system

Abstract

We provide the system. [Solution] The data processing device includes means for detecting missing values ​​from input tabular data, A means for compensating for the detected missing values ​​using an initial imputation means, A means for further enhancing the data enhanced by the initial enhancement means using a generative model, A means of outputting processed data, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot performed by at least one processor, the method including: receiving a user utterance; adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot; 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

Summary of the Invention

Problems to be Solved by the Invention

[0004] One of the main problems that data scientists face in improving the quality of analysis models is to efficiently and accurately complement missing values in data. Conventional manual complementation of missing values has problems of taking time, having uneven accuracy, and reducing the efficiency of analysis. Therefore, it is required to simultaneously improve the accuracy and efficiency in data processing.

[0005] ​The present invention provides a data processing system that detects missing values ​​from input tabular data, initially imputes the missing values ​​using an initial imputation means, and then further imputes them with high accuracy based on context using a generative model. As a result, missing values ​​are automatically and accurately imputed, improving the efficiency and accuracy of data processing.

[0006] A "data processing device" is a device that has the function of analyzing, transforming, and supplementing input data, and outputting it in a format suitable for the purpose.

[0007] "Tabular data" refers to data with a structure composed of rows and columns, and is a data format commonly used in databases, spreadsheets, and other applications.

[0008] "Missing values" refer to locations in a dataset where values ​​that should be present are missing.

[0009] "Means of detection" refers to methods or techniques for identifying and extracting missing values ​​within a dataset.

[0010] An "initial imputation method" is a method or technique that uses simple statistical methods to first fill in missing values ​​with representative values ​​such as the mean or mode.

[0011] A "generative model" is a technique that uses machine learning and artificial intelligence technologies to understand the context of data and generate appropriate values ​​to impute missing values.

[0012] "Means of high-precision imputation" refers to methods or techniques for generating more accurate data by using generative models to infer missing values ​​based on context.

[0013] "Means for outputting processed data" refers to a method or technique for saving or displaying processed data in a predetermined format. [Brief explanation of the drawing]

[0014] [Figure 1] It is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] It is a conceptual diagram showing an example of the main functions of a data processing device and a smart device according to the first embodiment. [Figure 3] It is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] It is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] It is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] It is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] It is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0015] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0016] First, the terms used in the following description will be explained.

[0017] In the following embodiments, a tagged processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0018] In the following embodiments, a tagged RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0019] In the following embodiments, a tagged storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[0020] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0022] [First Embodiment]

[0023] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0024] As shown in Figure 1, the 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.

[0025] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0026] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0027] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and 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.

[0028] 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 perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0030] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] As shown in Figure 2, in the data processing device 12, a specific processing 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" related to the technology of this 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 according to the specific processing program 56 executed on the RAM 30.

[0032] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0033] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0034] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0035] This invention is a system for efficiently and accurately imputing missing values ​​in tabular data input by a data processing device. The server first reads the data received in CSV file format and manages it as a data frame using the pandas library. Missing values ​​are detected from this data frame and imputed using initial statistical methods, using the mean for numerical data and the mode for categorical data.

[0036] Furthermore, the server performs completion using a generative model. Here, a generative model based on natural language processing techniques is used to more precisely complete missing values ​​based on the context within the dataset. In this process, a prompt is generated for each missing location and input into the generative model to generate a value. The generated value is also applied to locations that were not filled in by the initial completion.

[0037] On the terminal, the user inputs data, the server processes it, and then outputs the completed data to the terminal. This output data is saved in CSV file format as an analyzable dataset with all missing values ​​imputed.

[0038] As a concrete example, suppose a user uploads a CSV file containing "customer data" to the server. This file includes attributes such as "age," "gender," and "purchase history." The server detects missing data, performs initial imputation, and then uses a generative model to impute context-dependent missing values ​​such as "purchase history." Finally, the user on their device receives the data with the missing values ​​imputed, which they can then use for further analysis and model building.

[0039] Thus, the present invention provides a data analysis environment with high accuracy and efficiency through the automation of missing value imputation in data processing.

[0040] The following describes the processing flow.

[0041] Step 1:

[0042] The server receives tabular data in CSV format uploaded by the user and reads it as a dataframe using the pandas library. This dataframe contains various attributes, forming a data base that is useful for analysis.

[0043] Step 2:

[0044] The server analyzes the loaded data frame and performs a process to detect missing values ​​in each column. Based on the detection results, it prepares to perform initial imputation.

[0045] Step 3:

[0046] The server imputes missing values ​​in each column of the data frame according to the initial imputation method. Specifically, for numerical data, it calculates the mean and uses that value to fill in the missing values. For categorical data, it identifies the mode and fills in the corresponding values.

[0047] Step 4:

[0048] The server uses a generative model to fill in missing values ​​that were not properly filled in during the initial completion. At this stage, a generative model that considers the context between data attributes, based on natural language processing techniques, is used to create prompts about the missing locations and estimate appropriate values.

[0049] Step 5:

[0050] The server aggregates the data supplemented by the generative model and creates a processed, integrated data frame. This data frame is then modified into a complete tabular format.

[0051] Step 6:

[0052] The server saves the final data frame in CSV format and outputs it to the user's terminal. This output data is provided as a dataset with missing values ​​already imputed, ready for analysis and further data processing.

[0053] (Example 1)

[0054] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0055] Accurate and efficient imputation of missing values ​​in digital information is a critical challenge in data analysis and machine learning. However, traditional statistical methods of imputation fail to consider the context of the data, potentially resulting in reduced accuracy. Furthermore, manual imputation of missing values ​​is inefficient and time-consuming. Addressing these issues is essential.

[0056] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0057] In this invention, the server includes means for detecting missing values ​​from digital information, means for imputing missing values ​​using initial statistical methods, and means for imputing them with high accuracy based on natural language processing techniques using a generative model. This enables highly accurate imputation of missing values ​​that takes into account the context of the digital information.

[0058] "Digital information" refers to any format of information that is processed, transmitted, or stored electronically. This includes databases, electronic spreadsheet software, and CSV files.

[0059] "Missing values" refer to parts of a dataset where information that should be included is missing for some reason.

[0060] "Initial statistical methods" are basic statistical approaches used to impute missing data, and they involve calculating the mean or mode to fill in the gaps.

[0061] A "generative model" refers to a model that has the ability to learn from large amounts of data and generate new data. It is particularly applied to natural language processing and data completion.

[0062] "Natural language processing technology" refers to the technology that enables computers to understand, generate, and manipulate human language.

[0063] A "prompt statement" refers to a sentence or phrase that is input into a generative model, and is used to instruct the model on a specific task.

[0064] This invention is a system that uses a digital information processing device to efficiently and accurately impute missing values ​​in a dataset. Specific embodiments are described below.

[0065] First, the user provides the server with tabular information as digital data in the form of a CSV file. For example, this data might include "customer information" and attributes such as "age," "gender," and "purchase history." Upon receiving this file, the server uses the pandas library in the Python programming language to manage the data in the form of a data frame. Pandas is software designed to efficiently manipulate large amounts of data.

[0066] The server initiates a process to identify missing values ​​in the data frame, imputing missing values ​​in numerical data by applying the mean and missing values ​​in categorical data by applying the mode. This initial statistical method performs basic missing value imputation.

[0067] Next, the server utilizes a generative AI model to more precisely fill in missing values, taking into account the context of the digital information. This process employs natural language processing techniques to generate prompts for each missing value. An example prompt might be, "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to fill in this information." The generated prompts are input into the generative AI model, and the server calculates the appropriate completion values.

[0068] The server applies the calculated interpolation values ​​to the areas that were incomplete in the initial interpolation, creating a complete dataset. This data is then output to the terminal in CSV format for the user to receive. The user can then use the completed dataset for further detailed analysis or to train machine learning models.

[0069] Thus, the present invention improves the accuracy and efficiency of data analysis by effectively imputing missing values ​​in digital information.

[0070] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0071] Step 1:

[0072] The user uploads tabular digital data to the server in CSV file format. The server receives this file and manages the data in DataFrame format using the pandas library. The input is a user-provided CSV file, and the output is data in DataFrame format.

[0073] Step 2:

[0074] The server analyzes the data frame and detects missing data points. Specifically, it uses pandas functionality to scan each column and identify missing values. In this step, the data frame is used as input, and the output is the location information of the missing values.

[0075] Step 3:

[0076] The server imputes missing values ​​using initial statistical methods. For numerical data, it calculates the mean of the column; for categorical data, it calculates the mode. The input is the location information of the missing values ​​and a data frame, and the output is a data frame with the initial imputation performed.

[0077] Step 4:

[0078] The server uses a generative AI model to precisely impute missing values ​​based on the context within the dataset. For each missing value, it generates a prompt message, such as "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to impute this information," which is then input into the generative model. The input consists of the prompt message and an initially imputed dataframe, while the output is the final imputed dataframe.

[0079] Step 5:

[0080] The server outputs the final completed data in CSV format and sends it to the terminal. The user can then use the received completed data for further analysis or to train machine learning models. The input is the final data frame, and the output is a CSV file usable by the user.

[0081] (Application Example 1)

[0082] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0083] Missing customer information and purchase history in e-commerce can reduce the accuracy of product recommendations. This can hinder effective marketing to customers and potentially lead to decreased sales. Furthermore, manual data completion is time-consuming and labor-intensive, hindering efficient operations. This invention aims to solve these problems by quickly and accurately completing missing data values ​​and enabling the presentation of personalized product information.

[0084] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0085] In this invention, the server includes means for detecting missing values ​​from input tabular data using a data processing device, means for imputing the detected missing values ​​using an initial imputation means, means for further imputing the data imputed by the initial imputation means with higher accuracy using a generative model, and means for presenting personalized product information based on the processed data. This makes it possible to provide customized recommendations to consumers after imputing missing values.

[0086] A "data processing device" is a device that analyzes input tabular data and detects and imputes missing values.

[0087] "Missing values" refer to a state in a dataset where data that should be present is missing, indicating a lack of information.

[0088] "Initial imputation methods" refer to techniques that use basic statistical methods to perform initial imputation on detected missing values.

[0089] A "generative model" refers to a model that analyzes the context and characteristics of data and uses natural language processing techniques to accurately impute missing values.

[0090] "Personalized product information" refers to information about products and services that are optimized for each individual consumer, based on processed customer data.

[0091] "Means of output" refers to methods or devices for providing the supplemented data to external systems or users.

[0092] The system that realizes this invention operates using a server, user terminals, and a communication network.

[0093] First, the user sends a data file related to the e-commerce site from their device to the server. This data file often contains the customer's personal information and purchase history. Upon receiving this data, the server uses the pandas library to analyze the tabular data and detect missing values.

[0094] For initial completion, the server uses the Scikit-learn library to impute missing values, using the mean for numerical data and the mode for string data. After this initial completion, the server utilizes OpenAI's language generation model to perform more precise, context-based imputation of missing values. Specifically, it generates prompts corresponding to the missing data and inputs these into the generation model to produce more appropriate completion values.

[0095] Ultimately, personalized product information is resent to the user's device based on the data with missing values ​​filled in. This information helps users select the most suitable products on the e-commerce site. Users can use the filled-in data to refer to product ratings and predictions based on their past purchase history.

[0096] As a concrete example, suppose a user has a particular preference for eco-friendly products. In this case, the generative model would fill in the missing parts of the user's purchase history in a similar context and suggest recommended products related to eco-friendly items. An example of a prompt in this process would be, "Generate a recommendation for missing purchase history data for a customer who frequently buys eco-friendly products and has recently viewed outdoor adventure gear." This prompt enables the generative AI model to fill in the missing data and provide the user with appropriate product information.

