system

Abstract

We provide the system. [Solution] A data collection means that receives data provided by the user and temporarily stores it, A preprocessing means for the received data, which includes imputing missing values, correcting outliers, and removing duplicate data. A means for applying a machine learning algorithm that extracts features from preprocessed data, selects optimal parameters, and performs analysis, An analysis result generation means that outputs the analysis results obtained by the algorithm in a format that is easy for the user to understand, A feedback mechanism that records user activity results and sends them to a server, An algorithm update method that retrains the algorithm based on feedback 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 persona chatbot control method performed by at least one processor, including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In a conventional data analysis system, it is difficult to process a large amount of data quickly and accurately, and a lot of time and labor are required for preprocessing such as correcting missing values and outliers of data and eliminating duplicate data. Also, in the application of machine learning algorithms, it is difficult to extract appropriate feature quantities and select optimal parameters, resulting in a problem of low accuracy of analysis results. Therefore, there is a need for a system that improves the efficiency and accuracy of data analysis and enhances user convenience.

Means for Solving the Problems

[0005] The present invention solves the above problems with a system including data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, and algorithm update means. The data collection means collects data provided by the user, and the preprocessing means performs imputation of missing values, correction of outliers, and elimination of duplicate data. Next, the machine learning algorithm application means extracts features from the preprocessed data and selects the optimal parameters. The analysis result generation means outputs the results in a format that is easy for the user to understand, and the feedback means records the user's activity results and sends them to the server. Finally, the algorithm update means retrains the algorithm based on this feedback data to improve the accuracy of the next analysis. In this way, rapid and accurate analysis of data becomes possible, greatly improving user convenience.

[0006] A "data collection method" is a function for receiving and temporarily storing data provided by users.

[0007] "Preprocessing means" refers to functions for processing received data, such as imputing missing values, correcting outliers, and removing duplicate data.

[0008] "Machine learning algorithm application means" refers to a function that extracts features from preprocessed data, selects the optimal parameters, and applies an algorithm for analysis.

[0009] The "analysis result generation means" is a function that outputs the analysis results obtained by the algorithm in a format that is easy for the user to understand.

[0010] A "feedback mechanism" is a function that records the user's activity results and sends them to the server.

[0011] The "algorithm update mechanism" is a function that retrains the algorithm based on feedback data to improve the accuracy of the next analysis. [Brief explanation of the drawing]

[0012] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This 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] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This 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] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This 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] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14] This is a sequence diagram showing the processing flow of the data processing system in Application Example 2, which combines an emotion engine. [Modes for carrying out the invention]

[0013] An example of an embodiment of the system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0014] First, the terms used in the following description will be explained.

[0015] In the following embodiments, a numbered 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.

[0016] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0017] In the following embodiments, a numbered 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, and the like.

[0018] 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).

[0019] 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."

[0020] [First Embodiment]

[0021] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0022] 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.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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.

[0028] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] 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".

[0033] The invention of this system involves implementing a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, and algorithm update means. Each means works in cooperation to rapidly and accurately analyze large amounts of data, improving user convenience.

[0034] System program processing overview

[0035] Data collection

[0036] The server receives data provided by users. This data comes in various forms, such as sales data and customer support data, and is stored temporarily.

[0037] Pre-treatment

[0038] The terminal performs preprocessing on the data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0039] Application of machine learning algorithms

[0040] The server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. The results of this analysis form the basis for providing specific results tailored to the user's needs.

[0041] Generation of analysis results

[0042] The server generates reports, graphs, and other data in a user-friendly format based on analysis results obtained by machine learning algorithms. These generated analysis results are provided to the sales team and customer support team for use in their subsequent activities.

[0043] feedback

[0044] The terminal records user activity results and sends them to the server. For example, it collects information such as whether a product was purchased as a result of sales activities or whether an inquiry was resolved as a result of customer support. This information is used for subsequent analysis.

[0045] Algorithm update

[0046] The server retrains its machine learning algorithm based on feedback data, improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, gradually increasing the accuracy of its analyses.

[0047] Specific example

[0048] Example 1: Analysis of sales data

[0049] 1. Data Collection: The server receives customer information and transaction history from the sales team.

[0050] 2. Preprocessing: The terminal imputes missing values ​​in the transaction history with the average value of past data, corrects outliers, and integrates duplicate customer information.

[0051] 3. Application of machine learning algorithms: The server analyzes the customer's purchase history to identify purchasing patterns and generates a list of products to suggest next.

[0052] 4. Generation of analysis results: The server generates a list of optimal product suggestions and provides them to the sales team.

[0053] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result on the terminal. The feedback data is sent to the server.

[0054] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0055] Example 2: Analysis of customer support data

[0056] 1. Data collection: The server receives inquiry data from customer support.

[0057] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes.

[0058] 3. Application of machine learning algorithms: The server analyzes the inquiry history and proposes the optimal response method.

[0059] 4. Generation of analysis results: The server generates recommended actions and provides them to the support staff.

[0060] 5. Feedback: The user (support staff) implements the suggested solution and records the results on their device. The feedback data is sent to the server.

[0061] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0062] Thus, the system of the present invention efficiently executes a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, providing valuable information to the user.

[0063] The following describes the processing flow.

[0064] Step 1:

[0065] The server receives data from users. This data can be in various formats, such as sales data or customer support data. The server temporarily stores the received data.

[0066] Step 2:

[0067] The terminal performs preprocessing on the data received from the server. The terminal checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data and creates a consistent dataset.

[0068] Step 3:

[0069] The server applies machine learning algorithms to extract features from preprocessed data. Specifically, it analyzes data patterns and automatically selects important features. During this process, various parameters are adjusted and optimized.

[0070] Step 4:

[0071] The server generates reports and graphs to provide to users based on the analysis results. For example, it provides the sales team with a list of products to propose next, and the customer support team with the best course of action.

[0072] Step 5:

[0073] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these analysis results to carry out actual activities. For example, they might propose products to customers based on the suggestion list, or implement the most appropriate countermeasures.

[0074] Step 6:

[0075] The device records user activity. For example, it collects information such as whether a product was purchased or whether an inquiry was resolved. This information is used for subsequent analysis.

[0076] Step 7:

[0077] The terminal sends the recorded feedback data to the server. The server analyzes the feedback data and retrains the algorithm. Specifically, it compares the initial analysis results with the feedback results and makes adjustments necessary to improve the accuracy of the model.

[0078] Step 8:

[0079] The server prepares to apply the updated algorithm to the next analysis. This allows the system to continuously improve itself, gradually increasing the accuracy of the analysis.

[0080] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately, provide useful information to the user, and continuously improve the accuracy of the algorithm.

[0081] (Example 1)

[0082] 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."

[0083] Conventional data analysis systems struggle to analyze large amounts of data quickly and accurately, and are hampered by insufficient data preprocessing and feedback on analysis results. Furthermore, they lack mechanisms to effectively incorporate user feedback and improve analysis accuracy.

[0084] 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.

[0085] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by the terminal, data analysis means by the server, and user feedback recording means. This makes it possible to analyze large amounts of data quickly and accurately, and to improve the accuracy of the analysis by incorporating feedback from the user.

[0086] A "data collection method" is a means by which a server receives and temporarily stores data provided by a user.

[0087] "Preprocessing means" are methods for improving data quality by imputing missing data values, correcting outliers, and eliminating duplicate data.

[0088] A "machine learning algorithm application method" is a method of performing analysis by applying a machine learning algorithm to pre-processed data, extracting data features, and optimizing parameters.

[0089] "Analysis result generation means" refers to a means of generating reports and graphs in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0090] A "feedback mechanism" is a means of recording the user's activity results and sending that data to a server.

