means for deleting unnecessary columns from the integrated data, means for imputing missing values, means for normalizing the data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for exporting a list of predicted gifts and sending it to a user terminal, and means for displaying the predicted gifts on the user terminal. This enables efficient collection and preprocessing of a variety of user data and highly accurate personalized gift recommendations.

[0643] A "database" is a system that structures and stores information so that it can be easily retrieved and manipulated.

[0644] "User purchase history" refers to data about products and services purchased by a user in the past.

[0645] "User interest genres" is information about the fields and categories in which a user is interested.

[0646] An "answer to a question" is answer data selected or entered by a user in response to a specific question.

[0647] An "integrated dataset" is one in which data collected from multiple different data sources is combined into one dataset.

[0648] An "unnecessary column" is a data item that is not essential for analysis or processing.

[0649] A "missing value" is a location where data that should be present is missing.

[0650] "Imputation" is the operation of filling in missing data with appropriate values.

[0651] "Normalization" is a data conversion operation to standardize data with different scales or units.

[0652] A "machine learning model" is an algorithm that uses data to learn and make predictions and classifications.

[0653] "Training" is the process of feeding data to a machine learning model to improve its prediction and classification capabilities.

[0654] A "reward" is a gift or benefit given to a user when they take a specific action or fulfill a certain condition.

[0655] "Prediction" is the act of estimating and judging future events and outcomes based on past data.

[0656] "Export" is an operation to output data to the outside in a specific format.

[0657] A "user terminal" is a device such as a computer or smartphone operated by a user.

[0658] "Display" means visually showing information on the screen of a user's terminal.

[0659] To implement this invention, the server must first collect user data. Specifically, the server extracts the user's purchase history, genres of interest, and answers to questions from a database. The database used is a structured database such as MySQL or PostgreSQL.

[0660] The server combines the collected data into a single unified dataset, which can be done using a data management tool or library (e.g., the Pandas library in Python).

[0661] The server then preprocesses the integrated data. Specifically, it removes unnecessary columns, imputes missing values, and normalizes the data. For example, in the case of purchase history, unnecessary columns such as "purchase date" are removed, unanswered questions are imputed as "unanswered," and data is normalized to unify the units and scale of each item.

[0662] The server uses the preprocessed dataset to train a machine learning model. Algorithms used for this include random forests and support vector machines. The Python scikit-learn library is commonly used for training.

[0663] The server uses the trained model to predict suitable gifts for specific users. By specifying the user ID, the necessary features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0664] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0665] Illustrative prompt examples

[0666] Please recommend a gift based on the data for user ID 12345. The following data is available:

[0667] Purchase History: ['Books', 'Household Goods', 'Gadgets']

[0668] Interests: ['Sports', 'Cooking']

[0669] Answers to questions: { "Q1": "Yes", "Q2": "No", "Q3": "No"}

[0670] This allows for efficient collection and pre-processing of diverse user data, enabling highly accurate personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0671] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0672] Step 1:

[0673] Data collection

[0674] The server collects the user's purchase history, genre interests, and answers to questions from a database.

[0675] Input: User ID

[0676] Output: Dataset of purchase history, interest genres, and answers to questions

[0677] What happens: The server executes the SQL query to extract the required information from the database, and returns the results in the form of a DataFrame or a list.

[0678] Step 2:

[0679] Data integration

[0680] The server combines the collected data into a consolidated data set.

[0681] Input: Dataset of purchase history, genre interests, and answers to questions

[0682] Output: A consolidated dataset

[0683] Specific operation: The server uses Python's Pandas library to merge purchase history, interest genres, and answers to questions to generate a single data frame.

[0684] Step 3:

[0685] Data Preprocessing

[0686] The server removes unnecessary columns from the integrated data, imputes missing values, and normalizes the data.

[0687] Input: Integrated dataset

[0688] Output: Preprocessed dataset

[0689] Specific operation: The server uses Pandas to delete the "Purchase Date" column, complete unanswered items as "Unanswered," and normalize the data to unify the units and scale of each item.

[0690] Step 4:

[0691] Training a machine learning model

[0692] The server uses the preprocessed dataset to train a machine learning model.

[0693] Input: Preprocessed dataset

[0694] Output: A trained machine learning model

[0695] Specific operation: The server separates the data into training data and test data, and trains the model using the random forest algorithm in the scikit-learn library.

[0696] Step 5:

[0697] Gift predictions

[0698] The server uses the trained model to predict suitable gifts for a particular user.

[0699] Input: User ID, preprocessed dataset

[0700] Output: A list of recommended items

[0701] Specific operation: Based on the user ID, the server extracts the necessary features, inputs them into the model, and predicts the recommended items. For example, it might recommend "sports goods" or "cooking class tickets" to user ID 12345.

[0702] Step 6:

[0703] Submit a recommendation

[0704] The server sends a list of predicted gifts to the user's terminal.

[0705] Input: List of recommended items

[0706] Output: Data sent to the terminal

[0707] Specific operation: The server converts the list of recommended items into JSON format and sends it to the user's device.

[0708] Step 7:

[0709] Displaying recommended items

[0710] The terminal processes the received data and displays it to the user.

[0711] Input: Data sent from the server

[0712] Output: Recommended items displayed on the user's screen

[0713] Specific operation: The device parses the received JSON data and displays "Recommended gifts: sports merchandise, cooking class tickets" for user ID 12345 on the screen.

[0714] In this way, user data can be efficiently collected and preprocessed, and personalized gift recommendations can be made with high accuracy. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0715] (Application example 1)

[0716] 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."

[0717] In recent years, the use of online shopping sites has increased, leading to a growing demand for personalized recommendation systems for users. However, conventional recommendation systems have been unable to effectively utilize diverse user data, resulting in low recommendation accuracy and insufficient improvement in user satisfaction. Furthermore, there is a lack of a means to effectively display recommended items on the user's device, which reduces user convenience.

[0718] 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.

[0719] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user's terminal, means for displaying the predicted gift on the user's terminal, means for outputting a list of the predicted gift in JSON format, and means for providing a user interface for displaying the predicted gift on the user's terminal. This enables the integration and analysis of various user data to enable more accurate and personalized gift recommendations. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[0720] "User" means an individual or corporation that uses the online shopping site.

[0721] "Purchase history" refers to a record of past purchases of products and services made by a user on an online shopping site.

[0722] "Interest genre" is information about the categories of products and services in which a user is interested.

[0723] "Answers to questions" are data on answers given by users to specific questions.

[0724] "Data integration" is the process of combining multiple types of collected data into a single data set.

[0725] "Data preprocessing" is the process of removing unnecessary columns from the integrated data, filling in missing values, and normalizing the data.

[0726] A "machine learning model" is an algorithm that is trained with data to perform a specific task.

[0727] "Gift prediction" refers to using a machine learning model to predict the best gift for a specific user.

[0728] A "user terminal" is a device operated by a user (such as a smartphone, tablet, or PC).

[0729] "Export predicted gift list in JSON format" means exporting the list of items predicted by the machine learning model in JSON (JavaScript Object Notation) format.

[0730] A "user interface" is a screen or input means through which a user interacts with a digital system.

[0731] To implement this invention, a server first collects user data. This data includes the user's purchase history, genre interests, and answers to questions. This data is collected from user devices such as smartphones, tablets, and PCs and stored on the server. Examples of hardware used include servers, smartphones, and PCs. Examples of software used include Python, Pandas, Firebase, and JSON.

[0732] The server integrates the collected data into a single dataset. At this time, it deletes unnecessary columns (such as purchase date) and imputes missing values ​​as "unanswered." It also standardizes the units and scale of the data and normalizes it. This data preprocessing is performed using Python and Pandas.