[0097] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0098] Step 1:

[0099] The user uploads a data file containing customer information and purchase history from their device to the server. The input is a CSV data file, and the output is the original data file stored on the server. In this step, the server confirms the file transmission and notifies the user that it has been successfully received.

[0100] Step 2:

[0101] The server reads the received data file as a DataFrame using the pandas library. The input is data in CSV format, and the output is a DataFrame object. At this point, it detects missing values ​​in the data and lists which items are missing.

[0102] Step 3:

[0103] The server performs initial imputation on dataframes where missing values ​​are detected. Specifically, it uses the mean for numerical data and the mode for string data. It utilizes the Scikit-learn library to perform imputation according to the data type. The input is a dataframe with missing values, and the output is a dataframe with initial imputation applied.

[0104] Step 4:

[0105] The server performs context-based, high-precision missing value imputation on the data after initial imputation. This process utilizes OpenAI's generative AI model. During this process, prompt sentences based on the context of each missing value are generated and input into the generative model. The input consists of the initially imputed data frame and the generated prompt sentences, while the output is the data frame imputed by the generative model.

[0106] Step 5:

[0107] The server creates personalized product information based on complete data supplemented by a generative AI model. This takes into account the consumer's profile and purchase history. The input is a supplemented data frame, and the output is a list of personalized recommended products.

[0108] Step 6:

[0109] The server sends the final output, personalized product information, to the user's terminal. The user receives this information on their terminal and can select from the displayed product list. The input is personalized product information, and the output is the product list displayed on the user's terminal. In this step, the user is notified that new products have been recommended.

[0110] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0111] This invention relates to a system that combines a user emotion engine with a data processing device to accurately impute missing values. The server first receives tabular data provided by the user in CSV format and manages it as a data frame using the pandas library. Missing values ​​in this data frame are detected, and an initial imputation means imputes numerical data with the mean and categorical data with the mode.

[0112] Next, the server utilizes an emotion engine. The emotion engine analyzes user emotions in real time from user interactions and feedback, and incorporates this emotion data into the data completion process. Specifically, it analyzes the user's facial expressions and voice while manipulating data, and extracts emotions using natural language processing techniques. As a result, it fine-tunes the data completion strategy and methods during the final missing value completion using a generative model.

[0113] For example, if a user shows dissatisfaction with the data interpolation results, the emotion engine can sense this and instruct the generative model to repeat the interpolation or try a different estimation approach. This process enables customizable data interpolation that meets the user's needs and expectations.

[0114] Finally, the server outputs a data frame in CSV format, incorporating sentiment information and imputing missing values, which is then provided to the user for download. This processed data improves user satisfaction and is optimized for the specific purpose of data analysis.

[0115] Thus, the present invention provides a more intuitive and effective data processing environment by taking user emotions into consideration during data completion processing.

[0116] The following describes the processing flow.

[0117] Step 1:

[0118] The user uploads tabular data, including missing values, from their terminal to the server. The server reads this data as a DataFrame using the pandas library and stores it in a structured format.

[0119] Step 2:

[0120] The server analyzes each column in the data frame to check for missing values. It detects any identified missing values ​​and uses this information to prepare for initial imputation.

[0121] Step 3:

[0122] The server uses initial imputation methods to fill in missing values. For numerical data, it calculates the mean and uses that result to fill in missing values. For categorical data, it identifies the mode and similarly fills in missing values.

[0123] Step 4:

[0124] The server activates the emotion engine and collects user emotion data through interaction with the user's device. The emotion engine monitors the user's facial expressions, voice, input behavior, etc., and analyzes emotions in real time.

[0125] Step 5:

[0126] The server performs advanced missing value imputation using a generative model. In this process, it considers user sentiment data obtained from the sentiment engine. If the user indicates dissatisfaction or anxiety, the generative model adjusts its imputation methods and algorithms to attempt to imputate the data in a way that aligns with the user's expectations.

[0127] Step 6:

[0128] The server saves the dataframe, with all missing values ​​imputed, in CSV format. This data is now complete with all missing values ​​filled in, and users can download this file from their terminal for further analysis.

[0129] (Example 2)

[0130] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0131] Traditional methods for imputing missing information in tabular data often failed to consider the data's characteristics and context, making high-precision imputation difficult. Furthermore, a lack of approaches that considered user expectations and emotions meant that data imputation results did not always align with user intent.

[0132] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0133] In this invention, the server includes means for detecting missing information from tabular data input by a data processing unit, means for analyzing the user's emotional information using an emotional analysis engine and adjusting the data completion policy based on the analysis results, and means for completing the data to which the completion policy has been applied with high accuracy using a generative model. This enables high-precision completion of missing information and customization according to the user's emotions.

[0134] A "data processing unit" is a device that has the function of acquiring information from table-formatted data and managing and manipulating it.

[0135] "Missing information" refers to data elements that are missing in tabular data.

[0136] "Initial imputation techniques" are methods for temporarily filling in missing information using basic methods such as the mean or mode.

[0137] A "sentiment analysis engine" is a system that analyzes and extracts a user's emotions in real time from data and operational status.

[0138] A "generative model" is a model that uses machine learning and artificial intelligence technologies to learn data trends and perform more accurate data imputation.

[0139] A "prompt statement" is a form of text used to give instructions or make adjustments to a generative model.

[0140] This invention provides a novel data processing system that takes user sentiment into consideration in order to accurately fill in missing information in tabular data. This system involves a server, a terminal, and a user, each with a specific role.

[0141] When the server receives tabular data, it first uses the pandas library to manage the data as a data frame. It then detects missing information within the data frame and performs initial imputation using the mean for numerical data and the mode for categorical data.

[0142] Next, the device collects user emotion data. Specifically, it uses a webcam and microphone to capture the user's facial expressions and voice. This allows the user's emotional state to be analyzed in real time through an emotion analysis engine.

[0143] The sentiment information obtained by the sentiment analysis engine is sent to the server. Based on this information, the server adjusts the data completion strategy. Here, a generative AI model is used to instruct the model to perform completions that are appropriate to the user's emotions. In this process, specific instructions are given to the generative AI model using prompt statements. For example, an instruction such as "If the user is dissatisfied with the data completion result, try completing it using a different method" might be given.

[0144] Once the data completion is satisfactory to the user, the server outputs the completed dataset in CSV format, making it available for download to the user via their terminal. This achieves data completion that balances the user's subjective satisfaction with the objective accuracy of the data. By incorporating user emotions, this system enables more intuitive and appropriate data processing.

[0145] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0146] Step 1:

[0147] The server receives a tabular data file from the user. When a user uploads data in CSV format via a terminal, the server receives this file and converts it into a DataFrame using the pandas library. The input is a CSV file, and the output is a DataFrame in internal memory.

[0148] Step 2:

[0149] The server detects missing data within the data frame. This step applies an algorithm to check for missing values ​​in each column of the data frame. The input is the data frame, and the output is the data frame containing the missing data.

[0150] Step 3:

[0151] The server uses initial imputation techniques to fill in missing information. For numerical data, it calculates the mean of the column and applies it to the missing parts; for categorical data, it identifies the mode and applies it. The input is a data frame containing missing information, and the output is a data frame with initial imputation completed.

[0152] Step 4:

[0153] The device uses a webcam and microphone to capture facial expressions and voice to collect user emotions. The input is real-time emotional information from the user, and the output is digital data sent to an emotion analysis engine.

[0154] Step 5:

[0155] The server uses an emotion analysis engine to analyze the user's emotional information. It uses natural language processing techniques to quantify the emotional data and, based on the results, generates the data necessary to adjust the interpolation process. The input is the user's emotional data, and the output is data containing adjustment parameters.

[0156] Step 6:

[0157] The server sends prompt messages to the generating AI model using tuning parameters to perform final data completion. An example of a prompt message is, "The user's sentiment rating showed dissatisfaction; please apply a different completion method." The input is the tuning parameters and an initially completed data frame, and the output is the final completed data frame.

[0158] Step 7:

[0159] The server converts the final completed dataframe into CSV format and provides it to the user's terminal in a downloadable format. The input is the final completed dataframe, and the output is a CSV file.

[0160] (Application Example 2)

[0161] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0162] In data processing, imputing missing values ​​is a crucial task. However, traditional methods often process data mechanically without considering user emotions, potentially leading to decreased user satisfaction. Especially with emotionally charged data such as reviews and feedback, data imputation that aligns with user intent and needs is essential. The challenge lies in improving the user experience and enabling more intuitive and effective data analysis.

[0163] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0164] In this invention, the server includes means for detecting missing values ​​from input tabular data, means for imputing missing values ​​using initial imputation means, and means for acquiring user emotion information using emotion analysis technology and imputing based on the emotion information with high accuracy using a generative model. This makes it possible to imputate data in accordance with the user's emotions, thereby improving the user experience.

[0165] A "data processing device" is a device that analyzes and transforms input data and performs information processing according to its purpose.

[0166] A "missing value" is a state in a dataset where a value that should be present is missing.

[0167] "Initial imputation methods" are techniques for imputing missing values ​​using basic methods, such as using the mean for numerical data and the mode for categorical data.

[0168] "Emotion analysis technology" is a technology that analyzes a user's emotions from their facial expressions and voice, and grasps their emotional state in real time.

[0169] A "generative model" is a machine learning model that can learn patterns from large amounts of data and generate new data.

[0170] "High-precision imputation" means inferring and filling in missing values ​​with values ​​that are more accurate and in line with the user's intent.

[0171] "User emotional information" refers to data about the emotional state and characteristics obtained by analyzing the user's facial expressions, voice, etc.

[0172] A "data imputation policy" refers to the basic guidelines and strategies for how to imputate missing values.

[0173] In the system for implementing this invention, the server is primarily responsible for data processing. As a data processing device, the server receives tabular data provided by the user. This data is input in formats such as CSV and managed as a dataframe using the pandas library. The server first detects missing values ​​in this dataframe and performs initial imputation. For numerical data, the mean is used to imputate missing values, and for categorical data, the mode is used.

[0174] Next, the server executes emotion analysis technology. This technology analyzes facial expression and voice data collected from the user's terminal to obtain information about the user's emotions. A dedicated emotion engine is used for emotion analysis, analyzing emotions in real time. Based on this information, the generative AI model adjusts the missing value imputation strategy. Specifically, it dynamically changes the method of imputing missing values, taking into account the user's satisfaction level and emotional state. In addition, the generative AI model utilizes natural language processing technology to perform advanced imputation based on the context of the data.

[0175] The data processed in this way is ultimately output by the server in CSV format, which can be downloaded by the user. Through this process, users can experience customized data supplementation based on their emotions.

[0176] As a concrete example, consider a case where a user writes a product review with partially missing information. The application reads the user's emotion of "satisfaction" from their facial expressions and voice, and fills in the missing parts with positive content. An example of an input prompt for the generative AI model in this case could be: "Review text: This product is easy to use,," "User emotion: Satisfied," "Complete the missing feedback." This enables data completion optimized for each individual user.

[0177] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0178] Step 1:

[0179] The server receives tabular data in CSV format from the terminal. This data is converted into a DataFrame using the pandas library. The input is the raw data provided by the user, and the output is a digital DataFrame. The server analyzes this data and detects missing values.

[0180] Step 2:

[0181] The server imputes detected missing values ​​using an initial imputation mechanism. For numerical data, the mean is used; for categorical data, the mode is used. The input is a data frame containing missing values, and the output is a data frame with initial imputation performed. The server thus ensures the basic integrity of the data.

[0182] Step 3:

[0183] The user's device captures their facial expressions and voice data in real time and sends it to a server. This data is processed using emotion analysis technology and output as the user's emotional information. The input is the user's biometric data, and the output is the analyzed emotional data. The device then uses this to understand the user's emotions.