[0091] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0092] "Data preprocessing means by the terminal" refers to means by which the terminal preprocesses data received from the server, specifically performing missing value imputation, outlier correction, and duplicate data removal.

[0093] A "server-based data analysis method" refers to a method in which a server receives pre-processed data and performs analysis by applying machine learning algorithms.

[0094] A "user feedback recording means" is a means for a user to record the results of their activities, input that feedback data into a terminal, and send it to a server.

[0095] This system invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by a terminal, data analysis means by a server, and user feedback recording means. The program processing of this system will be described below in natural language.

[0096] First, the server receives a wide variety of data provided by the user. This data, such as sales data and customer support data, is temporarily stored. Next, the server sends the stored data to the terminal for preprocessing.

[0097] The terminal performs preprocessing on the data received from the server. Preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis. Once preprocessing is complete, the data is sent back to the server.

[0098] Next, the server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features and selects optimal parameters. The analysis results form the basis for providing specific results tailored to the user's needs.

[0099] The server generates reports and graphs in an easy-to-understand format for users based on analysis results obtained from machine learning algorithms. The generated analysis results are shared with the sales team and customer support team and used for subsequent activities. Users perform activities based on the reports provided by the server and record the results on their devices. The devices send the activity results to the server as feedback data.

[0100] The feedback data is used on the server to retrain the machine learning algorithm, improving the accuracy of subsequent analyses. This process allows the system to continuously improve itself, analyze large amounts of data quickly and accurately, and incorporate user feedback to further enhance analysis accuracy.

[0101] As a concrete example, let's consider the analysis of sales data. When a user uploads sales data to the system, the server receives the data. The server sends the received data to a terminal for preprocessing, where the terminal performs tasks such as imputing missing values, correcting outliers, and merging duplicate data. The preprocessed data is sent back to the server, which performs analysis using a machine learning algorithm. The analysis results are generated as a report and provided to the user. The user acts based on this report, records the results on their terminal, and sends them back to the server as feedback data. The server uses this feedback data to retrain its algorithm and improve the accuracy of the next analysis.

[0102] Example of a prompt:

[0103] Collect sales data, apply an algorithm to analyze purchasing patterns based on customer information and transaction history, and generate a list of products to recommend next. Output the results as a report and send it to the user via email. Retrain the algorithm using user feedback data to improve its accuracy.

[0104] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0105] System program processing flow

[0106] Step 1: Data Collection

[0107] 1. Providing user data

[0108] Users input necessary information, such as sales data and customer support data, into the system.

[0109] Input: Sales data, customer support data

[0110] Specific action: The sales representative uploads customer transaction data as a text file.

[0111] 2. Receiving data from the server

[0112] The server receives the uploaded data and stores it temporarily.

[0113] Output: Temporarily saved data file

[0114] Specific action: Receive a file named 'customers.csv' and save it to a specific folder on the server.

[0115] Step 2: Data Preprocessing

[0116] 1. Server data transmission

[0117] The server sends the stored data to the terminal for preprocessing.

[0118] Input: Temporarily saved data file

[0119] Output: Data file sent to the terminal

[0120] Specific action: Send 'customers.csv' to terminal A.

[0121] 2. Data preprocessing at the terminal

[0122] The terminal preprocesses the received data.

[0123] Input: 'customers.csv'

[0124] Output: Preprocessed data file ('data_processed.csv')

[0125] Specific actions: Imput missing values ​​with the mean of past data, correct outliers, and remove duplicate data.

[0126] Step 3: Applying the machine learning algorithm

[0127] 1. Data transmission from the device

[0128] The terminal sends the pre-processed data to the server.

[0129] Input: Preprocessed data file ('data_processed.csv')

[0130] Output: Preprocessed data file sent to the server

[0131] Specific action: Send 'data_processed.csv' to the server.

[0132] 2. Server algorithm application

[0133] The server analyzes the data using machine learning algorithms.

[0134] Input: Preprocessed data file ('data_processed.csv')

[0135] Output: Analysis results (list of proposed products, corresponding methods, etc.)

[0136] Specific operation: Analyzes purchasing patterns using 'data_processed.csv' and generates an optimal product suggestion list.

[0137] Step 4: Generating analysis results

[0138] 1. Server report generation

[0139] The server generates reports and graphs based on the analysis results.

[0140] Input: Analysis results

[0141] Output: Reports and graphs (e.g., 'proposal_list_report.pdf')

[0142] Specific operation: Output the analysis results as a PDF report, including graphed data.

[0143] 2. Server data transmission

[0144] The server sends the generated reports and graphs to the user.

[0145] Input: Reports and graphs

[0146] Output: Reports and graphs sent to the user

[0147] Specific action: Send 'proposal_list_report.pdf' to the user's email address.

[0148] Step 5: Feedback

[0149] 1. User activity log

[0150] Users perform activities based on the generated reports and record the results on their devices.

[0151] Input: Activity results based on reports and graphs

[0152] Output: Recorded activity data (e.g., 'activity_log.xlsx')

[0153] Specific operation: A sales representative proposes a product to a customer and enters the result (whether or not a purchase was made) into the terminal.

[0154] 2. Sending feedback data from the device

[0155] The device sends feedback data to the server.

[0156] Input: Recorded activity data (e.g., 'activity_log.xlsx')

[0157] Output: Feedback data sent to the server

[0158] Specific action: Send 'activity_log.xlsx' to the server.

[0159] Step 6: Update the algorithm

[0160] 1. Server retraining

[0161] The server retrains the machine learning algorithm based on the feedback data, improving the model's accuracy.

[0162] Input: Feedback data (e.g., 'activity_log.xlsx')

[0163] Output: Updated machine learning model

[0164] Specific operation: Readjust the algorithm parameters using 'activity_log.xlsx' and generate a new model.

[0165] (Application Example 1)

[0166] 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."

[0167] Traditional e-commerce sites lacked adequate systems for recommending the most suitable products based on user purchasing behavior and browsing history, making it difficult to accurately recommend products that would be beneficial to the user. As a result, user convenience decreased, and purchasing intent tended to decline. Furthermore, there was a lack of mechanisms to update algorithms and improve recommendation accuracy based on feedback.

[0168] 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.

[0169] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means for recommending the optimal product based on the user's purchasing behavior and browsing history. This makes it possible to accurately recommend the optimal product based on the user's purchasing behavior and browsing history, improving user convenience and potentially increasing purchasing intent. Furthermore, the accuracy of the algorithm can be continuously improved using feedback data.

[0170] "Data collection methods" refer to means of collecting various forms of data, such as users' purchasing behavior and browsing history.

[0171] "Preprocessing means" are methods for improving the quality of collected data by imputing missing values, correcting outliers, and eliminating duplicate data.

[0172] A "machine learning algorithm application method" is a means of applying a machine learning algorithm to pre-processed data, extracting data features, selecting optimal parameters, and performing analysis.

[0173] "Analysis result generation means" refers to a means of generating reports, recommendation lists, and other similar documents in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0174] A "feedback mechanism" is a means of improving the accuracy of analysis results by recording the user's activity results and sending those results to a server.

[0175] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0176] A "recommendation method" is a means of recommending the most suitable products based on a user's purchasing behavior and browsing history.

[0177] A system for carrying out this invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means.

[0178] The server has data collection capabilities to gather various types of data, such as user purchasing behavior and browsing history. This data is temporarily stored and used for subsequent data analysis.

[0179] The terminal has a preprocessing mechanism that performs preprocessing on data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0180] The server has a means of applying machine learning algorithms to pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. Specifically, it analyzes user interests and purchasing trends from purchase history and browsing history.