[0733] The preprocessed data is used to train a machine learning model on the server. Here, an algorithm such as random forest is used. The data is divided into training data and test data, and the model is trained using the training data. The software used includes Python's Scikit-learn.

[0734] Next, the server uses the trained machine learning model to predict the appropriate gift for the user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. The prediction results are exported in JSON format and sent to the user's device.

[0735] The user receives the predicted gift list and displays it through a user interface, such as a smartphone app or web browser, using React Native and HTML / CSS / JavaScript.

[0736] To give a specific example, a prompt for user ID 12345 might be, "Please recommend gift items based on user ID 12345's past purchase history and genre interests." Processing based on this prompt might result in, for example, "Recommended gifts: sports merchandise, cooking class tickets" being displayed on the user's device.

[0737] This makes it possible to integrate and analyze a variety of user data to recommend more accurate and personalized thank-you gifts. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0738] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0739] Step 1:

[0740] The server collects the user's purchase history, genre interests, and answers to questions from the user's device. The collected data is stored on the server. The input is raw data sent from the user's device, and the output is integrated data stored on the server. Specifically, the smartphone app sends data to the Firebase database, and the server receives the data.

[0741] Step 2:

[0742] The data collected by the server is integrated. At this time, unnecessary columns (such as purchase date) are deleted and missing values ​​are imputed as "unanswered." Furthermore, the data units and scales are unified and normalized. The input is raw data stored on the server, and the output is preprocessed data. Specifically, the data frame is manipulated using Python's Pandas library to fill in missing values.

[0743] Step 3:

[0744] The server uses the preprocessed data to train a machine learning model. It splits the data into training and test data and trains the model using an algorithm such as random forest. The input is the preprocessed data, and the output is the trained machine learning model. Specifically, it uses Scikit-learn functions to split the data and train the model.

[0745] Step 4:

[0746] The server uses a trained machine learning model to predict suitable gifts for the user. By specifying the user ID, features are extracted and input into the model to predict recommended items. The input is the user ID and feature data, and the output is a list of recommended gifts. Specifically, the feature data is input into the trained model to obtain the predicted results.

[0747] Step 5:

[0748] The server exports the predicted gift list in JSON format and sends it to the user's device. The input is the list of recommended gifts, and the output is JSON data. Specifically, the data is converted to JSON format using the Python json library and sent via the API.

[0749] Step 6:

[0750] The JSON data received by the user's device is parsed and displayed on the user interface. The input is the JSON data sent from the server, and the output is a list of gifts that is displayed in a form that the user can see. Specifically, the data is parsed using a framework such as React Native and displayed on the screen.

[0751] By going through the above steps, a system is created that can recommend personalized gifts to users.

[0752] 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.

[0753] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0754] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0755] After preprocessing is complete, the server trains the machine learning model by dividing the user's data into features and targets, and splitting them into training data and test data. For example, it uses a random forest algorithm to train the model.

[0756] The server then uses the trained model to predict suitable gifts for a specific user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. For example, it might recommend "sports merchandise" or "tickets to a cooking class" to user ID 12345.

[0757] Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. The server uses the emotion engine to collect the user's emotion data (facial recognition data, text analysis data, etc.) and integrates it with their purchase history, genre interests, and answers to questions. This generates integrated data that also takes emotion data into account.

[0758] The server trains a machine learning model based on the integrated data, including emotional data. This model can predict gifts based on the user's emotions. For example, if a user has recently been feeling stressed, the model will enhance recommendations for relaxation items.

[0759] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format or similar and sent to the device. The device analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed on the screen.

[0760] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their tastes and emotions, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0761] The processing flow will be explained below.

[0762] Step 1:

[0763] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0764] Step 2:

[0765] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0766] Step 3:

[0767] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0768] Step 4:

[0769] The server collects emotion data using an emotion engine that recognizes the user's emotions. Specifically, it analyzes the user's emotions using facial recognition data, text analysis data, etc., and obtains information such as "happiness," "sadness," and "stress." For example, the emotion of user ID 12345 may be analyzed as "I've been feeling stressed recently."

[0770] Step 5:

[0771] The server integrates the collected purchase history, genre interests, answers to questions, and emotional data. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0772] Step 6:

[0773] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "no response." Finally, it normalizes data with different scales and converts them into a format suitable for analysis and machine learning. For example, it quantifies and handles emotional data.

[0774] Step 7:

[0775] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0776] Step 8:

[0777] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports goods," "cooking class tickets," and "relaxation goods" are recommended based on past data and emotional data.

[0778] Step 9:

[0779] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" are displayed on the screen. Based on this information, the user can select products and services that match their preferences and emotions.

[0780] Example 2

[0781] 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."

[0782] Modern e-commerce sites and service providers are required to recommend personalized gifts to users. However, existing methods rely only on data such as a user's purchase history, genre interests, and answers to questions, which limits recommendation accuracy and makes it difficult to respond flexibly to users' emotions and current situations. Furthermore, these methods face challenges such as missing data, inconsistent scale, and low-quality training data for machine learning models.

[0783] 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.

[0784] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting suitable items for the user using the trained machine learning model, means for collecting user emotion data and incorporating it into the integrated data, means for retraining the machine learning model using the emotion data, means for transmitting the predicted items to the user terminal, and means for displaying the predicted items on the user terminal. This makes it possible to integrate multifaceted information about the user and recommend personalized gifts with high accuracy.

[0785] "User purchase history" refers to all information about purchases made by a user in the past, specifically data including the purchase date and time, purchased items, quantity, and price.

[0786] "User interest genres" refers to categories or themes that the user is interested in, and specifically includes information on sports, music, cooking, etc.

[0787] "Answers to user questions" refers to the answers users give to surveys and question forms, and is data that reflects the user's preferences and opinions.

[0788] "Means of integrating collected data" refers to the process of combining data obtained from multiple disparate sources into a single integrated data set.

[0789] "Means for preprocessing integrated data" refers to processes such as filling in missing values ​​and deleting unnecessary data, which are carried out to improve the accuracy and quality of the data.

[0790] "Means of training a machine learning model" refers to the process of using collected data to train a machine learning algorithm and prepare it to make predictions on new data in the future.

[0791] "Means for predicting suitable products for a user" means a process that uses a trained machine learning model to predict the products that a user will like or need based on the characteristics of that user.

[0792] "User emotional data" refers to data that indicates the user's emotional state, and specifically includes emotional information obtained from facial expression recognition data, text analysis results, and the like.

[0793] "Means for collecting and incorporating emotional data into aggregated data" refers to the process of collecting data about user emotions and combining it with other user information into a single aggregated data set.

[0794] "Means for retraining a machine learning model with emotion data" refers to the process of retraining a machine learning model using a new dataset that includes emotion data to improve the accuracy of predictions that take emotion into account.

[0795] "Means of sending to user terminal" refers to the process of transferring data from the server to the user's device (such as a smartphone or PC) via a network.

[0796] "Means for displaying predicted items on a user device" refers to a mechanism for visually displaying predicted gifts or recommended products on a user device.

[0797] To implement this invention, the server first collects user data, such as the user's purchase history, genre interests, and answers to questions, from a database. This process involves retrieving the data using SQL queries and organizing it using the Pandas library.

[0798] Next, the server preprocesses the collected data. Preprocessing includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "Purchase Date" column is deleted from purchase history data, and any items with missing values ​​are filled in with a default value such as "No Answer." Normalization is also performed to unify the scale of the numerical data to the range of 0 to 1. These preprocessing steps are performed using data processing libraries such as Pandas and NumPy.