[0184] Step 4:

[0185] The server adjusts the generative AI model using the provided sentiment data and performs highly accurate imputation of missing values. The generative AI model uses prompts based on the user's sentiment and performs deep contextual analysis using natural language processing techniques. The input is an initially imputed data frame and sentiment data, and the output is an imputed data frame optimized based on sentiment. The server generates data that matches the user's intent.

[0186] Step 5:

[0187] Ultimately, the server outputs a completed and optimized dataframe in CSV format, making it available for download to the user. The input is a dataframe with missing values ​​imputed, and the output is a well-structured data file for the user. The server thus provides an environment where users can easily utilize the data.

[0188] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating 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.

[0189] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0190] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0191] [Second Embodiment]

[0192] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0193] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0194] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0195] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0196] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0197] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0198] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0199] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0200] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0201] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0202] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0203] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. 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".

[0204] This invention is a system for efficiently and accurately imputing missing values ​​in tabular data input by a data processing device. The server first reads the data received in CSV file format and manages it as a data frame using the pandas library. Missing values ​​are detected from this data frame and imputed using initial statistical methods, using the mean for numerical data and the mode for categorical data.

[0205] Furthermore, the server performs completion using a generative model. Here, a generative model based on natural language processing techniques is used to more precisely complete missing values ​​based on the context within the dataset. In this process, a prompt is generated for each missing location and input into the generative model to generate a value. The generated value is also applied to locations that were not filled in by the initial completion.

[0206] On the terminal, the user inputs data, the server processes it, and then outputs the completed data to the terminal. This output data is saved in CSV file format as an analyzable dataset with all missing values ​​imputed.

[0207] As a concrete example, suppose a user uploads a CSV file containing "customer data" to the server. This file includes attributes such as "age," "gender," and "purchase history." The server detects missing data, performs initial imputation, and then uses a generative model to impute context-dependent missing values ​​such as "purchase history." Finally, the user on their device receives the data with the missing values ​​imputed, which they can then use for further analysis and model building.

[0208] Thus, the present invention provides a data analysis environment with high accuracy and efficiency through the automation of missing value imputation in data processing.

[0209] The following describes the processing flow.

[0210] Step 1:

[0211] The server receives tabular data in CSV format uploaded by the user and reads it as a dataframe using the pandas library. This dataframe contains various attributes, forming a data base that is useful for analysis.

[0212] Step 2:

[0213] The server analyzes the loaded data frame and performs a process to detect missing values ​​in each column. Based on the detection results, it prepares to perform initial imputation.

[0214] Step 3:

[0215] The server imputes missing values ​​in each column of the data frame according to the initial imputation method. Specifically, for numerical data, it calculates the mean and uses that value to fill in the missing values. For categorical data, it identifies the mode and fills in the corresponding values.

[0216] Step 4:

[0217] The server uses a generative model to fill in missing values ​​that were not properly filled in during the initial completion. At this stage, a generative model that considers the context between data attributes, based on natural language processing techniques, is used to create prompts about the missing locations and estimate appropriate values.

[0218] Step 5:

[0219] The server aggregates the data supplemented by the generative model and creates a processed, integrated data frame. This data frame is then modified into a complete tabular format.

[0220] Step 6:

[0221] The server saves the final data frame in CSV format and outputs it to the user's terminal. This output data is provided as a dataset with missing values ​​already imputed, ready for analysis and further data processing.

[0222] (Example 1)

[0223] Next, we will describe Example 1. 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."

[0224] Accurate and efficient imputation of missing values ​​in digital information is a critical challenge in data analysis and machine learning. However, traditional statistical methods of imputation fail to consider the context of the data, potentially resulting in reduced accuracy. Furthermore, manual imputation of missing values ​​is inefficient and time-consuming. Addressing these issues is essential.

[0225] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0226] In this invention, the server includes means for detecting missing values ​​from digital information, means for imputing missing values ​​using initial statistical methods, and means for imputing them with high accuracy based on natural language processing techniques using a generative model. This enables highly accurate imputation of missing values ​​that takes into account the context of the digital information.

[0227] "Digital information" refers to any format of information that is processed, transmitted, or stored electronically. This includes databases, electronic spreadsheet software, and CSV files.

[0228] "Missing values" refer to parts of a dataset where information that should be included is missing for some reason.

[0229] "Initial statistical methods" are basic statistical approaches used to impute missing data, and they involve calculating the mean or mode to fill in the gaps.

[0230] A "generative model" refers to a model that has the ability to learn from large amounts of data and generate new data. It is particularly applied to natural language processing and data completion.

[0231] "Natural language processing technology" refers to the technology that enables computers to understand, generate, and manipulate human language.

[0232] A "prompt statement" refers to a sentence or phrase that is input into a generative model, and is used to instruct the model on a specific task.

[0233] This invention is a system that uses a digital information processing device to efficiently and accurately impute missing values ​​in a dataset. Specific embodiments are described below.

[0234] First, the user provides the server with tabular information as digital data in the form of a CSV file. For example, this data might include "customer information" and attributes such as "age," "gender," and "purchase history." Upon receiving this file, the server uses the pandas library in the Python programming language to manage the data in the form of a data frame. Pandas is software designed to efficiently manipulate large amounts of data.

[0235] The server initiates a process to identify missing values ​​in the data frame, imputing missing values ​​in numerical data by applying the mean and missing values ​​in categorical data by applying the mode. This initial statistical method performs basic missing value imputation.

[0236] Next, the server utilizes a generative AI model to more precisely fill in missing values, taking into account the context of the digital information. This process employs natural language processing techniques to generate prompts for each missing value. An example prompt might be, "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to fill in this information." The generated prompts are input into the generative AI model, and the server calculates the appropriate completion values.

[0237] The server applies the calculated interpolation values ​​to the areas that were incomplete in the initial interpolation, creating a complete dataset. This data is then output to the terminal in CSV format for the user to receive. The user can then use the completed dataset for further detailed analysis or to train machine learning models.

[0238] Thus, the present invention improves the accuracy and efficiency of data analysis by effectively imputing missing values ​​in digital information.

[0239] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0240] Step 1:

[0241] The user uploads tabular digital data to the server in CSV file format. The server receives this file and manages the data in DataFrame format using the pandas library. The input is a user-provided CSV file, and the output is data in DataFrame format.

[0242] Step 2:

[0243] The server analyzes the data frame and detects missing data points. Specifically, it uses pandas functionality to scan each column and identify missing values. In this step, the data frame is used as input, and the output is the location information of the missing values.

[0244] Step 3:

[0245] The server imputes missing values ​​using initial statistical methods. For numerical data, it calculates the mean of the column; for categorical data, it calculates the mode. The input is the location information of the missing values ​​and a data frame, and the output is a data frame with the initial imputation performed.

[0246] Step 4:

[0247] The server uses a generative AI model to precisely impute missing values ​​based on the context within the dataset. For each missing value, it generates a prompt message, such as "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to impute this information," which is then input into the generative model. The input consists of the prompt message and an initially imputed dataframe, while the output is the final imputed dataframe.

[0248] Step 5:

[0249] The server outputs the final completed data in CSV format and sends it to the terminal. The user can then use the received completed data for further analysis or to train machine learning models. The input is the final data frame, and the output is a CSV file usable by the user.

[0250] (Application Example 1)

[0251] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0252] Missing customer information and purchase history in e-commerce can reduce the accuracy of product recommendations. This can hinder effective marketing to customers and potentially lead to decreased sales. Furthermore, manual data completion is time-consuming and labor-intensive, hindering efficient operations. This invention aims to solve these problems by quickly and accurately completing missing data values ​​and enabling the presentation of personalized product information.

[0253] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0254] In this invention, the server includes means for detecting missing values ​​from input tabular data using a data processing device, means for imputing the detected missing values ​​using an initial imputation means, means for further imputing the data imputed by the initial imputation means with higher accuracy using a generative model, and means for presenting personalized product information based on the processed data. This makes it possible to provide customized recommendations to consumers after imputing missing values.

[0255] A "data processing device" is a device that analyzes input tabular data and detects and imputes missing values.

[0256] "Missing values" refer to a state in a dataset where data that should be present is missing, indicating a lack of information.

[0257] "Initial imputation methods" refer to techniques that use basic statistical methods to perform initial imputation on detected missing values.

[0258] A "generative model" refers to a model that analyzes the context and characteristics of data and uses natural language processing techniques to accurately impute missing values.

[0259] "Personalized product information" refers to information about products and services that are optimized for each individual consumer, based on processed customer data.

[0260] "Means of output" refers to methods or devices for providing the supplemented data to external systems or users.

[0261] The system that realizes this invention operates using a server, user terminals, and a communication network.

[0262] First, the user sends a data file related to the e-commerce site from their device to the server. This data file often contains the customer's personal information and purchase history. Upon receiving this data, the server uses the pandas library to analyze the tabular data and detect missing values.

[0263] For initial completion, the server uses the Scikit-learn library to impute missing values, using the mean for numerical data and the mode for string data. After this initial completion, the server utilizes OpenAI's language generation model to perform more precise, context-based imputation of missing values. Specifically, it generates prompts corresponding to the missing data and inputs these into the generation model to produce more appropriate completion values.

[0264] Ultimately, personalized product information is resent to the user's device based on the data with missing values ​​filled in. This information helps users select the most suitable products on the e-commerce site. Users can use the filled-in data to refer to product ratings and predictions based on their past purchase history.

[0265] As a concrete example, suppose a user has a particular preference for eco-friendly products. In this case, the generative model would fill in the missing parts of the user's purchase history in a similar context and suggest recommended products related to eco-friendly items. An example of a prompt in this process would be, "Generate a recommendation for missing purchase history data for a customer who frequently buys eco-friendly products and has recently viewed outdoor adventure gear." This prompt enables the generative AI model to fill in the missing data and provide the user with appropriate product information.

[0266] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0267] Step 1:

[0268] The user uploads a data file containing customer information and purchase history from their device to the server. The input is a CSV data file, and the output is the original data file stored on the server. In this step, the server confirms the file transmission and notifies the user that it has been successfully received.

[0269] Step 2:

[0270] The server reads the received data file as a DataFrame using the pandas library. The input is data in CSV format, and the output is a DataFrame object. At this point, it detects missing values ​​in the data and lists which items are missing.

[0271] Step 3:

[0272] The server performs initial imputation on dataframes where missing values ​​are detected. Specifically, it uses the mean for numerical data and the mode for string data. It utilizes the Scikit-learn library to perform imputation according to the data type. The input is a dataframe with missing values, and the output is a dataframe with initial imputation applied.

[0273] Step 4:

[0274] The server performs context-based, high-precision missing value imputation on the data after initial imputation. This process utilizes OpenAI's generative AI model. During this process, prompt sentences based on the context of each missing value are generated and input into the generative model. The input consists of the initially imputed data frame and the generated prompt sentences, while the output is the data frame imputed by the generative model.

[0275] Step 5:

[0276] The server creates personalized product information based on complete data supplemented by a generative AI model. This takes into account the consumer's profile and purchase history. The input is a supplemented data frame, and the output is a list of personalized recommended products.

[0277] Step 6:

[0278] The server sends the final output, personalized product information, to the user's terminal. The user receives this information on their terminal and can select from the displayed product list. The input is personalized product information, and the output is the product list displayed on the user's terminal. In this step, the user is notified that new products have been recommended.

[0279] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0280] This invention relates to a system that combines a user emotion engine with a data processing device to accurately impute missing values. The server first receives tabular data provided by the user in CSV format and manages it as a data frame using the pandas library. Missing values ​​in this data frame are detected, and an initial imputation means imputes numerical data with the mean and categorical data with the mode.