[0181] The server has a means of generating analysis results that, based on the analysis results obtained by machine learning algorithms, produce reports and recommendation lists in a user-friendly format. The generated analysis results are provided to the user for viewing and recommend products they should purchase next.

[0182] The feedback mechanism records user activity and sends it to the server. For example, it collects information such as whether the user actually purchased or viewed a recommended product. This information is used for subsequent analysis.

[0183] The algorithm update mechanism plays a role in retraining the machine learning algorithm based on feedback data, thereby improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, and the accuracy of the analysis gradually increases.

[0184] The hardware required includes a server for data collection, preprocessing, and training and application of machine learning models. Furthermore, smartphones are used to collect user purchase and browsing history and display the analysis results. The software used also includes the following:

[0185] Python: The main programming language used in programs.

[0186] pandas: A library for data collection and preprocessing.

[0187] scikit-learn: A library used to apply machine learning algorithms.

[0188] The following is a specific example. In this example, a system is designed to recommend the most suitable products based on the user's purchase and browsing history.

[0189] ---

[0190] Design a system that recommends the most suitable products based on a user's purchase and browsing history. This system will collect large amounts of data, perform preprocessing, apply machine learning algorithms, and provide valuable information to the user. As an example, show how to recommend new products based on the categories of products the user has previously purchased.

[0191] ---

[0192] By implementing such a system, it is possible to increase users' purchasing intent and improve the convenience of online shopping sites.

[0193] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0194] Step 1:

[0195] The server collects data. Specifically, the server collects and temporarily stores users' purchasing behavior and browsing history. This includes information about the user's past purchase history and the product pages they have viewed. The input data is user behavior data, and the output is the temporarily stored raw data.

[0196] Step 2:

[0197] The terminal performs preprocessing. Using the raw data received from the server, it imputes missing values, corrects outliers, and removes duplicate data. Specifically, missing values ​​are imputed with the mean or median of past data, and outliers are corrected based on predefined rules. The input data is temporarily stored raw data, and the output is preprocessed data.

[0198] Step 3:

[0199] The server applies machine learning algorithms to pre-processed data. Specifically, it extracts data features, selects optimal parameters, and performs analysis. Specific algorithms used include random forests and neural networks. The input data is pre-processed data, and the output is the analysis result.

[0200] Step 4:

[0201] The server generates reports and recommendation lists in an easy-to-understand format for the user based on the analysis results. Specifically, it lists products that the user might be interested in and displays them in graphs and tables. The input data is the analysis results, and the output is the reports and recommendation lists provided to the user.

[0202] Step 5:

[0203] Users review the provided recommendation lists and reports and then click on or purchase products. The user's actions are recorded by their device and sent to the server as feedback. The input data is the user's actions, and the output is feedback data.

[0204] Step 6:

[0205] The server retrains its machine learning algorithm based on the feedback data. Specifically, it uses the collected feedback data to update the algorithm's parameters and improve the accuracy of the next analysis. The input data is the feedback data, and the output is the updated machine learning model.

[0206] 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.

[0207] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes the user's emotions. By having each means and the emotion engine work together, it is possible to analyze large amounts of data quickly and accurately, and to provide results that take the user's emotions into consideration.

[0208] System program processing overview

[0209] Data collection

[0210] The server receives and temporarily stores data provided by users. This data comes in various forms, such as sales data and customer support data. The emotion engine also collects emotion information from user input data and real-time interactions.

[0211] Pre-treatment

[0212] The terminal preprocesses the data received from the server. It checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data to create a consistent dataset. Furthermore, the sentiment engine analyzes the collected sentiment data and includes it in the preprocessing results.

[0213] Application of machine learning algorithms

[0214] The server applies a machine learning algorithm to the pre-processed data. The algorithm extracts data features, selects optimal parameters, and performs analysis. During this process, sentiment data provided by the sentiment engine is also considered as an element of the analysis.

[0215] Generation of analysis results

[0216] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on the emotional data provided by the emotion engine. For example, if the user is experiencing stress, a clearer and more concise report will be provided.

[0217] feedback

[0218] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these results to carry out their actual activities. The terminal records the user's activity results and emotional information and sends it to the server. This allows for an understanding of what emotional states influence activity results.

[0219] Algorithm update

[0220] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. At this time, data on user emotions is also included in the training dataset, allowing for a more accurate consideration of the influence of emotions in subsequent analyses.

[0221] Specific example

[0222] Example 1: Analysis of sales data

[0223] 1. Data Collection: The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data.

[0224] 2. Preprocessing: The terminal fills in missing values ​​in the transaction history, corrects outliers, and consolidates duplicate customer information. The sentiment engine analyzes sentiment data and incorporates it into the preprocessing.

[0225] 3. Application of machine learning algorithms: The server analyzes customer purchasing patterns and user sentiment data to generate a list of products to suggest next.

[0226] 4. Generation of analysis results: The server generates a list of emotionally sensitive product suggestions and provides them to the sales team.

[0227] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result and emotional state on the device. This feedback data is sent to the server.

[0228] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0229] Example 2: Analysis of customer support data

[0230] 1. Data Collection: The server receives inquiry data from customer support, and the sentiment engine collects the sentiment of the user being supported.

[0231] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into the preprocessing.

[0232] 3. Application of machine learning algorithms: The server analyzes the inquiry history and sentiment data to propose the optimal response method.

[0233] 4. Generation of analysis results: The server generates emotionally sensitive recommended responses and provides them to the support staff.

[0234] 5. Feedback: The support staff member implements the suggested solution and records the results and emotional state on their device. This feedback data is sent to the server.

[0235] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0236] Thus, the system of the present invention takes user emotions into consideration throughout a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, to provide more appropriate and effective results.

[0237] The following describes the processing flow.

[0238] Step 1:

[0239] The server receives sales data and customer support data from users. This data, such as transaction history and customer information, is temporarily stored on the server. The emotion engine collects text and voice data entered by users and extracts emotional information.

[0240] Step 2:

[0241] The terminal begins preprocessing the data received from the server. Specifically, it checks for missing values ​​and imputes them with the mean or median of past data as needed. Next, it detects outliers in the data and corrects them to within the range of past normal values. Furthermore, it removes duplicate data and generates a consistent dataset.

[0242] Step 3:

[0243] The emotion engine analyzes the collected emotion data. The emotion engine extracts emotional states such as positive, negative, and neutral from the user's text and voice, and adds this as emotion data to the preprocessed results.

[0244] Step 4:

[0245] The server applies machine learning algorithms to the pre-processed data and sentiment data. Here, it extracts key features from the data, selects optimal parameters, and performs analysis. By incorporating sentiment data as a feature, it enables more accurate analysis based on user emotions.

[0246] Step 5:

[0247] The server generates reports and suggestion lists for the user based on analysis results obtained by machine learning algorithms. These results are then customized to reflect the user's emotions based on sentiment data from the emotion engine. For example, if the user is feeling stressed, the results are presented in a simple and intuitive interface.

[0248] Step 6:

[0249] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then begins their actual activities based on the displayed analysis results. For example, a sales representative might propose products to a customer based on a proposal list, and a support staff member might implement the recommended course of action.

[0250] Step 7:

[0251] The device meticulously records the user's activity results, including specific outcomes such as the success or failure of product proposals by sales representatives and the success rate of inquiry resolution by support staff. Additionally, the emotion engine monitors the user's emotions in real time during their activity and updates the emotion data as needed.

[0252] Step 8:

[0253] The device sends the recorded feedback data to the server. The server analyzes the transmitted feedback data and checks for any gaps with the initial analysis results. In particular, it analyzes how emotional data influenced the activity results.