[0799] The server uses the preprocessed data to train a machine learning model. This process uses the Scikit-learn library. For example, the dataset is divided into features and targets, and then further divided into training data and test data. The model is trained using algorithms such as random forests, and then the model's applicability to new data is evaluated.

[0800] Using the trained model, the server predicts suitable items for the user. Specifically, features are extracted based on the specified user ID and input into the trained model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended.

[0801] The server also collects user emotional data and incorporates it into the integrated data. This process uses facial recognition and text analysis data. The collected emotional data is analyzed using an emotional engine and combined with purchase history, genre interests, and answers to questions. This creates a dataset that takes into account the user's emotions.

[0802] The server retrains the machine learning model using a new dataset that includes emotion data. For example, by calculating "stress levels" from facial recognition data and retraining the model using data that includes this, it becomes possible to make predictions that take user emotions into account.

[0803] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format and sent to the device. The user's device analyzes the received data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0804] This system allows users to easily find products and services that match their tastes and emotions, and also allows providers to improve user satisfaction.

[0805] Example prompt sentence:

[0806] "Generate a list of recommended gifts based on the user's recent interests and sentiment."

[0807] "Please recommend suitable items for user ID 12345, taking into account the sentiment data."

[0808] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0809] Step 1: Collect user data

[0810] The server extracts the user's purchase history, interest genres, and answers to questions from a database. The input is a query based on the user ID, and the output is a dataset containing purchase history data, interest genre data, and question-answer data. Specifically, it uses an SQL query to retrieve the required data and converts it into a Pandas data frame. For example, it uses the query SELECT FROM user_data WHERE user_id = '12345';

[0811] Step 2: Integrate the data

[0812] The server merges the collected data. The input is the multiple datasets collected in step 1, and the output is a single merged dataset. Specifically, it merges data frames using the Pandas library. For example, it performs an operation such as pd.merge(purchase_data, interest_data, on='user_id').

[0813] Step 3: Preprocessing the data

[0814] The server preprocesses the integrated data. The input is the integrated dataset, and the output is the preprocessed dataset. Preprocessing includes removing unnecessary columns, imputing missing values, and normalizing the data. Specifically, it uses Pandas to remove unnecessary columns and impute missing values. For example, it performs operations such as df.drop(columns=['purchase_date']) and df.fillna('Unanswered').

[0815] Step 4: Train the machine learning model

[0816] The server trains a machine learning model using the preprocessed data. The input is the preprocessed dataset, and the output is a trained machine learning model. Specifically, a random forest model is trained using the Scikit-learn library. For example, the data is split using the train_test_split function, and the model is trained using RandomForestClassifier.

[0817] Step 5: Predicting recommended items

[0818] The server uses the trained model to predict suitable items for a user. The input is the features of a specific user, and the output is a list of recommended items. Specifically, the features are input into the model, and the prediction result is obtained using the predict function. For example, "sports goods" and "tickets to a cooking class" are recommended for user ID 12345.

[0819] Step 6: Collect and integrate emotion data

[0820] The server collects user emotion data (e.g., facial recognition data, text analysis data) and incorporates it into integrated data. The input is emotion data, and the output is an integrated dataset including emotion data. Specifically, face recognition is performed using the OpenCV library, and analysis is performed using the emotion engine. For example, face information is acquired using cv2.VideoCapture, and the "stress level" is calculated using the emotion engine.

[0821] Step 7: Retrain the model with emotion data

[0822] The server retrains the machine learning model using a new dataset containing emotion data. The input is the integrated dataset containing emotion data, and the output is the retrained model. Specifically, the server again splits the data into training and test data using the Scikit-learn library, and then retrains the model.

[0823] Step 8: Notification of recommended items

[0824] The server sends the predicted gifts to the user's device. The input is a list of recommended items, and the output is the data sent to the user's device. Specifically, Flask or Django is used to export the data in JSON format. The user's device parses this data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[0825] (Application example 2)

[0826] 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."

[0827] Traditional recommendation systems only consider a user's purchase history and genre interests, and are unable to reflect the user's emotions and momentary state of mind, thereby failing to maximize the user experience. Furthermore, due to insufficient data preprocessing and missing value imputation, recommendation accuracy is low. Furthermore, traditional systems also suffer from inefficient data integration and machine learning model training.

[0828] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0829] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for collecting the user's emotional data, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user terminal, and means for displaying the predicted gift on the user terminal. This allows for more personalized recommendations that reflect the user's emotions and state of mind. Furthermore, data accuracy can be improved by filling in missing values ​​and normalizing them, thereby improving the accuracy of recommendations.

[0830] "Means for collecting user purchase history" refers to a method for retrieving and storing information about products and services that users have previously purchased from a database.

[0831] "Means for collecting user interest genres" refers to methods for identifying categories or topics that users are interested in and collecting that information.

[0832] A "means for collecting answers to user questions" is a method for collecting answers to questions posed to users and storing them in a database.

[0833] "Means for collecting user emotional data" refers to a method for analyzing emotions that can be read from a user's facial expressions and text, and collecting this as data.

[0834] "Means for integrating collected data" refers to a method for integrating information collected from different data sources, such as purchase history, genre interests, question answers, and emotional data, into a single dataset.

[0835] "Means for preprocessing the integrated data" refers to methods for organizing the data by removing unnecessary columns from the integrated dataset, imputing missing values, and normalizing the data.

[0836] A "means for training a machine learning model using preprocessed data" is a method for providing preprocessed data as input to a machine learning algorithm to train a predictive model.

[0837] "Means for predicting suitable gifts for users using a trained machine learning model" means a method for using a trained machine learning model to predict the best product or service for a particular user.

[0838] "Means for sending predicted gifts to a user's device" refers to a method for sending information about products or services predicted by a machine learning model to a user's device.

[0839] The "means for displaying predicted gifts on the user's device" is a method for displaying the transmitted list of predicted goods and services on the user's device.

[0840] This invention is a system that collects and integrates a user's purchase history, genre interests, answers to questions, and emotional data to recommend highly accurate personalized gifts. This system uses specific means to execute each step, with the aim of improving the user experience. Each means and its implementation method will be explained in detail below.

[0841] First, the server collects user data, including information about products and services the user has purchased in the past, the user's interests, questions the user has answered, and emotional data obtained through facial recognition and text analysis. This data is stored in a database and integrated as needed.

[0842] Next, the server integrates the collected data. The collected purchase history, genre interests, answers to questions, and sentiment data are compiled into a single integrated dataset. This integration is performed using Python's Pandas library, among other tools. The integrated data is then preprocessed, which involves removing unnecessary columns, filling in missing values, and normalizing the data.

[0843] The server uses the preprocessed data to train a machine learning model. For this training, the scikit-learn library is used, applying the random forest algorithm. The data is split into training and test data, and the accuracy of the model is verified as it progresses.

[0844] Once trained, the machine learning model is used to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model. For example, recommendations can be made for user ID 12345, such as "sports goods," "tickets to a cooking class," and "relaxation goods."

[0845] Furthermore, the server can make recommendations that take into account the user's emotions based on the integrated data, including emotional data. For example, if the emotional data indicates that the user has recently been feeling stressed, the server will strengthen recommendations for relaxation products.

[0846] The predicted gift is sent from the server to the user's device and displayed on the device. The data is exported in JSON format and analyzed on the user's smartphone or other device.

[0847] For example, if a user has frequently purchased fitness equipment in the past six months and has recently experienced increased stress, the system will recommend relaxation products, yoga mats, stretching equipment, etc.