[0281] Next, the server utilizes an emotion engine. The emotion engine analyzes user emotions in real time from user interactions and feedback, and incorporates this emotion data into the data completion process. Specifically, it analyzes the user's facial expressions and voice while manipulating data, and extracts emotions using natural language processing techniques. As a result, it fine-tunes the data completion strategy and methods during the final missing value completion using a generative model.

[0282] For example, when the user shows dissatisfaction with the interpolation result of the data, the emotion engine can sense this and give instructions to the generation model to try repeated complementation or different estimation approaches. Through this process, customizable data complementation according to the user's needs and expectations is realized.

[0283] Finally, the server outputs the data frame with missing values filled in, taking into account the emotion information, in CSV format and provides it in a format that the user can download. This processed data not only improves the user's satisfaction but also is in an optimized state according to the application of data analysis.

[0284] In this way, the present invention provides a more intuitive and effective data processing environment by considering the user's emotions in the data complementation process.

[0285] The following describes the processing flow.

[0286] Step 1:

[0287] The user uploads tabular data containing missing values from the terminal to the server. The server reads this data as a data frame using the pandas library and holds it in a structured format.

[0288] Step 2:

[0289] The server analyzes each column in the data frame and checks for the presence of missing values. The identified missing values are detected, and based on this information, preparations for initial complementation are made.

[0290] Step 3:

[0291] The server complements the missing values using the initial complementation means. For numerical data, the average value is calculated and used to fill in the missing values. In the case of categorical data, the most frequent value is identified and the missing values are complemented in the same way.

[0292] Step 4:

[0293] The server activates the emotion engine and collects user emotion data through interaction with the user's device. The emotion engine monitors the user's facial expressions, voice, input behavior, etc., and analyzes emotions in real time.

[0294] Step 5:

[0295] The server performs advanced missing value imputation using a generative model. In this process, it considers user sentiment data obtained from the sentiment engine. If the user indicates dissatisfaction or anxiety, the generative model adjusts its imputation methods and algorithms to attempt to imputate the data in a way that aligns with the user's expectations.

[0296] Step 6:

[0297] The server saves the dataframe, with all missing values ​​imputed, in CSV format. This data is now complete with all missing values ​​filled in, and users can download this file from their terminal for further analysis.

[0298] (Example 2)

[0299] Next, we will describe Example 2. 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".

[0300] Traditional methods for imputing missing information in tabular data often failed to consider the data's characteristics and context, making high-precision imputation difficult. Furthermore, a lack of approaches that considered user expectations and emotions meant that data imputation results did not always align with user intent.

[0301] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0302] In this invention, the server includes means for detecting missing information from table-form data input by a data processing unit, means for analyzing the user's sentiment information using a sentiment analysis engine and adjusting the data completion policy based on the analysis result, and means for accurately completing the data to which the completion policy is applied using a generation model. As a result, the missing information can be accurately completed, and customization according to the user's sentiment becomes possible.

[0303] The "data processing unit" is a device that has the function of acquiring information from table-form data and performing management and operations.

[0304] The "missing information" refers to the elements of data that are lacking in the table-form data.

[0305] The "initial completion technology" is a method of temporarily completing missing information using basic methods such as the average value or the most frequent value.

[0306] The "sentiment analysis engine" is a system for analyzing and extracting the user's sentiment in real time from data and operation situations.

[0307] The "generation model" is a model that uses machine learning and artificial intelligence technologies to learn the trends of data and perform more accurate data completion.

[0308] The "prompt sentence" is a form of sentence used to give operation instructions and adjustments to the generation model.

[0309] This invention provides a new data processing system that takes into account the user's sentiment in order to accurately complete the missing information in tabular data. This system involves a server, a terminal, and a user, each having a specific role.

[0310] When the server receives tabular data, it first uses the pandas library to manage the data as a data frame. It then detects missing information within the data frame and performs initial imputation using the mean for numerical data and the mode for categorical data.

[0311] Next, the device collects user emotion data. Specifically, it uses a webcam and microphone to capture the user's facial expressions and voice. This allows the user's emotional state to be analyzed in real time through an emotion analysis engine.

[0312] The sentiment information obtained by the sentiment analysis engine is sent to the server. Based on this information, the server adjusts the data completion strategy. Here, a generative AI model is used to instruct the model to perform completions that are appropriate to the user's emotions. In this process, specific instructions are given to the generative AI model using prompt statements. For example, an instruction such as "If the user is dissatisfied with the data completion result, try completing it using a different method" might be given.

[0313] Once the data completion is satisfactory to the user, the server outputs the completed dataset in CSV format, making it available for download to the user via their terminal. This achieves data completion that balances the user's subjective satisfaction with the objective accuracy of the data. By incorporating user emotions, this system enables more intuitive and appropriate data processing.

[0314] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0315] Step 1:

[0316] The server receives a tabular data file from the user. When a user uploads data in CSV format via a terminal, the server receives this file and converts it into a DataFrame using the pandas library. The input is a CSV file, and the output is a DataFrame in internal memory.

[0317] Step 2:

[0318] The server detects missing data within the data frame. This step applies an algorithm to check for missing values ​​in each column of the data frame. The input is the data frame, and the output is the data frame containing the missing data.

[0319] Step 3:

[0320] The server uses initial imputation techniques to fill in missing information. For numerical data, it calculates the mean of the column and applies it to the missing parts; for categorical data, it identifies the mode and applies it. The input is a data frame containing missing information, and the output is a data frame with initial imputation completed.

[0321] Step 4:

[0322] The device uses a webcam and microphone to capture facial expressions and voice to collect user emotions. The input is real-time emotional information from the user, and the output is digital data sent to an emotion analysis engine.

[0323] Step 5:

[0324] The server uses an emotion analysis engine to analyze the user's emotional information. It uses natural language processing techniques to quantify the emotional data and, based on the results, generates the data necessary to adjust the interpolation process. The input is the user's emotional data, and the output is data containing adjustment parameters.

[0325] Step 6:

[0326] The server sends prompt messages to the generating AI model using tuning parameters to perform final data completion. An example of a prompt message is, "The user's sentiment rating showed dissatisfaction; please apply a different completion method." The input is the tuning parameters and an initially completed data frame, and the output is the final completed data frame.

[0327] Step 7:

[0328] The server converts the final completed dataframe into CSV format and provides it to the user's terminal in a downloadable format. The input is the final completed dataframe, and the output is a CSV file.

[0329] (Application Example 2)

[0330] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0331] In data processing, imputing missing values ​​is a crucial task. However, traditional methods often process data mechanically without considering user emotions, potentially leading to decreased user satisfaction. Especially with emotionally charged data such as reviews and feedback, data imputation that aligns with user intent and needs is essential. The challenge lies in improving the user experience and enabling more intuitive and effective data analysis.

[0332] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0333] In this invention, the server includes means for detecting missing values ​​from input tabular data, means for imputing missing values ​​using initial imputation means, and means for acquiring user emotion information using emotion analysis technology and imputing based on the emotion information with high accuracy using a generative model. This makes it possible to imputate data in accordance with the user's emotions, thereby improving the user experience.

[0334] A "data processing device" is a device that analyzes and transforms input data and performs information processing according to its purpose.

[0335] A "missing value" is a state in a dataset where a value that should be present is missing.

[0336] "Initial imputation methods" are techniques for imputing missing values ​​using basic methods, such as using the mean for numerical data and the mode for categorical data.

[0337] "Emotion analysis technology" is a technology that analyzes a user's emotions from their facial expressions and voice, and grasps their emotional state in real time.

[0338] A "generative model" is a machine learning model that can learn patterns from large amounts of data and generate new data.

[0339] "High-precision imputation" means inferring and filling in missing values ​​with values ​​that are more accurate and in line with the user's intent.

[0340] "User emotional information" refers to data about the emotional state and characteristics obtained by analyzing the user's facial expressions, voice, etc.

[0341] A "data imputation policy" refers to the basic guidelines and strategies for how to imputate missing values.

[0342] In the system for implementing this invention, the server is primarily responsible for data processing. As a data processing device, the server receives tabular data provided by the user. This data is input in formats such as CSV and managed as a dataframe using the pandas library. The server first detects missing values ​​in this dataframe and performs initial imputation. For numerical data, the mean is used to imputate missing values, and for categorical data, the mode is used.

[0343] Next, the server executes emotion analysis technology. This technology analyzes facial expression and voice data collected from the user's terminal to obtain information about the user's emotions. A dedicated emotion engine is used for emotion analysis, analyzing emotions in real time. Based on this information, the generative AI model adjusts the missing value imputation strategy. Specifically, it dynamically changes the method of imputing missing values, taking into account the user's satisfaction level and emotional state. In addition, the generative AI model utilizes natural language processing technology to perform advanced imputation based on the context of the data.

[0344] The data processed in this way is ultimately output by the server in CSV format, which can be downloaded by the user. Through this process, users can experience customized data supplementation based on their emotions.

[0345] As a concrete example, consider a case where a user writes a product review with partially missing information. The application reads the user's emotion of "satisfaction" from their facial expressions and voice, and fills in the missing parts with positive content. An example of an input prompt for the generative AI model in this case could be: "Review text: This product is easy to use,," "User emotion: Satisfied," "Complete the missing feedback." This enables data completion optimized for each individual user.

[0346] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0347] Step 1:

[0348] The server receives tabular data in CSV format from the terminal. This data is converted into a DataFrame using the pandas library. The input is the raw data provided by the user, and the output is a digital DataFrame. The server analyzes this data and detects missing values.

[0349] Step 2:

[0350] The server imputes detected missing values ​​using an initial imputation mechanism. For numerical data, the mean is used; for categorical data, the mode is used. The input is a data frame containing missing values, and the output is a data frame with initial imputation performed. The server thus ensures the basic integrity of the data.

[0351] Step 3:

[0352] The user's device captures their facial expressions and voice data in real time and sends it to a server. This data is processed using emotion analysis technology and output as the user's emotional information. The input is the user's biometric data, and the output is the analyzed emotional data. The device then uses this to understand the user's emotions.

[0353] Step 4:

[0354] The server adjusts the generative AI model using the provided sentiment data and performs highly accurate imputation of missing values. The generative AI model uses prompts based on the user's sentiment and performs deep contextual analysis using natural language processing techniques. The input is an initially imputed data frame and sentiment data, and the output is an imputed data frame optimized based on sentiment. The server generates data that matches the user's intent.

[0355] Step 5:

[0356] Ultimately, the server outputs a completed and optimized dataframe in CSV format, making it available for download to the user. The input is a dataframe with missing values ​​imputed, and the output is a well-structured data file for the user. The server thus provides an environment where users can easily utilize the data.

[0357] 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 user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0358] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0359] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0360] [Third Embodiment]

[0361] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0362] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0363] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0364] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0365] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0366] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0367] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0368] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0369] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0370] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0371] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0372] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0373] This invention is a system for efficiently and accurately imputing missing values ​​in tabular data input by a data processing device. The server first reads the data received in CSV file format and manages it as a data frame using the pandas library. Missing values ​​are detected from this data frame and imputed using initial statistical methods, using the mean for numerical data and the mode for categorical data.

[0374] Furthermore, the server performs completion using a generative model. Here, a generative model based on natural language processing techniques is used to more precisely complete missing values ​​based on the context within the dataset. In this process, a prompt is generated for each missing location and input into the generative model to generate a value. The generated value is also applied to locations that were not filled in by the initial completion.

[0375] On the terminal, the user inputs data, the server processes it, and then outputs the completed data to the terminal. This output data is saved in CSV file format as an analyzable dataset with all missing values ​​imputed.

[0376] As a concrete example, suppose a user uploads a CSV file containing "customer data" to the server. This file includes attributes such as "age," "gender," and "purchase history." The server detects missing data, performs initial imputation, and then uses a generative model to impute context-dependent missing values ​​such as "purchase history." Finally, the user on their device receives the data with the missing values ​​imputed, which they can then use for further analysis and model building.