[0254] Step 9:

[0255] The server uses feedback data to retrain the machine learning algorithm. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. By incorporating sentiment data into the retraining process, the next analysis will more accurately reflect the user's emotions.

[0256] Step 10:

[0257] The server prepares to apply the updated algorithm to the next analysis. This process allows the system to continuously improve itself, progressively increasing the accuracy of the analysis.

[0258] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately and provide optimal results that take into account the user's feelings.

[0259] (Example 2)

[0260] 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".

[0261] Traditionally, data analysis systems have involved data preprocessing and the application of machine learning algorithms, but they have faced the challenge of failing to provide analysis results that take user emotions into account. In particular, analysis that ignores the impact of user emotional states on activity outcomes is insufficient in providing results that are appropriate for actual activities.

[0262] 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.

[0263] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and emotion engine means for recognizing user emotions. This makes it possible to incorporate user emotions into the analysis process and provide emotion-sensitive analysis results. Furthermore, by using user emotions and feedback data to retrain the algorithm, the accuracy of the model can be improved, and more appropriate suggestions and solutions can be provided.

[0264] "Data collection means" refers to a means of receiving and temporarily storing data provided by users or data generated within a system.

[0265] "Preprocessing means" refers to means of performing the following actions on the received data: imputing missing values, correcting outliers, removing duplicate data, and integrating sentiment data.

[0266] "Method for applying machine learning algorithms" refers to a method for applying machine learning algorithms to preprocessed data, extracting data features, and optimizing parameters while taking sentiment data into consideration.

[0267] "Analysis result generation means" refers to a means of generating reports and graphs for users based on analysis results obtained by machine learning algorithms.

[0268] A "feedback mechanism" is a means of recording the user's activity results and emotional information and sending it to a server.

[0269] An "algorithm update method" is a means of improving the accuracy of a model by retraining a machine learning algorithm based on feedback data.

[0270] An "emotion engine" is a means of collecting emotional information from user input data and real-time interactions, and incorporating it into the analysis process.

[0271] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. The following describes specific embodiments for implementing this system.

[0272] Data collection

[0273] The server receives sales data, customer support data, and other data provided by users and temporarily stores this data. The server receives data using communication methods such as REST APIs and WebSockets. The emotion engine collects emotion information from user input data and real-time interactions. This emotion data is collected using speech emotion recognition APIs and Sentiment Analysis APIs.

[0274] Pre-treatment

[0275] The terminal performs data cleaning on the data received from the server, checking for missing values ​​and imputing them as needed. Python libraries such as Pandas are used for this process. Furthermore, the terminal uses Scikit-learn's anomaly detection module and SQL queries to detect and correct outliers and remove duplicate data. The sentiment engine also analyzes the collected sentiment data and integrates it into the preprocessing results. Natural Language Toolkit (NLTK) and SpaCy are used for this task.

[0276] Application of machine learning algorithms

[0277] The server applies a machine learning algorithm to the preprocessed data. The algorithm extracts features from the data and selects optimal parameters, taking sentiment data into consideration. In this case, TensorFlow or Scikit-learn are used.

[0278] Generation of analysis results

[0279] Based on the analysis results obtained by the machine learning algorithm, the server generates reports and graphs for providing to the user. The reports and proposed contents are customized based on the sentiment data provided by the sentiment engine. For example, when the user is feeling stressed, a more understandable and concise report is provided. Specifically, graphs are created using Matplotlib or Seaborn, and reports are generated using a template engine such as Jinja2.

[0280] Feedback

[0281] The terminal displays the generated analysis results to the user. The user (salesperson or support person) conducts actual activities based on the results, records the activity results and sentiment information on the terminal, and sends them to the server. With this feedback data, it is possible to grasp in detail what kind of emotional state affects the activity results. Specifically, the terminal uses HTML / CSS to display the results and AJAX to send data to the server.

[0282] [[ID=]14Algorithm Update

[0283] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes adjustments necessary to improve the accuracy of the model. At this time, by including the user's sentiment data in the dataset, the influence of sentiment in the next analysis can be more accurately considered. TensorFlow or Keras is used for retraining.

[0284] Specific Example

[0285] Example 1: Analysis of Sales Data

[0286] The server receives customer information and transaction history from the sales team, and the sentiment engine collects the user's sentiment from the interaction data. In the preprocessing step, the terminal complements the missing values in the transaction history, corrects the outliers, and integrates the duplicate customer information. The sentiment engine analyzes the sentiment data and incorporates it into the preprocessing. By applying machine learning algorithms, the server analyzes the customer's purchase pattern and sentiment data, and then generates a list of products to be proposed. A product proposal list considering sentiment is provided to the sales team. The salesperson presents the proposed products to the customer and records the results and sentiment situation. These feedback data are sent to the server and used for the re-learning of the algorithm.

[0287] Example 2: Analysis of Customer Support Data

[0288] Inquiries data from customer support is received by the server, and the sentiment engine collects the sentiment of the users being supported. The terminal removes the noise in the inquiry content and extracts the important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into the preprocessing. The server analyzes the optimal response method from the inquiry history and sentiment data, and generates a recommended response method considering sentiment. The support staff executes the response method and records the results and sentiment situation. These feedback data are sent to the server and used for the re-learning of the algorithm.

[0289] Examples of Prompt Sentences

[0290] "Please generate product proposals based on the customer's purchase pattern from the sales data. This proposal is suitable when the user is feeling stressed."

[0291] The flow of the specific processing in Example 2 will be described using FIG. 13.

[0292] Step 1:

[0293] The server receives and temporarily stores sales data and customer support data provided by users. Specifically, the server receives data using REST APIs or WebSockets. The input in this process is data from the user, and the output is data stored on the server.

[0294] Step 2:

[0295] The server uses an emotion engine to collect user interaction data and analyze emotional information. The input is user interaction data, and the output is the analyzed emotional data. Specifically, the emotion engine uses the Sentiment Analysis API and the Speech Emotion Recognition API to collect and analyze emotional data.

[0296] Step 3:

[0297] The terminal performs data cleaning on the data received from the server, checking for and imputing missing values. The input is the data obtained from the server, and the output is the imputed data. Specifically, the terminal uses the Python library Pandas to imputate missing values.

[0298] Step 4:

[0299] The terminal detects and corrects anomalies in the data. The input is the augmented data, and the output is the corrected data. Specifically, the terminal uses the Scikit-learn anomaly detection module.

[0300] Step 5:

[0301] The terminal removes duplicate data and creates a consistent dataset. The input is corrected data, and the output is formatted data. Specifically, the terminal uses SQL queries and Pandas' duplicate detection methods.

[0302] Step 6:

[0303] The emotion engine analyzes the collected emotion data and integrates it into the pre - processing results. The input is the formatted data and the emotion data, and the output is the pre - processed data with the emotion data integrated. As a specific operation, the emotion engine uses Natural Language Toolkit (NLTK) and SpaCy to integrate the emotion data.

[0304] Step 7:

[0305] The server applies a machine learning algorithm to the pre - processed data, extracts the features of the data, and selects the optimal parameters. The input is the pre - processed data with the emotion data integrated, and the output is the analysis result. As a specific operation, the server uses TensorFlow and Scikit - learn to execute the machine learning algorithm.

[0306] Step 8:

[0307] Based on the analysis result obtained by the machine learning algorithm, the server generates reports and graphs for providing to the user. The input is the analysis result, and the output is the reports and graphs for the user. As a specific operation, the server uses Matplotlib and Seaborn to create graphs and uses a template engine such as Jinja2 to generate reports.