[0848] Examples of prompts to input to a generative AI model include:

[0849] "Please use user ID 12345's past purchase history, genre interests, question responses, and sentiment data to predict personalized product recommendations."

[0850] In this way, the system can integrate and analyze a variety of user data to recommend more accurate and personalized gifts, benefiting both users and gift givers.

[0851] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0852] Step 1:

[0853] The server collects the user's data.

[0854] Specifically, the server retrieves user purchase history, genre interests, answers to questions, and emotional data obtained through facial recognition and text analysis from the database. These data are read from separate tables or files and stored in a unified format.

[0855] Input: Raw data from databases and various APIs

[0856] Output: User data stored in a unified format

[0857] Step 2:

[0858] The server consolidates the collected data.

[0859] The collected purchase history, interest genres, question answers, and sentiment data are merged into a single data frame. This integration is performed using the Python Pandas library.

[0860] Input: User data collected in step 1

[0861] Output: A merged data frame

[0862] Step 3:

[0863] The server preprocesses the consolidated data.

[0864] Preprocessing involves deleting unnecessary columns, imputing missing values, and normalizing the data. Specifically, the "Purchase Date" column is deleted, NULL values ​​are imputed as "No Answer," and the units and scales of each data are unified.

[0865] Input: Merged data frame

[0866] Output: Preprocessed data frame

[0867] Step 4:

[0868] The server uses the preprocessed data to train a machine learning model.

[0869] The preprocessed data is divided into features and targets, and then split into training and test data. A model is trained using the random forest algorithm in scikit-learn.

[0870] Input: Preprocessed data frame

[0871] Output: A trained machine learning model

[0872] Step 5:

[0873] The server uses a trained machine learning model to predict suitable gifts for users.

[0874] By extracting features based on a specific user ID and inputting them into an existing model, the system predicts the recommended gift.

[0875] Input: User ID and preprocessed data frame

[0876] Output: A list of recommended gifts

[0877] Step 6:

[0878] The server sends the predicted gift to the user's device.

[0879] Export the list of recommended items in JSON format or other format and send it to the user's smartphone or other device.

[0880] Input: List of recommended gifts

[0881] Output: Data exported in JSON format

[0882] Step 7:

[0883] The terminal displays the predicted gift received from the server.

[0884] The received JSON data is analyzed and displayed on the screen for the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed.

[0885] Input: JSON data sent from the server

[0886] Output: A list of gifts displayed on the screen

[0887] 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.

[0888] 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.

[0889] 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.

[0890] [Fourth embodiment]

[0891] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0892] 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.

[0893] 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).

[0894] 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.

[0895] 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.

[0896] 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).

[0897] 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.

[0898] 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.

[0899] 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.

[0900] 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.

[0901] 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.

[0902] 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.

[0903] 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."

[0904] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[0905] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[0906] After preprocessing is complete, the server trains the machine learning model. It separates the user's data into features and targets, and splits them into training data and test data. It uses the training data to train a model using a machine learning algorithm such as random forest. The trained model is then used to predict rewards for the user.

[0907] The server then uses the trained model to predict suitable gifts for specific users. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0908] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. Specifically, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0909] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[0910] The processing flow will be explained below.

[0911] Step 1:

[0912] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[0913] Step 2:

[0914] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[0915] Step 3:

[0916] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[0917] Step 4:

[0918] The server integrates the collected purchase history, genre interests, and answers to questions. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[0919] Step 5:

[0920] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "No response." Finally, it normalizes data with different scales and converts it into a format suitable for analysis and machine learning.

[0921] Step 6:

[0922] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[0923] Step 7:

[0924] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, it predicts "sports goods" or "tickets to a cooking class."

[0925] Step 8:

[0926] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0927] Example 1

[0928] 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."

[0929] Conventional recommendation systems struggle to collect, integrate, and analyze diverse user data, making it difficult to provide accurate, personalized rewards to individual users. Furthermore, the accuracy of machine learning models was hindered by insufficient data imputation and normalization. This resulted in failures to increase user satisfaction and to provide optimal services to providers.

[0930] 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.

[0931] In this invention, the server includes means for collecting a user's purchase history from a database, means for collecting the user's genre interests from the database, means for collecting answers to the user's questions from the database, means for combining the collected data into an integrated dataset, means for deleting unnecessary columns from the integrated data, means for imputing missing values, means for normalizing the data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for exporting a list of predicted gifts and sending it to a user terminal, and means for displaying the predicted gifts on the user terminal. This enables efficient collection and preprocessing of a variety of user data and highly accurate personalized gift recommendations.

[0932] A "database" is a system that structures and stores information so that it can be easily retrieved and manipulated.

[0933] "User purchase history" refers to data about products and services purchased by a user in the past.

[0934] "User interest genres" is information about the fields and categories in which a user is interested.

[0935] An "answer to a question" is answer data selected or entered by a user in response to a specific question.

[0936] An "integrated dataset" is one in which data collected from multiple different data sources is combined into one dataset.

[0937] An "unnecessary column" is a data item that is not essential for analysis or processing.

[0938] A "missing value" is a location where data that should be present is missing.

[0939] "Imputation" is the operation of filling in missing data with appropriate values.

[0940] "Normalization" is a data conversion operation to standardize data with different scales or units.

[0941] A "machine learning model" is an algorithm that uses data to learn and make predictions and classifications.

[0942] "Training" is the process of feeding data to a machine learning model to improve its prediction and classification capabilities.

[0943] A "reward" is a gift or benefit given to a user when they take a specific action or fulfill a certain condition.

[0944] "Prediction" is the act of estimating and judging future events and outcomes based on past data.

[0945] "Export" is an operation to output data to the outside in a specific format.

[0946] A "user terminal" is a device such as a computer or smartphone operated by a user.

[0947] "Display" means visually showing information on the screen of a user's terminal.

[0948] To implement this invention, the server must first collect user data. Specifically, the server extracts the user's purchase history, genres of interest, and answers to questions from a database. The database used is a structured database such as MySQL or PostgreSQL.

[0949] The server combines the collected data into a single unified dataset, which can be done using a data management tool or library (e.g., the Pandas library in Python).

[0950] The server then preprocesses the integrated data. Specifically, it removes unnecessary columns, imputes missing values, and normalizes the data. For example, in the case of purchase history, unnecessary columns such as "purchase date" are removed, unanswered questions are imputed as "unanswered," and data is normalized to unify the units and scale of each item.

[0951] The server uses the preprocessed dataset to train a machine learning model. Algorithms used for this include random forests and support vector machines. The Python scikit-learn library is commonly used for training.

[0952] The server uses the trained model to predict suitable gifts for specific users. By specifying the user ID, the necessary features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended based on past purchase history, genre interests, and answers to questions.

[0953] The prediction results are sent from the server to the device. The list of recommended items is exported in JSON format or similar and sent to the device. The device processes the received data and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports merchandise, cooking class tickets" is displayed on the screen.

[0954] Illustrative prompt examples

[0955] Please recommend a gift based on the data for user ID 12345. The following data is available:

[0956] Purchase History: ['Books', 'Household Goods', 'Gadgets']

[0957] Interests: ['Sports', 'Cooking']

[0958] Answers to questions: { "Q1": "Yes", "Q2": "No", "Q3": "No"}

[0959] This allows for efficient collection and pre-processing of diverse user data, enabling highly accurate personalized gift recommendations. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[0960] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0961] Step 1:

[0962] Data collection

[0963] The server collects the user's purchase history, genre interests, and answers to questions from a database.