[0377] Thus, the present invention provides a data analysis environment with high accuracy and efficiency through the automation of missing value imputation in data processing.

[0378] The following describes the processing flow.

[0379] Step 1:

[0380] The server receives tabular data in CSV format uploaded by the user and reads it as a dataframe using the pandas library. This dataframe contains various attributes, forming a data base that is useful for analysis.

[0381] Step 2:

[0382] The server analyzes the loaded data frame and performs a process to detect missing values ​​in each column. Based on the detection results, it prepares to perform initial imputation.

[0383] Step 3:

[0384] The server imputes missing values ​​in each column of the data frame according to the initial imputation method. Specifically, for numerical data, it calculates the mean and uses that value to fill in the missing values. For categorical data, it identifies the mode and fills in the corresponding values.

[0385] Step 4:

[0386] The server uses a generative model to fill in missing values ​​that were not properly filled in during the initial completion. At this stage, a generative model that considers the context between data attributes, based on natural language processing techniques, is used to create prompts about the missing locations and estimate appropriate values.

[0387] Step 5:

[0388] The server aggregates the data supplemented by the generative model and creates a processed, integrated data frame. This data frame is then modified into a complete tabular format.

[0389] Step 6:

[0390] The server saves the final data frame in CSV format and outputs it to the user's terminal. This output data is provided as a dataset with missing values ​​already imputed, ready for analysis and further data processing.

[0391] (Example 1)

[0392] Next, we will describe Example 1. 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."

[0393] Accurate and efficient imputation of missing values ​​in digital information is a critical challenge in data analysis and machine learning. However, traditional statistical methods of imputation fail to consider the context of the data, potentially resulting in reduced accuracy. Furthermore, manual imputation of missing values ​​is inefficient and time-consuming. Addressing these issues is essential.

[0394] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0395] In this invention, the server includes means for detecting missing values ​​from digital information, means for imputing missing values ​​using initial statistical methods, and means for imputing them with high accuracy based on natural language processing techniques using a generative model. This enables highly accurate imputation of missing values ​​that takes into account the context of the digital information.

[0396] "Digital information" refers to any format of information that is processed, transmitted, or stored electronically. This includes databases, electronic spreadsheet software, and CSV files.

[0397] "Missing values" refer to parts of a dataset where information that should be included is missing for some reason.

[0398] "Initial statistical methods" are basic statistical approaches used to impute missing data, and they involve calculating the mean or mode to fill in the gaps.

[0399] A "generative model" refers to a model that has the ability to learn from large amounts of data and generate new data. It is particularly applied to natural language processing and data completion.

[0400] "Natural language processing technology" refers to the technology that enables computers to understand, generate, and manipulate human language.

[0401] A "prompt statement" refers to a sentence or phrase that is input into a generative model, and is used to instruct the model on a specific task.

[0402] This invention is a system that uses a digital information processing device to efficiently and accurately impute missing values ​​in a dataset. Specific embodiments are described below.

[0403] First, the user provides the server with tabular information as digital data in the form of a CSV file. For example, this data might include "customer information" and attributes such as "age," "gender," and "purchase history." Upon receiving this file, the server uses the pandas library in the Python programming language to manage the data in the form of a data frame. Pandas is software designed to efficiently manipulate large amounts of data.

[0404] The server initiates a process to identify missing values ​​in the data frame, imputing missing values ​​in numerical data by applying the mean and missing values ​​in categorical data by applying the mode. This initial statistical method performs basic missing value imputation.

[0405] Next, the server utilizes a generative AI model to more precisely fill in missing values, taking into account the context of the digital information. This process employs natural language processing techniques to generate prompts for each missing value. An example prompt might be, "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to fill in this information." The generated prompts are input into the generative AI model, and the server calculates the appropriate completion values.

[0406] The server applies the calculated interpolation values ​​to the areas that were incomplete in the initial interpolation, creating a complete dataset. This data is then output to the terminal in CSV format for the user to receive. The user can then use the completed dataset for further detailed analysis or to train machine learning models.

[0407] Thus, the present invention improves the accuracy and efficiency of data analysis by effectively imputing missing values ​​in digital information.

[0408] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0409] Step 1:

[0410] The user uploads tabular digital data to the server in CSV file format. The server receives this file and manages the data in DataFrame format using the pandas library. The input is a user-provided CSV file, and the output is data in DataFrame format.

[0411] Step 2:

[0412] The server analyzes the data frame and detects missing data points. Specifically, it uses pandas functionality to scan each column and identify missing values. In this step, the data frame is used as input, and the output is the location information of the missing values.

[0413] Step 3:

[0414] The server imputes missing values ​​using initial statistical methods. For numerical data, it calculates the mean of the column; for categorical data, it calculates the mode. The input is the location information of the missing values ​​and a data frame, and the output is a data frame with the initial imputation performed.

[0415] Step 4:

[0416] The server uses a generative AI model to precisely impute missing values ​​based on the context within the dataset. For each missing value, it generates a prompt message, such as "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to impute this information," which is then input into the generative model. The input consists of the prompt message and an initially imputed dataframe, while the output is the final imputed dataframe.

[0417] Step 5:

[0418] The server outputs the final completed data in CSV format and sends it to the terminal. The user can then use the received completed data for further analysis or to train machine learning models. The input is the final data frame, and the output is a CSV file usable by the user.

[0419] (Application Example 1)

[0420] Next, we will explain Application Example 1. In the following explanation, 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."

[0421] Missing customer information and purchase history in e-commerce can reduce the accuracy of product recommendations. This can hinder effective marketing to customers and potentially lead to decreased sales. Furthermore, manual data completion is time-consuming and labor-intensive, hindering efficient operations. This invention aims to solve these problems by quickly and accurately completing missing data values ​​and enabling the presentation of personalized product information.

[0422] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0423] In this invention, the server includes means for detecting missing values ​​from input tabular data using a data processing device, means for imputing the detected missing values ​​using an initial imputation means, means for further imputing the data imputed by the initial imputation means with higher accuracy using a generative model, and means for presenting personalized product information based on the processed data. This makes it possible to provide customized recommendations to consumers after imputing missing values.

[0424] A "data processing device" is a device that analyzes input tabular data and detects and imputes missing values.

[0425] "Missing values" refer to a state in a dataset where data that should be present is missing, indicating a lack of information.

[0426] "Initial imputation methods" refer to techniques that use basic statistical methods to perform initial imputation on detected missing values.

[0427] A "generative model" refers to a model that analyzes the context and characteristics of data and uses natural language processing techniques to accurately impute missing values.

[0428] "Personalized product information" refers to information about products and services that are optimized for each individual consumer, based on processed customer data.

[0429] "Means of output" refers to methods or devices for providing the supplemented data to external systems or users.

[0430] The system that realizes this invention operates using a server, user terminals, and a communication network.

[0431] First, the user sends a data file related to the e-commerce site from their device to the server. This data file often contains the customer's personal information and purchase history. Upon receiving this data, the server uses the pandas library to analyze the tabular data and detect missing values.

[0432] For initial completion, the server uses the Scikit-learn library to impute missing values, using the mean for numerical data and the mode for string data. After this initial completion, the server utilizes OpenAI's language generation model to perform more precise, context-based imputation of missing values. Specifically, it generates prompts corresponding to the missing data and inputs these into the generation model to produce more appropriate completion values.

[0433] Ultimately, personalized product information is resent to the user's device based on the data with missing values ​​filled in. This information helps users select the most suitable products on the e-commerce site. Users can use the filled-in data to refer to product ratings and predictions based on their past purchase history.

[0434] As a concrete example, suppose a user has a particular preference for eco-friendly products. In this case, the generative model would fill in the missing parts of the user's purchase history in a similar context and suggest recommended products related to eco-friendly items. An example of a prompt in this process would be, "Generate a recommendation for missing purchase history data for a customer who frequently buys eco-friendly products and has recently viewed outdoor adventure gear." This prompt enables the generative AI model to fill in the missing data and provide the user with appropriate product information.

[0435] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0436] Step 1:

[0437] The user uploads a data file containing customer information and purchase history from their device to the server. The input is a CSV data file, and the output is the original data file stored on the server. In this step, the server confirms the file transmission and notifies the user that it has been successfully received.

[0438] Step 2:

[0439] The server reads the received data file as a DataFrame using the pandas library. The input is data in CSV format, and the output is a DataFrame object. At this point, it detects missing values ​​in the data and lists which items are missing.

[0440] Step 3:

[0441] The server performs initial imputation on dataframes where missing values ​​are detected. Specifically, it uses the mean for numerical data and the mode for string data. It utilizes the Scikit-learn library to perform imputation according to the data type. The input is a dataframe with missing values, and the output is a dataframe with initial imputation applied.

[0442] Step 4:

[0443] The server performs context-based, high-precision missing value imputation on the data after initial imputation. This process utilizes OpenAI's generative AI model. During this process, prompt sentences based on the context of each missing value are generated and input into the generative model. The input consists of the initially imputed data frame and the generated prompt sentences, while the output is the data frame imputed by the generative model.

[0444] Step 5:

[0445] The server creates personalized product information based on complete data supplemented by a generative AI model. This takes into account the consumer's profile and purchase history. The input is a supplemented data frame, and the output is a list of personalized recommended products.

[0446] Step 6:

[0447] The server sends the final output, personalized product information, to the user's terminal. The user receives this information on their terminal and can select from the displayed product list. The input is personalized product information, and the output is the product list displayed on the user's terminal. In this step, the user is notified that new products have been recommended.

[0448] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0449] This invention relates to a system that combines a user emotion engine with a data processing device to accurately impute missing values. The server first receives tabular data provided by the user in CSV format and manages it as a data frame using the pandas library. Missing values ​​in this data frame are detected, and an initial imputation means imputes numerical data with the mean and categorical data with the mode.

[0450] Next, the server utilizes an emotion engine. The emotion engine analyzes user emotions in real time from user interactions and feedback, and incorporates this emotion data into the data completion process. Specifically, it analyzes the user's facial expressions and voice while manipulating data, and extracts emotions using natural language processing techniques. As a result, it fine-tunes the data completion strategy and methods during the final missing value completion using a generative model.

[0451] For example, if a user shows dissatisfaction with the data interpolation results, the emotion engine can sense this and instruct the generative model to repeat the interpolation or try a different estimation approach. This process enables customizable data interpolation that meets the user's needs and expectations.

[0452] Finally, the server outputs a data frame in CSV format, incorporating sentiment information and imputing missing values, which is then provided to the user for download. This processed data improves user satisfaction and is optimized for the specific purpose of data analysis.

[0453] Thus, the present invention provides a more intuitive and effective data processing environment by taking user emotions into consideration during data completion processing.

[0454] The following describes the processing flow.

[0455] Step 1:

[0456] The user uploads tabular data, including missing values, from their terminal to the server. The server reads this data as a DataFrame using the pandas library and stores it in a structured format.

[0457] Step 2:

[0458] The server analyzes each column in the data frame to check for missing values. It detects any identified missing values ​​and uses this information to prepare for initial imputation.

[0459] Step 3:

[0460] The server uses initial imputation methods to fill in missing values. For numerical data, it calculates the mean and uses that result to fill in missing values. For categorical data, it identifies the mode and similarly fills in missing values.

[0461] Step 4:

[0462] The server activates the emotion engine and collects user emotion data through interaction with the user's device. The emotion engine monitors the user's facial expressions, voice, input behavior, etc., and analyzes emotions in real time.

[0463] Step 5:

[0464] The server performs advanced missing value imputation using a generative model. In this process, it considers user sentiment data obtained from the sentiment engine. If the user indicates dissatisfaction or anxiety, the generative model adjusts its imputation methods and algorithms to attempt to imputate the data in a way that aligns with the user's expectations.