[0308] Step 9:

[0309] The terminal displays the generated analysis result to the user. The input is the reports and graphs for the user, and the output is the feedback from the user. As a specific operation, the terminal uses HTML / CSS to display the result and uses AJAX to send the user's feedback to the server.

[0310] Step 10:

[0311] The server retrains the machine learning algorithm based on feedback data. The input is the feedback data, and the output is the updated model. Specifically, the server uses TensorFlow or Keras to retrain the algorithm and improve the model's accuracy.

[0312] In this way, through a series of steps, the system can perform data preprocessing, analysis, result generation, user feedback, and model updates, enabling it to provide more accurate and effective results that take user sentiment into account.

[0313] (Application Example 2)

[0314] 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".

[0315] Traditional data analysis systems provided results based purely on data analysis, without considering user emotions. Therefore, if a user was experiencing a particular emotional state, the analysis results and recommendations may not be optimal, potentially significantly detracting from the user experience. Especially in e-commerce sites, product recommendations and customer support tailored to customer emotions are crucial, and a system capable of effectively achieving this was needed.

[0316] 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.

[0317] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, means including an emotion engine that recognizes user emotions, and means for customizing analysis results based on real-time emotion data. This makes it possible to provide analysis results and suggestions that take user emotions into consideration.

[0318] "Data collection means" refers to devices or mechanisms that collect and temporarily store user behavior data and real-time emotional data.

[0319] "Preprocessing means" refers to devices or mechanisms for preprocessing collected data, such as imputing missing values, correcting outliers, removing duplicate data, and preprocessing sentiment data.

[0320] A "machine learning algorithm application means" refers to a device or mechanism that applies a machine learning algorithm to preprocessed data, extracts data features, and optimizes parameters.

[0321] "Analysis result generation means" refers to a device or mechanism that generates reports and proposals for users based on analysis results obtained by machine learning algorithms.

[0322] A "feedback mechanism" is a device or mechanism that displays the generated analysis results and suggestions to the user, records the user's activity results and emotional information, and transmits it to the system.

[0323] An "algorithm update mechanism" is a device or mechanism that retrains a machine learning algorithm based on feedback data to improve the accuracy of the model.

[0324] An "emotion engine" is a device or mechanism that analyzes user input data and interactions to recognize the user's emotions.

[0325] "Means for customizing analysis results based on real-time emotional data" refers to devices or mechanisms that adjust analysis results and suggestions according to the user's emotional state, based on real-time recognized user emotional data.

[0326] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. By coordinating these means, it is possible to rapidly and accurately analyze large amounts of data and provide results that take user emotions into consideration.

[0327] System Overview

[0328] 1. Data collection:

[0329] The server collects and temporarily stores user behavior data (such as product browsing and purchase history) and real-time sentiment data. The sentiment engine uses the Emotion_recognition API to perform facial recognition and voice analysis of the user and obtain real-time sentiment data.

[0330] 2. Pre-processing:

[0331] The terminal preprocesses the data received from the server. Missing values ​​are imputed with the mean or median, and outliers are corrected using standard methods. Duplicate data is removed to create a consistent dataset. Sentiment data acquired by the sentiment engine is also incorporated into the preprocessing.

[0332] 3. Application of machine learning algorithms:

[0333] The server applies machine learning algorithms to the pre-processed data. It extracts data features using algorithms such as RandomForestClassifier, selects optimal parameters, and performs analysis. Sentiment data is also considered an important factor during this process.

[0334] 4. Generating analysis results:

[0335] The server generates product suggestions and reports based on analysis results obtained by machine learning algorithms. If a user is experiencing stress, concise and easy-to-understand suggestions are generated based on data from the emotion engine. For example, a prompt such as "Generate an easy-to-understand product description for a user experiencing stress" might be used.

[0336] 5. Feedback:

[0337] Users view and consider suggested products, providing feedback including their emotional state during the process. The device records this feedback data and sends it to the server. This data is then used for future analysis.

[0338] 6. Algorithm update:

[0339] The server retrains the machine learning algorithm based on feedback data to improve the model's accuracy. It compares the feedback results with the initial analysis results and makes necessary adjustments. Since data on user emotions is also included in the training dataset, the influence of emotions on subsequent analysis results can be more accurately understood.

[0340] In this way, analysis systems that incorporate emotion recognition are an effective means of improving the user experience. A concrete example is a system used on an e-commerce site that analyzes customer purchasing behavior and emotional data to provide optimal product recommendations.

[0341] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0342] Step 1:

[0343] Data collection

[0344] Input: User behavior data (product browsing and purchase history), sentiment data

[0345] Specific actions:

[0346] The server collects user behavior data. This behavior data includes information such as which pages the user viewed, which products they added to their cart, and which products they purchased. Furthermore, it uses an emotion engine to perform facial recognition and voice analysis of users via the Emotion_recognition API, collecting real-time emotion data.

[0347] Output: Collected behavioral and emotional data

[0348] Step 2:

[0349] Pre-treatment

[0350] Input: Collected behavioral data, emotional data

[0351] Specific actions:

[0352] The terminal performs missing value imputation, outlier correction, and duplicate data removal. Specifically, it imputates missing values ​​in behavioral data with the mean or median, and corrects outliers using standard methods. Duplicate data is removed to create a consistent dataset. Furthermore, sentiment data collected by the sentiment engine is also incorporated into the preprocessing.

[0353] Output: Consistent preprocessed data

[0354] Step 3:

[0355] Application of machine learning algorithms

[0356] Input: Preprocessed data (behavioral data, emotional data)

[0357] Specific actions:

[0358] The server applies machine learning algorithms to the preprocessed data. Specifically, it uses RandomForestClassifier to extract data features and selects optimal parameters. It also includes sentiment data in the analysis and generates an optimal model based on the characteristics of the data.

[0359] Output: Analysis results (e.g., a list of optimal product suggestions for each user)

[0360] Step 4:

[0361] Generation of analysis results

[0362] Input: Analysis results obtained from machine learning algorithms, real-time sentiment data

[0363] Specific actions:

[0364] The server generates reports and suggestions for the user based on analysis results obtained by machine learning algorithms. Based on real-time data from the emotion engine, it provides concise and easy-to-understand suggestions if the user is experiencing stress. For example, it might use a prompt such as, "Generate an easy-to-understand product description for a user who is experiencing stress."

[0365] Output: Customized proposals and reports

[0366] Step 5:

[0367] feedback

[0368] Input: Provided suggestions, user actions, sentiment data

[0369] Specific actions:

[0370] Users consider the suggested products, make purchases, or continue browsing. Their actions and emotional states are collected as feedback. The device records this feedback data and sends it to the server. This allows for the accumulation of important data for future analysis.

[0371] Output: User feedback data

[0372] Step 6:

[0373] Algorithm update

[0374] Input: Feedback data, analysis result data

[0375] Specific actions:

[0376] The server retrains its machine learning algorithm based on feedback data received from users. It compares the initial analysis results with the feedback results, makes necessary adjustments, and improves the algorithm's accuracy. By incorporating user sentiment data into the retraining process, it can provide more accurate results in subsequent analyses.

[0377] Output: Updated machine learning model

[0378] 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.

[0379] 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.

[0380] 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.

[0381] [Second Embodiment]

[0382] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0383] 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.

[0384] 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).

[0385] 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.

[0386] 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.

[0387] 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).

[0388] 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.

[0389] 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.

[0390] 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.

[0391] 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.

[0392] 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.

[0393] 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".

[0394] The invention of this system involves implementing a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, and algorithm update means. Each means works in cooperation to rapidly and accurately analyze large amounts of data, improving user convenience.

[0395] System program processing overview

[0396] Data collection

[0397] The server receives data provided by users. This data comes in various forms, such as sales data and customer support data, and is stored temporarily.