[0964] Input: User ID

[0965] Output: Dataset of purchase history, interest genres, and answers to questions

[0966] What happens: The server executes the SQL query to extract the required information from the database, and returns the results in the form of a DataFrame or a list.

[0967] Step 2:

[0968] Data integration

[0969] The server combines the collected data into a consolidated data set.

[0970] Input: Dataset of purchase history, genre interests, and answers to questions

[0971] Output: A consolidated dataset

[0972] Specific operation: The server uses Python's Pandas library to merge purchase history, interest genres, and answers to questions to generate a single data frame.

[0973] Step 3:

[0974] Data Preprocessing

[0975] The server removes unnecessary columns from the integrated data, imputes missing values, and normalizes the data.

[0976] Input: Integrated dataset

[0977] Output: Preprocessed dataset

[0978] Specific operation: The server uses Pandas to delete the "Purchase Date" column, complete unanswered items as "Unanswered," and normalize the data to unify the units and scale of each item.

[0979] Step 4:

[0980] Training a machine learning model

[0981] The server uses the preprocessed dataset to train a machine learning model.

[0982] Input: Preprocessed dataset

[0983] Output: A trained machine learning model

[0984] Specific operation: The server separates the data into training data and test data, and trains the model using the random forest algorithm in the scikit-learn library.

[0985] Step 5:

[0986] Gift predictions

[0987] The server uses the trained model to predict suitable gifts for a particular user.

[0988] Input: User ID, preprocessed dataset

[0989] Output: A list of recommended items

[0990] Specific operation: Based on the user ID, the server extracts the necessary features, inputs them into the model, and predicts the recommended items. For example, it might recommend "sports goods" or "cooking class tickets" to user ID 12345.

[0991] Step 6:

[0992] Submit a recommendation

[0993] The server sends a list of predicted gifts to the user's terminal.

[0994] Input: List of recommended items

[0995] Output: Data sent to the terminal

[0996] Specific operation: The server converts the list of recommended items into JSON format and sends it to the user's device.

[0997] Step 7:

[0998] Displaying recommended items

[0999] The terminal processes the received data and displays it to the user.

[1000] Input: Data sent from the server

[1001] Output: Recommended items displayed on the user's screen

[1002] Specific operation: The device parses the received JSON data and displays "Recommended gifts: sports merchandise, cooking class tickets" for user ID 12345 on the screen.

[1003] In this way, user data can be efficiently collected and preprocessed, and personalized gift recommendations can be made with high accuracy. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[1004] (Application example 1)

[1005] 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."

[1006] In recent years, the use of online shopping sites has increased, leading to a growing demand for personalized recommendation systems for users. However, conventional recommendation systems have been unable to effectively utilize diverse user data, resulting in low recommendation accuracy and insufficient improvement in user satisfaction. Furthermore, there is a lack of a means to effectively display recommended items on the user's device, which reduces user convenience.

[1007] 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.

[1008] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user's terminal, means for displaying the predicted gift on the user's terminal, means for outputting a list of the predicted gift in JSON format, and means for providing a user interface for displaying the predicted gift on the user's terminal. This enables the integration and analysis of various user data to enable more accurate and personalized gift recommendations. Users can easily find products and services that suit their preferences, and providers can also improve user satisfaction.

[1009] "User" means an individual or corporation that uses the online shopping site.

[1010] "Purchase history" refers to a record of past purchases of products and services made by a user on an online shopping site.

[1011] "Interest genre" is information about the categories of products and services in which a user is interested.

[1012] "Answers to questions" are data on answers given by users to specific questions.

[1013] "Data integration" is the process of combining multiple types of collected data into a single data set.

[1014] "Data preprocessing" is the process of removing unnecessary columns from the integrated data, filling in missing values, and normalizing the data.

[1015] A "machine learning model" is an algorithm that is trained with data to perform a specific task.

[1016] "Gift prediction" refers to using a machine learning model to predict the best gift for a specific user.

[1017] A "user terminal" is a device operated by a user (such as a smartphone, tablet, or PC).

[1018] "Export predicted gift list in JSON format" means exporting the list of items predicted by the machine learning model in JSON (JavaScript Object Notation) format.

[1019] A "user interface" is a screen or input means through which a user interacts with a digital system.

[1020] To implement this invention, a server first collects user data. This data includes the user's purchase history, genre interests, and answers to questions. This data is collected from user devices such as smartphones, tablets, and PCs and stored on the server. Examples of hardware used include servers, smartphones, and PCs. Examples of software used include Python, Pandas, Firebase, and JSON.

[1021] The server integrates the collected data into a single dataset. At this time, it deletes unnecessary columns (such as purchase date) and imputes missing values ​​as "unanswered." It also standardizes the units and scale of the data and normalizes it. This data preprocessing is performed using Python and Pandas.

[1022] The preprocessed data is used to train a machine learning model on the server. Here, an algorithm such as random forest is used. The data is divided into training data and test data, and the model is trained using the training data. The software used includes Python's Scikit-learn.

[1023] Next, the server uses the trained machine learning model to predict the appropriate gift for the user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. The prediction results are exported in JSON format and sent to the user's device.

[1024] The user receives the predicted gift list and displays it through a user interface, such as a smartphone app or web browser, using React Native and HTML / CSS / JavaScript.

[1025] To give a specific example, a prompt for user ID 12345 might be, "Please recommend gift items based on user ID 12345's past purchase history and genre interests." Processing based on this prompt might result in, for example, "Recommended gifts: sports merchandise, cooking class tickets" being displayed on the user's device.

[1026] This makes it possible to integrate and analyze a variety of user data to recommend more accurate and personalized thank-you gifts. Users can easily find products and services that match their preferences, and providers can also improve user satisfaction.

[1027] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1028] Step 1:

[1029] The server collects the user's purchase history, genre interests, and answers to questions from the user's device. The collected data is stored on the server. The input is raw data sent from the user's device, and the output is integrated data stored on the server. Specifically, the smartphone app sends data to the Firebase database, and the server receives the data.

[1030] Step 2:

[1031] The data collected by the server is integrated. At this time, unnecessary columns (such as purchase date) are deleted and missing values ​​are imputed as "unanswered." Furthermore, the data units and scales are unified and normalized. The input is raw data stored on the server, and the output is preprocessed data. Specifically, the data frame is manipulated using Python's Pandas library to fill in missing values.

[1032] Step 3:

[1033] The server uses the preprocessed data to train a machine learning model. It splits the data into training and test data and trains the model using an algorithm such as random forest. The input is the preprocessed data, and the output is the trained machine learning model. Specifically, it uses Scikit-learn functions to split the data and train the model.

[1034] Step 4:

[1035] The server uses a trained machine learning model to predict suitable gifts for the user. By specifying the user ID, features are extracted and input into the model to predict recommended items. The input is the user ID and feature data, and the output is a list of recommended gifts. Specifically, the feature data is input into the trained model to obtain the predicted results.

[1036] Step 5:

[1037] The server exports the predicted gift list in JSON format and sends it to the user's device. The input is the list of recommended gifts, and the output is JSON data. Specifically, the data is converted to JSON format using the Python json library and sent via the API.

[1038] Step 6:

[1039] The JSON data received by the user's device is parsed and displayed on the user interface. The input is the JSON data sent from the server, and the output is a list of gifts that is displayed in a form that the user can see. Specifically, the data is parsed using a framework such as React Native and displayed on the screen.

[1040] By going through the above steps, a system is created that can recommend personalized gifts to users.

[1041] 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.