[0465] Step 6:

[0466] The server saves the dataframe, with all missing values ​​imputed, in CSV format. This data is now complete with all missing values ​​filled in, and users can download this file from their terminal for further analysis.

[0467] (Example 2)

[0468] Next, we will describe Example 2. 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."

[0469] Traditional methods for imputing missing information in tabular data often failed to consider the data's characteristics and context, making high-precision imputation difficult. Furthermore, a lack of approaches that considered user expectations and emotions meant that data imputation results did not always align with user intent.

[0470] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0471] In this invention, the server includes means for detecting missing information from tabular data input by a data processing unit, means for analyzing the user's emotional information using an emotional analysis engine and adjusting the data completion policy based on the analysis results, and means for completing the data to which the completion policy has been applied with high accuracy using a generative model. This enables high-precision completion of missing information and customization according to the user's emotions.

[0472] A "data processing unit" is a device that has the function of acquiring information from table-formatted data and managing and manipulating it.

[0473] "Missing information" refers to data elements that are missing in tabular data.

[0474] "Initial imputation techniques" are methods for temporarily filling in missing information using basic methods such as the mean or mode.

[0475] A "sentiment analysis engine" is a system that analyzes and extracts a user's emotions in real time from data and operational status.

[0476] A "generative model" is a model that uses machine learning and artificial intelligence technologies to learn data trends and perform more accurate data imputation.

[0477] A "prompt statement" is a form of text used to give instructions or make adjustments to a generative model.

[0478] This invention provides a novel data processing system that takes user sentiment into consideration in order to accurately fill in missing information in tabular data. This system involves a server, a terminal, and a user, each with a specific role.

[0479] When the server receives tabular data, it first uses the pandas library to manage the data as a data frame. It then detects missing information within the data frame and performs initial imputation using the mean for numerical data and the mode for categorical data.

[0480] Next, the device collects user emotion data. Specifically, it uses a webcam and microphone to capture the user's facial expressions and voice. This allows the user's emotional state to be analyzed in real time through an emotion analysis engine.

[0481] The sentiment information obtained by the sentiment analysis engine is sent to the server. Based on this information, the server adjusts the data completion strategy. Here, a generative AI model is used to instruct the model to perform completions that are appropriate to the user's emotions. In this process, specific instructions are given to the generative AI model using prompt statements. For example, an instruction such as "If the user is dissatisfied with the data completion result, try completing it using a different method" might be given.

[0482] Once the data completion is satisfactory to the user, the server outputs the completed dataset in CSV format, making it available for download to the user via their terminal. This achieves data completion that balances the user's subjective satisfaction with the objective accuracy of the data. By incorporating user emotions, this system enables more intuitive and appropriate data processing.

[0483] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0484] Step 1:

[0485] The server receives a tabular data file from the user. When a user uploads data in CSV format via a terminal, the server receives this file and converts it into a DataFrame using the pandas library. The input is a CSV file, and the output is a DataFrame in internal memory.

[0486] Step 2:

[0487] The server detects missing data within the data frame. This step applies an algorithm to check for missing values ​​in each column of the data frame. The input is the data frame, and the output is the data frame containing the missing data.

[0488] Step 3:

[0489] The server uses initial imputation techniques to fill in missing information. For numerical data, it calculates the mean of the column and applies it to the missing parts; for categorical data, it identifies the mode and applies it. The input is a data frame containing missing information, and the output is a data frame with initial imputation completed.

[0490] Step 4:

[0491] The device uses a webcam and microphone to capture facial expressions and voice to collect user emotions. The input is real-time emotional information from the user, and the output is digital data sent to an emotion analysis engine.

[0492] Step 5:

[0493] The server uses an emotion analysis engine to analyze the user's emotional information. It uses natural language processing techniques to quantify the emotional data and, based on the results, generates the data necessary to adjust the interpolation process. The input is the user's emotional data, and the output is data containing adjustment parameters.

[0494] Step 6:

[0495] The server sends prompt messages to the generating AI model using tuning parameters to perform final data completion. An example of a prompt message is, "The user's sentiment rating showed dissatisfaction; please apply a different completion method." The input is the tuning parameters and an initially completed data frame, and the output is the final completed data frame.

[0496] Step 7:

[0497] The server converts the final completed dataframe into CSV format and provides it to the user's terminal in a downloadable format. The input is the final completed dataframe, and the output is a CSV file.

[0498] (Application Example 2)

[0499] Next, we will explain application example 2. In the following explanation, 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."

[0500] In data processing, imputing missing values ​​is a crucial task. However, traditional methods often process data mechanically without considering user emotions, potentially leading to decreased user satisfaction. Especially with emotionally charged data such as reviews and feedback, data imputation that aligns with user intent and needs is essential. The challenge lies in improving the user experience and enabling more intuitive and effective data analysis.

[0501] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0502] In this invention, the server includes means for detecting missing values ​​from input tabular data, means for imputing missing values ​​using initial imputation means, and means for acquiring user emotion information using emotion analysis technology and imputing based on the emotion information with high accuracy using a generative model. This makes it possible to imputate data in accordance with the user's emotions, thereby improving the user experience.

[0503] A "data processing device" is a device that analyzes and transforms input data and performs information processing according to its purpose.

[0504] A "missing value" is a state in a dataset where a value that should be present is missing.

[0505] "Initial imputation methods" are techniques for imputing missing values ​​using basic methods, such as using the mean for numerical data and the mode for categorical data.

[0506] "Emotion analysis technology" is a technology that analyzes a user's emotions from their facial expressions and voice, and grasps their emotional state in real time.

[0507] A "generative model" is a machine learning model that can learn patterns from large amounts of data and generate new data.

[0508] "High-precision imputation" means inferring and filling in missing values ​​with values ​​that are more accurate and in line with the user's intent.

[0509] "User emotional information" refers to data about the emotional state and characteristics obtained by analyzing the user's facial expressions, voice, etc.

[0510] A "data imputation policy" refers to the basic guidelines and strategies for how to imputate missing values.

[0511] In the system for implementing this invention, the server is primarily responsible for data processing. As a data processing device, the server receives tabular data provided by the user. This data is input in formats such as CSV and managed as a dataframe using the pandas library. The server first detects missing values ​​in this dataframe and performs initial imputation. For numerical data, the mean is used to imputate missing values, and for categorical data, the mode is used.

[0512] Next, the server executes emotion analysis technology. This technology analyzes facial expression and voice data collected from the user's terminal to obtain information about the user's emotions. A dedicated emotion engine is used for emotion analysis, analyzing emotions in real time. Based on this information, the generative AI model adjusts the missing value imputation strategy. Specifically, it dynamically changes the method of imputing missing values, taking into account the user's satisfaction level and emotional state. In addition, the generative AI model utilizes natural language processing technology to perform advanced imputation based on the context of the data.

[0513] The data processed in this way is ultimately output by the server in CSV format, which can be downloaded by the user. Through this process, users can experience customized data supplementation based on their emotions.

[0514] As a concrete example, consider a case where a user writes a product review with partially missing information. The application reads the user's emotion of "satisfaction" from their facial expressions and voice, and fills in the missing parts with positive content. An example of an input prompt for the generative AI model in this case could be: "Review text: This product is easy to use,," "User emotion: Satisfied," "Complete the missing feedback." This enables data completion optimized for each individual user.

[0515] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0516] Step 1:

[0517] The server receives tabular data in CSV format from the terminal. This data is converted into a DataFrame using the pandas library. The input is the raw data provided by the user, and the output is a digital DataFrame. The server analyzes this data and detects missing values.

[0518] Step 2:

[0519] The server imputes detected missing values ​​using an initial imputation mechanism. For numerical data, the mean is used; for categorical data, the mode is used. The input is a data frame containing missing values, and the output is a data frame with initial imputation performed. The server thus ensures the basic integrity of the data.

[0520] Step 3:

[0521] The user's device captures their facial expressions and voice data in real time and sends it to a server. This data is processed using emotion analysis technology and output as the user's emotional information. The input is the user's biometric data, and the output is the analyzed emotional data. The device then uses this to understand the user's emotions.

[0522] Step 4:

[0523] The server adjusts the generative AI model using the provided sentiment data and performs highly accurate imputation of missing values. The generative AI model uses prompts based on the user's sentiment and performs deep contextual analysis using natural language processing techniques. The input is an initially imputed data frame and sentiment data, and the output is an imputed data frame optimized based on sentiment. The server generates data that matches the user's intent.

[0524] Step 5:

[0525] Ultimately, the server outputs a completed and optimized dataframe in CSV format, making it available for download to the user. The input is a dataframe with missing values ​​imputed, and the output is a well-structured data file for the user. The server thus provides an environment where users can easily utilize the data.

[0526] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0527] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0528] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0529] [Fourth Embodiment]

[0530] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0531] As shown in Figure 7, the 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.

[0532] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0533] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0534] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0535] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0536] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0537] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive 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 robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0538] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0539] The specific processing program 56 is an example of a "program" relating to the technology of this 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.

[0540] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0541] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. 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 processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0542] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0543] This invention is a system for efficiently and accurately imputing missing values ​​in tabular data input by a data processing device. The server first reads the data received in CSV file format and manages it as a data frame using the pandas library. Missing values ​​are detected from this data frame and imputed using initial statistical methods, using the mean for numerical data and the mode for categorical data.

[0544] Furthermore, the server performs completion using a generative model. Here, a generative model based on natural language processing techniques is used to more precisely complete missing values ​​based on the context within the dataset. In this process, a prompt is generated for each missing location and input into the generative model to generate a value. The generated value is also applied to locations that were not filled in by the initial completion.

[0545] On the terminal, the user inputs data, the server processes it, and then outputs the completed data to the terminal. This output data is saved in CSV file format as an analyzable dataset with all missing values ​​imputed.

[0546] As a concrete example, suppose a user uploads a CSV file containing "customer data" to the server. This file includes attributes such as "age," "gender," and "purchase history." The server detects missing data, performs initial imputation, and then uses a generative model to impute context-dependent missing values ​​such as "purchase history." Finally, the user on their device receives the data with the missing values ​​imputed, which they can then use for further analysis and model building.

[0547] Thus, the present invention provides a data analysis environment with high accuracy and efficiency through the automation of missing value imputation in data processing.

[0548] The following describes the processing flow.

[0549] Step 1:

[0550] The server receives tabular data in CSV format uploaded by the user and reads it as a dataframe using the pandas library. This dataframe contains various attributes, forming a data base that is useful for analysis.

[0551] Step 2:

[0552] The server analyzes the loaded data frame and performs a process to detect missing values ​​in each column. Based on the detection results, it prepares to perform initial imputation.

[0553] Step 3:

[0554] The server imputes missing values ​​in each column of the data frame according to the initial imputation method. Specifically, for numerical data, it calculates the mean and uses that value to fill in the missing values. For categorical data, it identifies the mode and fills in the corresponding values.

[0555] Step 4:

[0556] The server uses a generative model to fill in missing values ​​that were not properly filled in during the initial completion. At this stage, a generative model that considers the context between data attributes, based on natural language processing techniques, is used to create prompts about the missing locations and estimate appropriate values.

[0557] Step 5:

[0558] The server aggregates the data supplemented by the generative model and creates a processed, integrated data frame. This data frame is then modified into a complete tabular format.

[0559] Step 6:

[0560] The server saves the final data frame in CSV format and outputs it to the user's terminal. This output data is provided as a dataset with missing values ​​already imputed, ready for analysis and further data processing.