[0398] Pre-treatment

[0399] The terminal performs preprocessing on the data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0400] Application of machine learning algorithms

[0401] The server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. The results of this analysis form the basis for providing specific results tailored to the user's needs.

[0402] Generation of analysis results

[0403] The server generates reports, graphs, and other data in a user-friendly format based on analysis results obtained by machine learning algorithms. These generated analysis results are provided to the sales team and customer support team for use in their subsequent activities.

[0404] feedback

[0405] The terminal records user activity results and sends them to the server. For example, it collects information such as whether a product was purchased as a result of sales activities or whether an inquiry was resolved as a result of customer support. This information is used for subsequent analysis.

[0406] Algorithm update

[0407] The server retrains its machine learning algorithm based on feedback data, improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, gradually increasing the accuracy of its analyses.

[0408] Specific example

[0409] Example 1: Analysis of sales data

[0410] 1. Data Collection: The server receives customer information and transaction history from the sales team.

[0411] 2. Preprocessing: The terminal imputes missing values ​​in the transaction history with the average value of past data, corrects outliers, and integrates duplicate customer information.

[0412] 3. Application of machine learning algorithms: The server analyzes the customer's purchase history to identify purchasing patterns and generates a list of products to suggest next.

[0413] 4. Generation of analysis results: The server generates a list of optimal product suggestions and provides them to the sales team.

[0414] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result on the terminal. The feedback data is sent to the server.

[0415] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0416] Example 2: Analysis of customer support data

[0417] 1. Data collection: The server receives inquiry data from customer support.

[0418] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes.

[0419] 3. Application of machine learning algorithms: The server analyzes the inquiry history and proposes the optimal response method.

[0420] 4. Generation of analysis results: The server generates recommended actions and provides them to the support staff.

[0421] 5. Feedback: The user (support staff) implements the suggested solution and records the results on their device. The feedback data is sent to the server.

[0422] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0423] Thus, the system of the present invention efficiently executes a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, providing valuable information to the user.

[0424] The following describes the processing flow.

[0425] Step 1:

[0426] The server receives data from users. This data can be in various formats, such as sales data or customer support data. The server temporarily stores the received data.

[0427] Step 2:

[0428] The terminal performs preprocessing on the data received from the server. The terminal checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data and creates a consistent dataset.

[0429] Step 3:

[0430] The server applies machine learning algorithms to extract features from preprocessed data. Specifically, it analyzes data patterns and automatically selects important features. During this process, various parameters are adjusted and optimized.

[0431] Step 4:

[0432] The server generates reports and graphs to provide to users based on the analysis results. For example, it provides the sales team with a list of products to propose next, and the customer support team with the best course of action.

[0433] Step 5:

[0434] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these analysis results to carry out actual activities. For example, they might propose products to customers based on the suggestion list, or implement the most appropriate countermeasures.

[0435] Step 6:

[0436] The device records user activity. For example, it collects information such as whether a product was purchased or whether an inquiry was resolved. This information is used for subsequent analysis.

[0437] Step 7:

[0438] The terminal sends the recorded feedback data to the server. The server analyzes the feedback data and retrains the algorithm. Specifically, it compares the initial analysis results with the feedback results and makes adjustments necessary to improve the accuracy of the model.

[0439] Step 8:

[0440] The server prepares to apply the updated algorithm to the next analysis. This allows the system to continuously improve itself, gradually increasing the accuracy of the analysis.

[0441] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately, provide useful information to the user, and continuously improve the accuracy of the algorithm.

[0442] (Example 1)

[0443] 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."

[0444] Conventional data analysis systems struggle to analyze large amounts of data quickly and accurately, and are hampered by insufficient data preprocessing and feedback on analysis results. Furthermore, they lack mechanisms to effectively incorporate user feedback and improve analysis accuracy.

[0445] 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.

[0446] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by the terminal, data analysis means by the server, and user feedback recording means. This makes it possible to analyze large amounts of data quickly and accurately, and to improve the accuracy of the analysis by incorporating feedback from the user.

[0447] A "data collection method" is a means by which a server receives and temporarily stores data provided by a user.

[0448] "Preprocessing means" are methods for improving data quality by imputing missing data values, correcting outliers, and eliminating duplicate data.

[0449] A "machine learning algorithm application method" is a method of performing analysis by applying a machine learning algorithm to pre-processed data, extracting data features, and optimizing parameters.

[0450] "Analysis result generation means" refers to a means of generating reports and graphs in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0451] A "feedback mechanism" is a means of recording the user's activity results and sending that data to a server.

[0452] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0453] "Data preprocessing means by the terminal" refers to means by which the terminal preprocesses data received from the server, specifically performing missing value imputation, outlier correction, and duplicate data removal.

[0454] A "server-based data analysis method" refers to a method in which a server receives pre-processed data and performs analysis by applying machine learning algorithms.

[0455] A "user feedback recording means" is a means for a user to record the results of their activities, input that feedback data into a terminal, and send it to a server.

[0456] This system invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by a terminal, data analysis means by a server, and user feedback recording means. The program processing of this system will be described below in natural language.

[0457] First, the server receives a wide variety of data provided by the user. This data, such as sales data and customer support data, is temporarily stored. Next, the server sends the stored data to the terminal for preprocessing.

[0458] The terminal performs preprocessing on the data received from the server. Preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis. Once preprocessing is complete, the data is sent back to the server.

[0459] Next, the server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features and selects optimal parameters. The analysis results form the basis for providing specific results tailored to the user's needs.

[0460] The server generates reports and graphs in an easy-to-understand format for users based on analysis results obtained from machine learning algorithms. The generated analysis results are shared with the sales team and customer support team and used for subsequent activities. Users perform activities based on the reports provided by the server and record the results on their devices. The devices send the activity results to the server as feedback data.

[0461] The feedback data is used on the server to retrain the machine learning algorithm, improving the accuracy of subsequent analyses. This process allows the system to continuously improve itself, analyze large amounts of data quickly and accurately, and incorporate user feedback to further enhance analysis accuracy.

[0462] As a concrete example, let's consider the analysis of sales data. When a user uploads sales data to the system, the server receives the data. The server sends the received data to a terminal for preprocessing, where the terminal performs tasks such as imputing missing values, correcting outliers, and merging duplicate data. The preprocessed data is sent back to the server, which performs analysis using a machine learning algorithm. The analysis results are generated as a report and provided to the user. The user acts based on this report, records the results on their terminal, and sends them back to the server as feedback data. The server uses this feedback data to retrain its algorithm and improve the accuracy of the next analysis.

[0463] Example of a prompt:

[0464] Collect sales data, apply an algorithm to analyze purchasing patterns based on customer information and transaction history, and generate a list of products to recommend next. Output the results as a report and send it to the user via email. Retrain the algorithm using user feedback data to improve its accuracy.

[0465] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0466] System program processing flow

[0467] Step 1: Data Collection

[0468] 1. Providing user data

[0469] Users input necessary information, such as sales data and customer support data, into the system.

[0470] Input: Sales data, customer support data

[0471] Specific action: The sales representative uploads customer transaction data as a text file.

[0472] 2. Receiving data from the server

[0473] The server receives the uploaded data and stores it temporarily.

[0474] Output: Temporarily saved data file

[0475] Specific action: Receive a file named 'customers.csv' and save it to a specific folder on the server.

[0476] Step 2: Data Preprocessing

[0477] 1. Server data transmission

[0478] The server sends the stored data to the terminal for preprocessing.

[0479] Input: Temporarily saved data file

[0480] Output: Data file sent to the terminal

[0481] Specific action: Send 'customers.csv' to terminal A.