[1042] To implement this invention, the server first collects user data, including the user's purchase history, genre interests, and answers to questions. Specifically, this information stored in the database is extracted and combined to generate integrated data.

[1043] Next, the server performs preprocessing on the data. This includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "purchase date" in the purchase history is deleted, and unanswered questions are marked as "unanswered." The data is also normalized by standardizing its units and scale.

[1044] After preprocessing is complete, the server trains the machine learning model by dividing the user's data into features and targets, and splitting them into training data and test data. For example, it uses a random forest algorithm to train the model.

[1045] The server then uses the trained model to predict suitable gifts for a specific user. It extracts features by specifying the user ID and inputs them into the model to predict recommended items. For example, it might recommend "sports merchandise" or "tickets to a cooking class" to user ID 12345.

[1046] Furthermore, this system is combined with an emotion engine that recognizes the user's emotions. The server uses the emotion engine to collect the user's emotion data (facial recognition data, text analysis data, etc.) and integrates it with their purchase history, genre interests, and answers to questions. This generates integrated data that also takes emotion data into account.

[1047] The server trains a machine learning model based on the integrated data, including emotional data. This model can predict gifts based on the user's emotions. For example, if a user has recently been feeling stressed, the model will enhance recommendations for relaxation items.

[1048] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format or similar and sent to the device. The device analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed on the screen.

[1049] This system integrates and analyzes a variety of user data, enabling more accurate and personalized gift recommendations. Users can easily find products and services that match their tastes and emotions, and providers can also improve user satisfaction. In this way, a system that benefits both users and providers can be realized.

[1050] The processing flow will be explained below.

[1051] Step 1:

[1052] The server connects to the database and collects the user's purchase history. Specifically, it extracts information such as the user ID, purchased items, and purchase date and time. For example, it obtains "item A, item B" as the purchase history for user ID 12345.

[1053] Step 2:

[1054] The server collects the user's interest genres. Specifically, it retrieves information about the categories and fields of interest from the database. For example, it retrieves "sports, cooking" as the interest genres for user ID 12345.

[1055] Step 3:

[1056] The server collects answers to the user's questions. Specifically, it retrieves answers to past surveys and feedback from a database. For example, it retrieves the answer "I'm interested in Japanese food" from user ID 12345.

[1057] Step 4:

[1058] The server collects emotion data using an emotion engine that recognizes the user's emotions. Specifically, it analyzes the user's emotions using facial recognition data, text analysis data, etc., and obtains information such as "happiness," "sadness," and "stress." For example, the emotion of user ID 12345 may be analyzed as "I've been feeling stressed recently."

[1059] Step 5:

[1060] The server integrates the collected purchase history, genre interests, answers to questions, and emotional data. Specifically, this information is linked by a common user ID and compiled into a single data set. This allows for integrated data for each user ID.

[1061] Step 6:

[1062] The server preprocesses the integrated data. First, it removes unnecessary columns (e.g., purchase date and time). Next, it imputes missing values ​​with appropriate values ​​such as "no response." Finally, it normalizes data with different scales and converts them into a format suitable for analysis and machine learning. For example, it quantifies and handles emotional data.

[1063] Step 7:

[1064] The server trains a machine learning model using the preprocessed data. It separates the user data into features and targets (gifts), and splits them into training data and test data. For example, it trains the model using a random forest algorithm.

[1065] Step 8:

[1066] The server uses a trained model to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model to predict recommended items. For example, for user ID 12345, "sports goods," "cooking class tickets," and "relaxation goods" are recommended based on past data and emotional data.

[1067] Step 9:

[1068] The server exports the prediction results in JSON format or similar and sends them to the user's device. The device then analyzes the data received from the server and displays it to the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" are displayed on the screen. Based on this information, the user can select products and services that match their preferences and emotions.

[1069] Example 2

[1070] 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."

[1071] Modern e-commerce sites and service providers are required to recommend personalized gifts to users. However, existing methods rely only on data such as a user's purchase history, genre interests, and answers to questions, which limits recommendation accuracy and makes it difficult to respond flexibly to users' emotions and current situations. Furthermore, these methods face challenges such as missing data, inconsistent scale, and low-quality training data for machine learning models.

[1072] 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.

[1073] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting suitable items for the user using the trained machine learning model, means for collecting user emotion data and incorporating it into the integrated data, means for retraining the machine learning model using the emotion data, means for transmitting the predicted items to the user terminal, and means for displaying the predicted items on the user terminal. This makes it possible to integrate multifaceted information about the user and recommend personalized gifts with high accuracy.

[1074] "User purchase history" refers to all information about purchases made by a user in the past, specifically data including the purchase date and time, purchased items, quantity, and price.

[1075] "User interest genres" refers to categories or themes that the user is interested in, and specifically includes information on sports, music, cooking, etc.

[1076] "Answers to user questions" refers to the answers users give to surveys and question forms, and is data that reflects the user's preferences and opinions.

[1077] "Means of integrating collected data" refers to the process of combining data obtained from multiple disparate sources into a single integrated data set.

[1078] "Means for preprocessing integrated data" refers to processes such as filling in missing values ​​and deleting unnecessary data, which are carried out to improve the accuracy and quality of the data.

[1079] "Means of training a machine learning model" refers to the process of using collected data to train a machine learning algorithm and prepare it to make predictions on new data in the future.

[1080] "Means for predicting suitable products for a user" means a process that uses a trained machine learning model to predict the products that a user will like or need based on the characteristics of that user.

[1081] "User emotional data" refers to data that indicates the user's emotional state, and specifically includes emotional information obtained from facial expression recognition data, text analysis results, and the like.

[1082] "Means for collecting and incorporating emotional data into aggregated data" refers to the process of collecting data about user emotions and combining it with other user information into a single aggregated data set.

[1083] "Means for retraining a machine learning model with emotion data" refers to the process of retraining a machine learning model using a new dataset that includes emotion data to improve the accuracy of predictions that take emotion into account.

[1084] "Means of sending to user terminal" refers to the process of transferring data from the server to the user's device (such as a smartphone or PC) via a network.

[1085] "Means for displaying predicted items on a user device" refers to a mechanism for visually displaying predicted gifts or recommended products on a user device.

[1086] To implement this invention, the server first collects user data, such as the user's purchase history, genre interests, and answers to questions, from a database. This process involves retrieving the data using SQL queries and organizing it using the Pandas library.

[1087] Next, the server preprocesses the collected data. Preprocessing includes deleting unnecessary columns, filling in missing values, and normalizing the data. For example, the "Purchase Date" column is deleted from purchase history data, and any items with missing values ​​are filled in with a default value such as "No Answer." Normalization is also performed to unify the scale of the numerical data to the range of 0 to 1. These preprocessing steps are performed using data processing libraries such as Pandas and NumPy.

[1088] The server uses the preprocessed data to train a machine learning model. This process uses the Scikit-learn library. For example, the dataset is divided into features and targets, and then further divided into training data and test data. The model is trained using algorithms such as random forests, and then the model's applicability to new data is evaluated.

[1089] Using the trained model, the server predicts suitable items for the user. Specifically, features are extracted based on the specified user ID and input into the trained model to predict recommended items. For example, for user ID 12345, "sports-related goods" and "tickets to a cooking class" are recommended.

[1090] The server also collects user emotional data and incorporates it into the integrated data. This process uses facial recognition and text analysis data. The collected emotional data is analyzed using an emotional engine and combined with purchase history, genre interests, and answers to questions. This creates a dataset that takes into account the user's emotions.

[1091] The server retrains the machine learning model using a new dataset that includes emotion data. For example, by calculating "stress levels" from facial recognition data and retraining the model using data that includes this, it becomes possible to make predictions that take user emotions into account.