[0561] (Example 1)

[0562] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0563] Accurate and efficient imputation of missing values ​​in digital information is a critical challenge in data analysis and machine learning. However, traditional statistical methods of imputation fail to consider the context of the data, potentially resulting in reduced accuracy. Furthermore, manual imputation of missing values ​​is inefficient and time-consuming. Addressing these issues is essential.

[0564] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0565] In this invention, the server includes means for detecting missing values ​​from digital information, means for imputing missing values ​​using initial statistical methods, and means for imputing them with high accuracy based on natural language processing techniques using a generative model. This enables highly accurate imputation of missing values ​​that takes into account the context of the digital information.

[0566] "Digital information" refers to any format of information that is processed, transmitted, or stored electronically. This includes databases, electronic spreadsheet software, and CSV files.

[0567] "Missing values" refer to parts of a dataset where information that should be included is missing for some reason.

[0568] "Initial statistical methods" are basic statistical approaches used to impute missing data, and they involve calculating the mean or mode to fill in the gaps.

[0569] A "generative model" refers to a model that has the ability to learn from large amounts of data and generate new data. It is particularly applied to natural language processing and data completion.

[0570] "Natural language processing technology" refers to the technology that enables computers to understand, generate, and manipulate human language.

[0571] A "prompt statement" refers to a sentence or phrase that is input into a generative model, and is used to instruct the model on a specific task.

[0572] This invention is a system that uses a digital information processing device to efficiently and accurately impute missing values ​​in a dataset. Specific embodiments are described below.

[0573] First, the user provides the server with tabular information as digital data in the form of a CSV file. For example, this data might include "customer information" and attributes such as "age," "gender," and "purchase history." Upon receiving this file, the server uses the pandas library in the Python programming language to manage the data in the form of a data frame. Pandas is software designed to efficiently manipulate large amounts of data.

[0574] The server initiates a process to identify missing values ​​in the data frame, imputing missing values ​​in numerical data by applying the mean and missing values ​​in categorical data by applying the mode. This initial statistical method performs basic missing value imputation.

[0575] Next, the server utilizes a generative AI model to more precisely fill in missing values, taking into account the context of the digital information. This process employs natural language processing techniques to generate prompts for each missing value. An example prompt might be, "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to fill in this information." The generated prompts are input into the generative AI model, and the server calculates the appropriate completion values.

[0576] The server applies the calculated interpolation values ​​to the areas that were incomplete in the initial interpolation, creating a complete dataset. This data is then output to the terminal in CSV format for the user to receive. The user can then use the completed dataset for further detailed analysis or to train machine learning models.

[0577] Thus, the present invention improves the accuracy and efficiency of data analysis by effectively imputing missing values ​​in digital information.

[0578] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0579] Step 1:

[0580] The user uploads tabular digital data to the server in CSV file format. The server receives this file and manages the data in DataFrame format using the pandas library. The input is a user-provided CSV file, and the output is data in DataFrame format.

[0581] Step 2:

[0582] The server analyzes the data frame and detects missing data points. Specifically, it uses pandas functionality to scan each column and identify missing values. In this step, the data frame is used as input, and the output is the location information of the missing values.

[0583] Step 3:

[0584] The server imputes missing values ​​using initial statistical methods. For numerical data, it calculates the mean of the column; for categorical data, it calculates the mode. The input is the location information of the missing values ​​and a data frame, and the output is a data frame with the initial imputation performed.

[0585] Step 4:

[0586] The server uses a generative AI model to precisely impute missing values ​​based on the context within the dataset. For each missing value, it generates a prompt message, such as "This dataset contains missing values ​​in the purchase history of male customers in their 30s. Please generate appropriate values ​​to impute this information," which is then input into the generative model. The input consists of the prompt message and an initially imputed dataframe, while the output is the final imputed dataframe.

[0587] Step 5:

[0588] The server outputs the final completed data in CSV format and sends it to the terminal. The user can then use the received completed data for further analysis or to train machine learning models. The input is the final data frame, and the output is a CSV file usable by the user.

[0589] (Application Example 1)

[0590] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0591] Missing customer information and purchase history in e-commerce can reduce the accuracy of product recommendations. This can hinder effective marketing to customers and potentially lead to decreased sales. Furthermore, manual data completion is time-consuming and labor-intensive, hindering efficient operations. This invention aims to solve these problems by quickly and accurately completing missing data values ​​and enabling the presentation of personalized product information.

[0592] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0593] In this invention, the server includes means for detecting missing values ​​from input tabular data using a data processing device, means for imputing the detected missing values ​​using an initial imputation means, means for further imputing the data imputed by the initial imputation means with higher accuracy using a generative model, and means for presenting personalized product information based on the processed data. This makes it possible to provide customized recommendations to consumers after imputing missing values.

[0594] A "data processing device" is a device that analyzes input tabular data and detects and imputes missing values.

[0595] "Missing values" refer to a state in a dataset where data that should be present is missing, indicating a lack of information.

[0596] "Initial imputation methods" refer to techniques that use basic statistical methods to perform initial imputation on detected missing values.

[0597] A "generative model" refers to a model that analyzes the context and characteristics of data and uses natural language processing techniques to accurately impute missing values.

[0598] "Personalized product information" refers to information about products and services that are optimized for each individual consumer, based on processed customer data.

[0599] "Means of output" refers to methods or devices for providing the supplemented data to external systems or users.

[0600] The system that realizes this invention operates using a server, user terminals, and a communication network.

[0601] First, the user sends a data file related to the e-commerce site from their device to the server. This data file often contains the customer's personal information and purchase history. Upon receiving this data, the server uses the pandas library to analyze the tabular data and detect missing values.

[0602] For initial completion, the server uses the Scikit-learn library to impute missing values, using the mean for numerical data and the mode for string data. After this initial completion, the server utilizes OpenAI's language generation model to perform more precise, context-based imputation of missing values. Specifically, it generates prompts corresponding to the missing data and inputs these into the generation model to produce more appropriate completion values.

[0603] Ultimately, personalized product information is resent to the user's device based on the data with missing values ​​filled in. This information helps users select the most suitable products on the e-commerce site. Users can use the filled-in data to refer to product ratings and predictions based on their past purchase history.

[0604] As a concrete example, suppose a user has a particular preference for eco-friendly products. In this case, the generative model would fill in the missing parts of the user's purchase history in a similar context and suggest recommended products related to eco-friendly items. An example of a prompt in this process would be, "Generate a recommendation for missing purchase history data for a customer who frequently buys eco-friendly products and has recently viewed outdoor adventure gear." This prompt enables the generative AI model to fill in the missing data and provide the user with appropriate product information.

[0605] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0606] Step 1:

[0607] The user uploads a data file containing customer information and purchase history from their device to the server. The input is a CSV data file, and the output is the original data file stored on the server. In this step, the server confirms the file transmission and notifies the user that it has been successfully received.

[0608] Step 2:

[0609] The server reads the received data file as a DataFrame using the pandas library. The input is data in CSV format, and the output is a DataFrame object. At this point, it detects missing values ​​in the data and lists which items are missing.

[0610] Step 3:

[0611] The server performs initial imputation on dataframes where missing values ​​are detected. Specifically, it uses the mean for numerical data and the mode for string data. It utilizes the Scikit-learn library to perform imputation according to the data type. The input is a dataframe with missing values, and the output is a dataframe with initial imputation applied.

[0612] Step 4:

[0613] The server performs context-based, high-precision missing value imputation on the data after initial imputation. This process utilizes OpenAI's generative AI model. During this process, prompt sentences based on the context of each missing value are generated and input into the generative model. The input consists of the initially imputed data frame and the generated prompt sentences, while the output is the data frame imputed by the generative model.

[0614] Step 5:

[0615] The server creates personalized product information based on complete data supplemented by a generative AI model. This takes into account the consumer's profile and purchase history. The input is a supplemented data frame, and the output is a list of personalized recommended products.

[0616] Step 6:

[0617] The server sends the final output, personalized product information, to the user's terminal. The user receives this information on their terminal and can select from the displayed product list. The input is personalized product information, and the output is the product list displayed on the user's terminal. In this step, the user is notified that new products have been recommended.

[0618] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0619] This invention relates to a system that combines a user emotion engine with a data processing device to accurately impute missing values. The server first receives tabular data provided by the user in CSV format and manages it as a data frame using the pandas library. Missing values ​​in this data frame are detected, and an initial imputation means imputes numerical data with the mean and categorical data with the mode.

[0620] Next, the server utilizes an emotion engine. The emotion engine analyzes user emotions in real time from user interactions and feedback, and incorporates this emotion data into the data completion process. Specifically, it analyzes the user's facial expressions and voice while manipulating data, and extracts emotions using natural language processing techniques. As a result, it fine-tunes the data completion strategy and methods during the final missing value completion using a generative model.

[0621] For example, if a user shows dissatisfaction with the data interpolation results, the emotion engine can sense this and instruct the generative model to repeat the interpolation or try a different estimation approach. This process enables customizable data interpolation that meets the user's needs and expectations.

[0622] Finally, the server outputs a data frame in CSV format, incorporating sentiment information and imputing missing values, which is then provided to the user for download. This processed data improves user satisfaction and is optimized for the specific purpose of data analysis.

[0623] Thus, the present invention provides a more intuitive and effective data processing environment by taking user emotions into consideration during data completion processing.

[0624] The following describes the processing flow.

[0625] Step 1:

[0626] The user uploads tabular data, including missing values, from their terminal to the server. The server reads this data as a DataFrame using the pandas library and stores it in a structured format.

[0627] Step 2:

[0628] The server analyzes each column in the data frame to check for missing values. It detects any identified missing values ​​and uses this information to prepare for initial imputation.

[0629] Step 3:

[0630] The server uses initial imputation methods to fill in missing values. For numerical data, it calculates the mean and uses that result to fill in missing values. For categorical data, it identifies the mode and similarly fills in missing values.

[0631] Step 4:

[0632] The server activates the emotion engine and collects user emotion data through interaction with the user's device. The emotion engine monitors the user's facial expressions, voice, input behavior, etc., and analyzes emotions in real time.

[0633] Step 5:

[0634] The server performs advanced missing value imputation using a generative model. In this process, it considers user sentiment data obtained from the sentiment engine. If the user indicates dissatisfaction or anxiety, the generative model adjusts its imputation methods and algorithms to attempt to imputate the data in a way that aligns with the user's expectations.

[0635] Step 6:

[0636] The server saves the dataframe, with all missing values ​​imputed, in CSV format. This data is now complete with all missing values ​​filled in, and users can download this file from their terminal for further analysis.

[0637] (Example 2)

[0638] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0639] Traditional methods for imputing missing information in tabular data often failed to consider the data's characteristics and context, making high-precision imputation difficult. Furthermore, a lack of approaches that considered user expectations and emotions meant that data imputation results did not always align with user intent.

[0640] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0641] In this invention, the server includes means for detecting missing information from tabular data input by a data processing unit, means for analyzing the user's emotional information using an emotional analysis engine and adjusting the data completion policy based on the analysis results, and means for completing the data to which the completion policy has been applied with high accuracy using a generative model. This enables high-precision completion of missing information and customization according to the user's emotions.

[0642] A "data processing unit" is a device that has the function of acquiring information from table-formatted data and managing and manipulating it.

[0643] "Missing information" refers to data elements that are missing in tabular data.

[0644] "Initial imputation techniques" are methods for temporarily filling in missing information using basic methods such as the mean or mode.

[0645] A "sentiment analysis engine" is a system that analyzes and extracts a user's emotions in real time from data and operational status.

[0646] A "generative model" is a model that uses machine learning and artificial intelligence technologies to learn data trends and perform more accurate data imputation.

[0647] A "prompt statement" is a form of text used to give instructions or make adjustments to a generative model.

[0648] This invention provides a novel data processing system that takes user sentiment into consideration in order to accurately fill in missing information in tabular data. This system involves a server, a terminal, and a user, each with a specific role.