[0482] 2. Data preprocessing at the terminal

[0483] The terminal preprocesses the received data.

[0484] Input: 'customers.csv'

[0485] Output: Preprocessed data file ('data_processed.csv')

[0486] Specific actions: Imput missing values ​​with the mean of past data, correct outliers, and remove duplicate data.

[0487] Step 3: Applying the machine learning algorithm

[0488] 1. Data transmission from the device

[0489] The terminal sends the pre-processed data to the server.

[0490] Input: Preprocessed data file ('data_processed.csv')

[0491] Output: Preprocessed data file sent to the server

[0492] Specific action: Send 'data_processed.csv' to the server.

[0493] 2. Server algorithm application

[0494] The server analyzes the data using machine learning algorithms.

[0495] Input: Preprocessed data file ('data_processed.csv')

[0496] Output: Analysis results (list of proposed products, corresponding methods, etc.)

[0497] Specific operation: Analyzes purchasing patterns using 'data_processed.csv' and generates an optimal product suggestion list.

[0498] Step 4: Generating analysis results

[0499] 1. Server report generation

[0500] The server generates reports and graphs based on the analysis results.

[0501] Input: Analysis results

[0502] Output: Reports and graphs (e.g., 'proposal_list_report.pdf')

[0503] Specific operation: Output the analysis results as a PDF report, including graphed data.

[0504] 2. Server data transmission

[0505] The server sends the generated reports and graphs to the user.

[0506] Input: Reports and graphs

[0507] Output: Reports and graphs sent to the user

[0508] Specific action: Send 'proposal_list_report.pdf' to the user's email address.

[0509] Step 5: Feedback

[0510] 1. User activity log

[0511] Users perform activities based on the generated reports and record the results on their devices.

[0512] Input: Activity results based on reports and graphs

[0513] Output: Recorded activity data (e.g., 'activity_log.xlsx')

[0514] Specific operation: A sales representative proposes a product to a customer and enters the result (whether or not a purchase was made) into the terminal.

[0515] 2. Sending feedback data from the device

[0516] The device sends feedback data to the server.

[0517] Input: Recorded activity data (e.g., 'activity_log.xlsx')

[0518] Output: Feedback data sent to the server

[0519] Specific action: Send 'activity_log.xlsx' to the server.

[0520] Step 6: Update the algorithm

[0521] 1. Server retraining

[0522] The server retrains the machine learning algorithm based on the feedback data, improving the model's accuracy.

[0523] Input: Feedback data (e.g., 'activity_log.xlsx')

[0524] Output: Updated machine learning model

[0525] Specific operation: Readjust the algorithm parameters using 'activity_log.xlsx' and generate a new model.

[0526] (Application Example 1)

[0527] 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."

[0528] Traditional e-commerce sites lacked adequate systems for recommending the most suitable products based on user purchasing behavior and browsing history, making it difficult to accurately recommend products that would be beneficial to the user. As a result, user convenience decreased, and purchasing intent tended to decline. Furthermore, there was a lack of mechanisms to update algorithms and improve recommendation accuracy based on feedback.

[0529] 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.

[0530] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means for recommending the optimal product based on the user's purchasing behavior and browsing history. This makes it possible to accurately recommend the optimal product based on the user's purchasing behavior and browsing history, improving user convenience and potentially increasing purchasing intent. Furthermore, the accuracy of the algorithm can be continuously improved using feedback data.

[0531] "Data collection methods" refer to means of collecting various forms of data, such as users' purchasing behavior and browsing history.

[0532] "Preprocessing means" are methods for improving the quality of collected data by imputing missing values, correcting outliers, and eliminating duplicate data.

[0533] A "machine learning algorithm application method" is a means of applying a machine learning algorithm to pre-processed data, extracting data features, selecting optimal parameters, and performing analysis.

[0534] "Analysis result generation means" refers to a means of generating reports, recommendation lists, and other similar documents in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0535] A "feedback mechanism" is a means of improving the accuracy of analysis results by recording the user's activity results and sending those results to a server.

[0536] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0537] A "recommendation method" is a means of recommending the most suitable products based on a user's purchasing behavior and browsing history.

[0538] A system for carrying out this invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means.

[0539] The server has data collection capabilities to gather various types of data, such as user purchasing behavior and browsing history. This data is temporarily stored and used for subsequent data analysis.

[0540] The terminal has a preprocessing mechanism that performs preprocessing on data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0541] The server has a means of applying machine learning algorithms to pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. Specifically, it analyzes user interests and purchasing trends from purchase history and browsing history.

[0542] The server has a means of generating analysis results that, based on the analysis results obtained by machine learning algorithms, produce reports and recommendation lists in a user-friendly format. The generated analysis results are provided to the user for viewing and recommend products they should purchase next.

[0543] The feedback mechanism records user activity and sends it to the server. For example, it collects information such as whether the user actually purchased or viewed a recommended product. This information is used for subsequent analysis.

[0544] The algorithm update mechanism plays a role in retraining the machine learning algorithm based on feedback data, thereby improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, and the accuracy of the analysis gradually increases.

[0545] The hardware required includes a server for data collection, preprocessing, and training and application of machine learning models. Furthermore, smartphones are used to collect user purchase and browsing history and display the analysis results. The software used also includes the following:

[0546] Python: The main programming language used in programs.

[0547] pandas: A library for data collection and preprocessing.

[0548] scikit-learn: A library used to apply machine learning algorithms.

[0549] The following is a specific example. In this example, a system is designed to recommend the most suitable products based on the user's purchase and browsing history.

[0550] ---

[0551] Design a system that recommends the most suitable products based on a user's purchase and browsing history. This system will collect large amounts of data, perform preprocessing, apply machine learning algorithms, and provide valuable information to the user. As an example, show how to recommend new products based on the categories of products the user has previously purchased.

[0552] ---

[0553] By implementing such a system, it is possible to increase users' purchasing intent and improve the convenience of online shopping sites.

[0554] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0555] Step 1:

[0556] The server collects data. Specifically, the server collects and temporarily stores users' purchasing behavior and browsing history. This includes information about the user's past purchase history and the product pages they have viewed. The input data is user behavior data, and the output is the temporarily stored raw data.

[0557] Step 2:

[0558] The terminal performs preprocessing. Using the raw data received from the server, it imputes missing values, corrects outliers, and removes duplicate data. Specifically, missing values ​​are imputed with the mean or median of past data, and outliers are corrected based on predefined rules. The input data is temporarily stored raw data, and the output is preprocessed data.

[0559] Step 3:

[0560] The server applies machine learning algorithms to pre-processed data. Specifically, it extracts data features, selects optimal parameters, and performs analysis. Specific algorithms used include random forests and neural networks. The input data is pre-processed data, and the output is the analysis result.

[0561] Step 4:

[0562] The server generates reports and recommendation lists in an easy-to-understand format for the user based on the analysis results. Specifically, it lists products that the user might be interested in and displays them in graphs and tables. The input data is the analysis results, and the output is the reports and recommendation lists provided to the user.

[0563] Step 5:

[0564] Users review the provided recommendation lists and reports and then click on or purchase products. The user's actions are recorded by their device and sent to the server as feedback. The input data is the user's actions, and the output is feedback data.

[0565] Step 6:

[0566] The server retrains its machine learning algorithm based on the feedback data. Specifically, it uses the collected feedback data to update the algorithm's parameters and improve the accuracy of the next analysis. The input data is the feedback data, and the output is the updated machine learning model.

[0567] 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.

[0568] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes the user's emotions. By having each means and the emotion engine work together, it is possible to analyze large amounts of data quickly and accurately, and to provide results that take the user's emotions into consideration.