[1092] The predicted gifts are sent from the server to the user's device. The list of recommended items is exported in JSON format and sent to the device. The user's device analyzes the received data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[1093] This system allows users to easily find products and services that match their tastes and emotions, and also allows providers to improve user satisfaction.

[1094] Example prompt sentence:

[1095] "Generate a list of recommended gifts based on the user's recent interests and sentiment."

[1096] "Please recommend suitable items for user ID 12345, taking into account the sentiment data."

[1097] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1098] Step 1: Collect user data

[1099] The server extracts the user's purchase history, interest genres, and answers to questions from a database. The input is a query based on the user ID, and the output is a dataset containing purchase history data, interest genre data, and question-answer data. Specifically, it uses an SQL query to retrieve the required data and converts it into a Pandas data frame. For example, it uses the query SELECT FROM user_data WHERE user_id = '12345';

[1100] Step 2: Integrate the data

[1101] The server merges the collected data. The input is the multiple datasets collected in step 1, and the output is a single merged dataset. Specifically, it merges data frames using the Pandas library. For example, it performs an operation such as pd.merge(purchase_data, interest_data, on='user_id').

[1102] Step 3: Preprocessing the data

[1103] The server preprocesses the integrated data. The input is the integrated dataset, and the output is the preprocessed dataset. Preprocessing includes removing unnecessary columns, imputing missing values, and normalizing the data. Specifically, it uses Pandas to remove unnecessary columns and impute missing values. For example, it performs operations such as df.drop(columns=['purchase_date']) and df.fillna('Unanswered').

[1104] Step 4: Train the machine learning model

[1105] The server trains a machine learning model using the preprocessed data. The input is the preprocessed dataset, and the output is a trained machine learning model. Specifically, a random forest model is trained using the Scikit-learn library. For example, the data is split using the train_test_split function, and the model is trained using RandomForestClassifier.

[1106] Step 5: Predicting recommended items

[1107] The server uses the trained model to predict suitable items for a user. The input is the features of a specific user, and the output is a list of recommended items. Specifically, the features are input into the model, and the prediction result is obtained using the predict function. For example, "sports goods" and "tickets to a cooking class" are recommended for user ID 12345.

[1108] Step 6: Collect and integrate emotion data

[1109] The server collects user emotion data (e.g., facial recognition data, text analysis data) and incorporates it into integrated data. The input is emotion data, and the output is an integrated dataset including emotion data. Specifically, face recognition is performed using the OpenCV library, and analysis is performed using the emotion engine. For example, face information is acquired using cv2.VideoCapture, and the "stress level" is calculated using the emotion engine.

[1110] Step 7: Retrain the model with emotion data

[1111] The server retrains the machine learning model using a new dataset containing emotion data. The input is the integrated dataset containing emotion data, and the output is the retrained model. Specifically, the server again splits the data into training and test data using the Scikit-learn library, and then retrains the model.

[1112] Step 8: Notification of recommended items

[1113] The server sends the predicted gifts to the user's device. The input is a list of recommended items, and the output is the data sent to the user's device. Specifically, Flask or Django is used to export the data in JSON format. The user's device parses this data and displays it on the screen. For example, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed for user ID 12345.

[1114] (Application example 2)

[1115] 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."

[1116] Traditional recommendation systems only consider a user's purchase history and genre interests, and are unable to reflect the user's emotions and momentary state of mind, thereby failing to maximize the user experience. Furthermore, due to insufficient data preprocessing and missing value imputation, recommendation accuracy is low. Furthermore, traditional systems also suffer from inefficient data integration and machine learning model training.

[1117] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1118] In this invention, the server includes means for collecting a user's purchase history, means for collecting the user's genres of interest, means for collecting answers to the user's questions, means for collecting the user's emotional data, means for integrating the collected data, means for preprocessing the integrated data, means for training a machine learning model using the preprocessed data, means for predicting a gift suitable for the user using the trained machine learning model, means for sending the predicted gift to the user terminal, and means for displaying the predicted gift on the user terminal. This allows for more personalized recommendations that reflect the user's emotions and state of mind. Furthermore, data accuracy can be improved by filling in missing values ​​and normalizing them, thereby improving the accuracy of recommendations.

[1119] "Means for collecting user purchase history" refers to a method for retrieving and storing information about products and services that users have previously purchased from a database.

[1120] "Means for collecting user interest genres" refers to methods for identifying categories or topics that users are interested in and collecting that information.

[1121] A "means for collecting answers to user questions" is a method for collecting answers to questions posed to users and storing them in a database.

[1122] "Means for collecting user emotional data" refers to a method for analyzing emotions that can be read from a user's facial expressions and text, and collecting this as data.

[1123] "Means for integrating collected data" refers to a method for integrating information collected from different data sources, such as purchase history, genre interests, question answers, and emotional data, into a single dataset.

[1124] "Means for preprocessing the integrated data" refers to methods for organizing the data by removing unnecessary columns from the integrated dataset, imputing missing values, and normalizing the data.

[1125] A "means for training a machine learning model using preprocessed data" is a method for providing preprocessed data as input to a machine learning algorithm to train a predictive model.

[1126] "Means for predicting suitable gifts for users using a trained machine learning model" means a method for using a trained machine learning model to predict the best product or service for a particular user.

[1127] "Means for sending predicted gifts to a user's device" refers to a method for sending information about products or services predicted by a machine learning model to a user's device.

[1128] The "means for displaying predicted gifts on the user's device" is a method for displaying the transmitted list of predicted goods and services on the user's device.

[1129] This invention is a system that collects and integrates a user's purchase history, genre interests, answers to questions, and emotional data to recommend highly accurate personalized gifts. This system uses specific means to execute each step, with the aim of improving the user experience. Each means and its implementation method will be explained in detail below.

[1130] First, the server collects user data, including information about products and services the user has purchased in the past, the user's interests, questions the user has answered, and emotional data obtained through facial recognition and text analysis. This data is stored in a database and integrated as needed.

[1131] Next, the server integrates the collected data. The collected purchase history, genre interests, answers to questions, and sentiment data are compiled into a single integrated dataset. This integration is performed using Python's Pandas library, among other tools. The integrated data is then preprocessed, which involves removing unnecessary columns, filling in missing values, and normalizing the data.

[1132] The server uses the preprocessed data to train a machine learning model. For this training, the scikit-learn library is used, applying the random forest algorithm. The data is split into training and test data, and the accuracy of the model is verified as it progresses.

[1133] Once trained, the machine learning model is used to predict suitable gifts for a specific user. By specifying the user ID, features are extracted and input into the model. For example, recommendations can be made for user ID 12345, such as "sports goods," "tickets to a cooking class," and "relaxation goods."

[1134] Furthermore, the server can make recommendations that take into account the user's emotions based on the integrated data, including emotional data. For example, if the emotional data indicates that the user has recently been feeling stressed, the server will strengthen recommendations for relaxation products.

[1135] The predicted gift is sent from the server to the user's device and displayed on the device. The data is exported in JSON format and analyzed on the user's smartphone or other device.

[1136] For example, if a user has frequently purchased fitness equipment in the past six months and has recently experienced increased stress, the system will recommend relaxation products, yoga mats, stretching equipment, etc.

[1137] Examples of prompts to input to a generative AI model include:

[1138] "Please use user ID 12345's past purchase history, genre interests, question responses, and sentiment data to predict personalized product recommendations."

[1139] In this way, the system can integrate and analyze a variety of user data to recommend more accurate and personalized gifts, benefiting both users and gift givers.