[0649] When the server receives tabular data, it first uses the pandas library to manage the data as a data frame. It then detects missing information within the data frame and performs initial imputation using the mean for numerical data and the mode for categorical data.

[0650] Next, the device collects user emotion data. Specifically, it uses a webcam and microphone to capture the user's facial expressions and voice. This allows the user's emotional state to be analyzed in real time through an emotion analysis engine.

[0651] The sentiment information obtained by the sentiment analysis engine is sent to the server. Based on this information, the server adjusts the data completion strategy. Here, a generative AI model is used to instruct the model to perform completions that are appropriate to the user's emotions. In this process, specific instructions are given to the generative AI model using prompt statements. For example, an instruction such as "If the user is dissatisfied with the data completion result, try completing it using a different method" might be given.

[0652] Once the data completion is satisfactory to the user, the server outputs the completed dataset in CSV format, making it available for download to the user via their terminal. This achieves data completion that balances the user's subjective satisfaction with the objective accuracy of the data. By incorporating user emotions, this system enables more intuitive and appropriate data processing.

[0653] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0654] Step 1:

[0655] The server receives a tabular data file from the user. When a user uploads data in CSV format via a terminal, the server receives this file and converts it into a DataFrame using the pandas library. The input is a CSV file, and the output is a DataFrame in internal memory.

[0656] Step 2:

[0657] The server detects missing data within the data frame. This step applies an algorithm to check for missing values ​​in each column of the data frame. The input is the data frame, and the output is the data frame containing the missing data.

[0658] Step 3:

[0659] The server uses initial imputation techniques to fill in missing information. For numerical data, it calculates the mean of the column and applies it to the missing parts; for categorical data, it identifies the mode and applies it. The input is a data frame containing missing information, and the output is a data frame with initial imputation completed.

[0660] Step 4:

[0661] The device uses a webcam and microphone to capture facial expressions and voice to collect user emotions. The input is real-time emotional information from the user, and the output is digital data sent to an emotion analysis engine.

[0662] Step 5:

[0663] The server uses an emotion analysis engine to analyze the user's emotional information. It uses natural language processing techniques to quantify the emotional data and, based on the results, generates the data necessary to adjust the interpolation process. The input is the user's emotional data, and the output is data containing adjustment parameters.

[0664] Step 6:

[0665] The server sends prompt messages to the generating AI model using tuning parameters to perform final data completion. An example of a prompt message is, "The user's sentiment rating showed dissatisfaction; please apply a different completion method." The input is the tuning parameters and an initially completed data frame, and the output is the final completed data frame.

[0666] Step 7:

[0667] The server converts the final completed dataframe into CSV format and provides it to the user's terminal in a downloadable format. The input is the final completed dataframe, and the output is a CSV file.

[0668] (Application Example 2)

[0669] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0670] In data processing, imputing missing values ​​is a crucial task. However, traditional methods often process data mechanically without considering user emotions, potentially leading to decreased user satisfaction. Especially with emotionally charged data such as reviews and feedback, data imputation that aligns with user intent and needs is essential. The challenge lies in improving the user experience and enabling more intuitive and effective data analysis.

[0671] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0672] In this invention, the server includes means for detecting missing values ​​from input tabular data, means for imputing missing values ​​using initial imputation means, and means for acquiring user emotion information using emotion analysis technology and imputing based on the emotion information with high accuracy using a generative model. This makes it possible to imputate data in accordance with the user's emotions, thereby improving the user experience.

[0673] A "data processing device" is a device that analyzes and transforms input data and performs information processing according to its purpose.

[0674] A "missing value" is a state in a dataset where a value that should be present is missing.

[0675] "Initial imputation methods" are techniques for imputing missing values ​​using basic methods, such as using the mean for numerical data and the mode for categorical data.

[0676] "Emotion analysis technology" is a technology that analyzes a user's emotions from their facial expressions and voice, and grasps their emotional state in real time.

[0677] A "generative model" is a machine learning model that can learn patterns from large amounts of data and generate new data.

[0678] "High-precision imputation" means inferring and filling in missing values ​​with values ​​that are more accurate and in line with the user's intent.

[0679] "User emotional information" refers to data about the emotional state and characteristics obtained by analyzing the user's facial expressions, voice, etc.

[0680] A "data imputation policy" refers to the basic guidelines and strategies for how to imputate missing values.

[0681] In the system for implementing this invention, the server is primarily responsible for data processing. As a data processing device, the server receives tabular data provided by the user. This data is input in formats such as CSV and managed as a dataframe using the pandas library. The server first detects missing values ​​in this dataframe and performs initial imputation. For numerical data, the mean is used to imputate missing values, and for categorical data, the mode is used.

[0682] Next, the server executes emotion analysis technology. This technology analyzes facial expression and voice data collected from the user's terminal to obtain information about the user's emotions. A dedicated emotion engine is used for emotion analysis, analyzing emotions in real time. Based on this information, the generative AI model adjusts the missing value imputation strategy. Specifically, it dynamically changes the method of imputing missing values, taking into account the user's satisfaction level and emotional state. In addition, the generative AI model utilizes natural language processing technology to perform advanced imputation based on the context of the data.

[0683] The data processed in this way is ultimately output by the server in CSV format, which can be downloaded by the user. Through this process, users can experience customized data supplementation based on their emotions.

[0684] As a concrete example, consider a case where a user writes a product review with partially missing information. The application reads the user's emotion of "satisfaction" from their facial expressions and voice, and fills in the missing parts with positive content. An example of an input prompt for the generative AI model in this case could be: "Review text: This product is easy to use,," "User emotion: Satisfied," "Complete the missing feedback." This enables data completion optimized for each individual user.

[0685] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0686] Step 1:

[0687] The server receives tabular data in CSV format from the terminal. This data is converted into a DataFrame using the pandas library. The input is the raw data provided by the user, and the output is a digital DataFrame. The server analyzes this data and detects missing values.

[0688] Step 2:

[0689] The server imputes detected missing values ​​using an initial imputation mechanism. For numerical data, the mean is used; for categorical data, the mode is used. The input is a data frame containing missing values, and the output is a data frame with initial imputation performed. The server thus ensures the basic integrity of the data.

[0690] Step 3:

[0691] The user's device captures their facial expressions and voice data in real time and sends it to a server. This data is processed using emotion analysis technology and output as the user's emotional information. The input is the user's biometric data, and the output is the analyzed emotional data. The device then uses this to understand the user's emotions.

[0692] Step 4:

[0693] The server adjusts the generative AI model using the provided sentiment data and performs highly accurate imputation of missing values. The generative AI model uses prompts based on the user's sentiment and performs deep contextual analysis using natural language processing techniques. The input is an initially imputed data frame and sentiment data, and the output is an imputed data frame optimized based on sentiment. The server generates data that matches the user's intent.

[0694] Step 5:

[0695] Ultimately, the server outputs a completed and optimized dataframe in CSV format, making it available for download to the user. The input is a dataframe with missing values ​​imputed, and the output is a well-structured data file for the user. The server thus provides an environment where users can easily utilize the data.

[0696] 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 controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0697] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0698] 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 this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0699] Furthermore, the emotion identification model 59, acting 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 a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0700] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0701] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0702] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0703] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0704] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is 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 the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0705] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0706] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0707] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0708] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0709] 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.

[0710] Furthermore, it is not necessary to store the entirety 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 the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0711] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0712] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of 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). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0713] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0714] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0715] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0716] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[0717] The following is further disclosed regarding the embodiments described above.

[0718] (Claim 1)

[0719] The data processing device includes means for detecting missing values ​​from input tabular data,

[0720] A means for compensating for the detected missing values ​​using an initial imputation means,

[0721] A means for further enhancing the data enhanced by the initial enhancement means using a generative model,

[0722] A means of outputting processed data,

[0723] A system that includes this.

[0724] (Claim 2)

[0725] The system according to claim 1, wherein the initial imputation means imputes missing values ​​using the mean for numerical data and the mode for categorical data.

[0726] (Claim 3)

[0727] The system according to claim 1, wherein the generative model uses natural language processing techniques to impute missing values ​​based on the context of the data.

[0728] "Example 1"

[0729] (Claim 1)

[0730] A data processing device provides means for detecting missing values ​​from input digital information,

[0731] Means for imputing the detected missing values ​​using an initial statistical method,

[0732] A means for further enhancing digital information, which has been supplemented by the aforementioned initial statistical method based on natural language processing technology, using a generative model,

[0733] A means of outputting processed information,

[0734] A system that includes this.

[0735] (Claim 2)

[0736] The system according to claim 1, wherein the initial statistical method imputes missing values ​​using the mean for numerical data and the mode for piecewise data.

[0737] (Claim 3)

[0738] The system according to claim 1, wherein the generation model imputes missing values ​​based on the context of the digital information using prompt statements.

[0739] "Application Example 1"

[0740] (Claim 1)

[0741] The data processing device includes means for detecting missing values ​​from input tabular data,

[0742] A means for compensating for the detected missing values ​​using an initial imputation means,

[0743] A means for further enhancing the data enhanced by the initial enhancement means using a generative model,

[0744] A means of presenting personalized product information based on processed data,

[0745] A means of outputting processed data,

[0746] A system that includes this.

[0747] (Claim 2)

[0748] The system according to claim 1, wherein the initial imputation means imputes missing values ​​using the mean for numerical data and the mode for categorical data.

[0749] (Claim 3)

[0750] The system according to claim 1, wherein the generative model uses language processing techniques to impute missing values ​​based on the context of the data.

[0751] "Example 2 of combining an emotion engine"

[0752] (Claim 1)

[0753] The data processing unit includes means for detecting missing information from input table-formatted data,

[0754] Means for supplementing the detected missing information using initial supplementation techniques,

[0755] A means of analyzing user emotional information using an emotion analysis engine and adjusting the data supplementation strategy based on the analysis results,

[0756] A method of using generative models to accurately complete data to which a completion policy has been applied,

[0757] A means of outputting processed information,

[0758] A system that includes this.

[0759] (Claim 2)

[0760] The system according to claim 1, wherein the initial imputation technique uses the mean value for numerical information and the mode for classification information to imputate missing information.

[0761] (Claim 3)

[0762] The system according to claim 1, wherein the generation model adjusts the completion policy using prompt statements and completes missing information.

[0763] "Application example 2 when combining with an emotional engine"

[0764] (Claim 1)

[0765] The data processing device includes means for detecting missing values ​​from input tabular data,

[0766] A means for compensating for the detected missing values ​​using an initial imputation means,

[0767] Using emotion analysis technology, we obtain the user's emotional information.

[0768] A means for further enhancing the data enhanced by the initial enhancement means using a generative model, based on acquired sentiment information,

[0769] A means of outputting processed data,

[0770] A system that includes this.

[0771] (Claim 2)

[0772] The system according to claim 1, wherein the initial imputation means imputes missing values ​​using the mean for numerical data and the mode for categorical data.

[0773] (Claim 3)

[0774] The system according to claim 1, wherein the generative model uses natural language processing techniques to impute missing values ​​based on the context of the data and adjusts the data imputation strategy using prompt sentences that correspond to the user's emotional state. [Explanation of Symbols]

[0775] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. The data processing device includes means for detecting missing values ​​from input tabular data, A means for compensating for the detected missing values ​​using an initial imputation means, A means for further enhancing the data enhanced by the initial enhancement means using a generative model, A means of outputting processed data, A system that includes this.

2. The system according to claim 1, wherein the initial imputation means imputes missing values ​​using the mean for numerical data and the mode for categorical data.

3. The system according to claim 1, wherein the generative model uses natural language processing techniques to impute missing values ​​based on the context of the data.

Images