[1140] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1141] Step 1:

[1142] The server collects the user's data.

[1143] Specifically, the server retrieves user purchase history, genre interests, answers to questions, and emotional data obtained through facial recognition and text analysis from the database. These data are read from separate tables or files and stored in a unified format.

[1144] Input: Raw data from databases and various APIs

[1145] Output: User data stored in a unified format

[1146] Step 2:

[1147] The server consolidates the collected data.

[1148] The collected purchase history, interest genres, question answers, and sentiment data are merged into a single data frame. This integration is performed using the Python Pandas library.

[1149] Input: User data collected in step 1

[1150] Output: A merged data frame

[1151] Step 3:

[1152] The server preprocesses the consolidated data.

[1153] Preprocessing involves deleting unnecessary columns, imputing missing values, and normalizing the data. Specifically, the "Purchase Date" column is deleted, NULL values ​​are imputed as "No Answer," and the units and scales of each data are unified.

[1154] Input: Merged data frame

[1155] Output: Preprocessed data frame

[1156] Step 4:

[1157] The server uses the preprocessed data to train a machine learning model.

[1158] The preprocessed data is divided into features and targets, and then split into training and test data. A model is trained using the random forest algorithm in scikit-learn.

[1159] Input: Preprocessed data frame

[1160] Output: A trained machine learning model

[1161] Step 5:

[1162] The server uses a trained machine learning model to predict suitable gifts for users.

[1163] By extracting features based on a specific user ID and inputting them into an existing model, the system predicts the recommended gift.

[1164] Input: User ID and preprocessed data frame

[1165] Output: A list of recommended gifts

[1166] Step 6:

[1167] The server sends the predicted gift to the user's device.

[1168] Export the list of recommended items in JSON format or other format and send it to the user's smartphone or other device.

[1169] Input: List of recommended gifts

[1170] Output: Data exported in JSON format

[1171] Step 7:

[1172] The terminal displays the predicted gift received from the server.

[1173] The received JSON data is analyzed and displayed on the screen for the user. For example, for user ID 12345, "Recommended gifts: sports goods, cooking class tickets, and relaxation goods" is displayed.

[1174] Input: JSON data sent from the server

[1175] Output: A list of gifts displayed on the screen

[1176] 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.

[1177] 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.

[1178] 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.

[1179] 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.

[1180] 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.

[1181] 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.

[1182] 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).

[1183] 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.

[1184] 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."

[1185] 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.

[1186] 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).

[1187] 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.

[1188] 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.

[1189] 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.

[1190] 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.

[1191] 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.

[1192] 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.

[1193] 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.

[1194] 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.

[1195] 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.

[1196] 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.

[1197] The following is further disclosed regarding the above embodiment.

[1198] (Claim 1)

[1199] A means of collecting user purchase history;

[1200] A means of collecting user interest genres,

[1201] a means of collecting answers to user questions;

[1202] a means of integrating the collected data;

[1203] a means for preprocessing the integrated data;

[1204] means for training a machine learning model using the preprocessed data;

[1205] A means for predicting suitable rewards for users using a trained machine learning model; and

[1206] means for transmitting the predicted gift to the user terminal;

[1207] A means of displaying predicted gifts on the user's device

[1208] A system including:

[1209] (Claim 2)

[1210] The system according to claim 1, further comprising means for imputing missing values ​​in the collected data.

[1211] (Claim 3)

[1212] 10. The system of claim 1, further comprising means for normalizing the collected data.

[1213] "Example 1"

[1214] (Claim 1)

[1215] A means of collecting user purchase history from a database;

[1216] A means of collecting user interest genres from the database;

[1217] a means for collecting answers to user questions from a database;

[1218] a means for combining the collected data into an integrated data set;

[1219] A means to remove unnecessary columns from the consolidated data;

[1220] a means of imputing missing values;

[1221] a means of normalizing the data;

[1222] means for training a machine learning model using the preprocessed data;

[1223] A means for predicting suitable rewards for users using a trained machine learning model; and

[1224] A means for exporting and sending the predicted gift list to a user device;

[1225] A means of displaying predicted gifts on the user's device

[1226] A system including:

[1227] (Claim 2)

[1228] The system of claim 1, further comprising means for imputing missing values ​​as "no response."

[1229] (Claim 3)

[1230] 10. The system of claim 1, further comprising: means for converting the exported data into a JSON format.

[1231] "Application Example 1"

[1232] (Claim 1)

[1233] A means of collecting user purchase history;

[1234] A means of collecting user interest genres,

[1235] a means of collecting answers to user questions;

[1236] a means of integrating the collected data;

[1237] a means for preprocessing the integrated data;

[1238] means for training a machine learning model using the preprocessed data;

[1239] A means for predicting suitable rewards for users using a trained machine learning model; and

[1240] means for transmitting the predicted gift to the user terminal;

[1241] a means for displaying the predicted gift on the user terminal;

[1242] A way to output the predicted gift list in JSON format,

[1243] means for providing a user interface for displaying the predicted gift on a user's terminal;

[1244] A system including:

[1245] (Claim 2)

[1246] The system according to claim 1, further comprising means for imputing missing values ​​in the collected data.

[1247] (Claim 3)

[1248] 10. The system of claim 1, further comprising means for normalizing the collected data.

[1249] "Example 2: Combining Emotion Engines"

[1250] (Claim 1)

[1251] A means of collecting user purchase history;

[1252] A means of collecting user interest genres,

[1253] a means of collecting answers to user questions;

[1254] a means of integrating the collected data;

[1255] a means for preprocessing the integrated data;

[1256] means for training a machine learning model using the preprocessed data;

[1257] a means for predicting suitable items for a user using a trained machine learning model;

[1258] A means of collecting user sentiment data and incorporating it into the integrated data;

[1259] a means for retraining a machine learning model using the emotion data; and

[1260] means for transmitting the predicted items to a user terminal;

[1261] A means for displaying predicted items on a user terminal

[1262] A system including:

[1263] (Claim 2)

[1264] The system according to claim 1, further comprising means for imputing missing values ​​in the collected data.

[1265] (Claim 3)

[1266] 10. The system of claim 1, further comprising means for normalizing the collected data.

[1267] "Application example 2 when combining emotion engines"

[1268] (Claim 1)

[1269] A means of collecting user purchase history;

[1270] A means of collecting user interest genres,

[1271] a means of collecting answers to user questions;

[1272] a means for collecting user emotional data;

[1273] a means of integrating the collected data;

[1274] a means for preprocessing the integrated data;

[1275] means for training a machine learning model using the preprocessed data;

[1276] A means for predicting suitable rewards for users using a trained machine learning model; and

[1277] means for transmitting the predicted gift to the user terminal;

[1278] A means of displaying predicted gifts on the user's device

[1279] A system including:

[1280] (Claim 2)

[1281] The system according to claim 1, further comprising means for imputing missing values ​​in the collected data.

[1282] (Claim 3)

[1283] 10. The system of claim 1, further comprising means for normalizing the collected data. [Explanation of symbols]

[1284] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of collecting user purchase history; A means of collecting user interest genres, a means of collecting answers to user questions; a means of integrating the collected data; a means for preprocessing the integrated data; means for training a machine learning model using the preprocessed data; A means for predicting suitable rewards for users using a trained machine learning model; and means for transmitting the predicted gift to the user terminal; A means of displaying predicted gifts on the user's device A system including:

2. The system of claim 1 further comprising means for imputing missing values ​​in the collected data.

3. 10. The system of claim 1, further comprising means for normalizing the collected data.

Images