[0569] System program processing overview

[0570] Data collection

[0571] The server receives and temporarily stores data provided by users. This data comes in various forms, such as sales data and customer support data. The emotion engine also collects emotion information from user input data and real-time interactions.

[0572] Pre-treatment

[0573] The terminal preprocesses the data received from the server. It checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data to create a consistent dataset. Furthermore, the sentiment engine analyzes the collected sentiment data and includes it in the preprocessing results.

[0574] Application of machine learning algorithms

[0575] The server applies a machine learning algorithm to the pre-processed data. The algorithm extracts data features, selects optimal parameters, and performs analysis. During this process, sentiment data provided by the sentiment engine is also considered as an element of the analysis.

[0576] Generation of analysis results

[0577] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on the emotional data provided by the emotion engine. For example, if the user is experiencing stress, a clearer and more concise report will be provided.

[0578] feedback

[0579] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these results to carry out their actual activities. The terminal records the user's activity results and emotional information and sends it to the server. This allows for an understanding of what emotional states influence activity results.

[0580] Algorithm update

[0581] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. At this time, data on user emotions is also included in the training dataset, allowing for a more accurate consideration of the influence of emotions in subsequent analyses.

[0582] Specific example

[0583] Example 1: Analysis of sales data

[0584] 1. Data Collection: The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data.

[0585] 2. Preprocessing: The terminal fills in missing values ​​in the transaction history, corrects outliers, and consolidates duplicate customer information. The sentiment engine analyzes sentiment data and incorporates it into the preprocessing.

[0586] 3. Application of machine learning algorithms: The server analyzes customer purchasing patterns and user sentiment data to generate a list of products to suggest next.

[0587] 4. Generation of analysis results: The server generates a list of emotionally sensitive product suggestions and provides them to the sales team.

[0588] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result and emotional state on the device. This feedback data is sent to the server.

[0589] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0590] Example 2: Analysis of customer support data

[0591] 1. Data Collection: The server receives inquiry data from customer support, and the sentiment engine collects the sentiment of the user being supported.

[0592] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into the preprocessing.

[0593] 3. Application of machine learning algorithms: The server analyzes the inquiry history and sentiment data to propose the optimal response method.

[0594] 4. Generation of analysis results: The server generates emotionally sensitive recommended responses and provides them to the support staff.

[0595] 5. Feedback: The support staff member implements the suggested solution and records the results and emotional state on their device. This feedback data is sent to the server.

[0596] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0597] Thus, the system of the present invention takes user emotions into consideration throughout a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, to provide more appropriate and effective results.

[0598] The following describes the processing flow.

[0599] Step 1:

[0600] The server receives sales data and customer support data from users. This data, such as transaction history and customer information, is temporarily stored on the server. The emotion engine collects text and voice data entered by users and extracts emotional information.

[0601] Step 2:

[0602] The terminal begins preprocessing the data received from the server. Specifically, it checks for missing values ​​and imputes them with the mean or median of past data as needed. Next, it detects outliers in the data and corrects them to within the range of past normal values. Furthermore, it removes duplicate data and generates a consistent dataset.

[0603] Step 3:

[0604] The emotion engine analyzes the collected emotion data. The emotion engine extracts emotional states such as positive, negative, and neutral from the user's text and voice, and adds this as emotion data to the preprocessed results.

[0605] Step 4:

[0606] The server applies machine learning algorithms to the pre-processed data and sentiment data. Here, it extracts key features from the data, selects optimal parameters, and performs analysis. By incorporating sentiment data as a feature, it enables more accurate analysis based on user emotions.

[0607] Step 5:

[0608] The server generates reports and suggestion lists for the user based on analysis results obtained by machine learning algorithms. These results are then customized to reflect the user's emotions based on sentiment data from the emotion engine. For example, if the user is feeling stressed, the results are presented in a simple and intuitive interface.

[0609] Step 6:

[0610] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then begins their actual activities based on the displayed analysis results. For example, a sales representative might propose products to a customer based on a proposal list, and a support staff member might implement the recommended course of action.

[0611] Step 7:

[0612] The device meticulously records the user's activity results, including specific outcomes such as the success or failure of product proposals by sales representatives and the success rate of inquiry resolution by support staff. Additionally, the emotion engine monitors the user's emotions in real time during their activity and updates the emotion data as needed.

[0613] Step 8:

[0614] The device sends the recorded feedback data to the server. The server analyzes the transmitted feedback data and checks for any gaps with the initial analysis results. In particular, it analyzes how emotional data influenced the activity results.

[0615] Step 9:

[0616] The server uses feedback data to retrain the machine learning algorithm. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. By incorporating sentiment data into the retraining process, the next analysis will more accurately reflect the user's emotions.

[0617] Step 10:

[0618] The server prepares to apply the updated algorithm to the next analysis. This process allows the system to continuously improve itself, progressively increasing the accuracy of the analysis.

[0619] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately and provide optimal results that take into account the user's feelings.

[0620] (Example 2)

[0621] 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".

[0622] Traditionally, data analysis systems have involved data preprocessing and the application of machine learning algorithms, but they have faced the challenge of failing to provide analysis results that take user emotions into account. In particular, analysis that ignores the impact of user emotional states on activity outcomes is insufficient in providing results that are appropriate for actual activities.

[0623] 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.

[0624] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and emotion engine means for recognizing user emotions. This makes it possible to incorporate user emotions into the analysis process and provide emotion-sensitive analysis results. Furthermore, by using user emotions and feedback data to retrain the algorithm, the accuracy of the model can be improved, and more appropriate suggestions and solutions can be provided.

[0625] "Data collection means" refers to a means of receiving and temporarily storing data provided by users or data generated within a system.

[0626] "Preprocessing means" refers to means of performing the following actions on the received data: imputing missing values, correcting outliers, removing duplicate data, and integrating sentiment data.

[0627] "Method for applying machine learning algorithms" refers to a method for applying machine learning algorithms to preprocessed data, extracting data features, and optimizing parameters while taking sentiment data into consideration.

[0628] "Analysis result generation means" refers to a means of generating reports and graphs for users based on analysis results obtained by machine learning algorithms.

[0629] A "feedback mechanism" is a means of recording the user's activity results and emotional information and sending it to a server.

[0630] An "algorithm update method" is a means of improving the accuracy of a model by retraining a machine learning algorithm based on feedback data.

[0631] An "emotion engine" is a means of collecting emotional information from user input data and real-time interactions, and incorporating it into the analysis process.

[0632] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. The following describes specific embodiments for implementing this system.

[0633] Data collection

[0634] The server receives sales data, customer support data, and other data provided by users and temporarily stores this data. The server receives data using communication methods such as REST APIs and WebSockets. The emotion engine collects emotion information from user input data and real-time interactions. This emotion data is collected using speech emotion recognition APIs and Sentiment Analysis APIs.

[0635] Pre-treatment

[0636] The terminal performs data cleaning on the data received from the server, checking for missing values ​​and imputing them as needed. Python libraries such as Pandas are used for this process. Furthermore, the terminal uses Scikit-learn's anomaly detection module and SQL queries to detect and correct outliers and remove duplicate data. The sentiment engine also analyzes the collected sentiment data and integrates it into the preprocessing results. Natural Language Toolkit (NLTK) and SpaCy are used for this task.

[0637] Application of machine learning algorithms

[0638] The server applies a machine learning algorithm to the preprocessed data. The algorithm extracts features from the data and selects optimal parameters, taking sentiment data into consideration. In this case, TensorFlow or Scikit-learn are used.

[0639] Generation of analysis results

[0640] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on sentiment data provided by the sentiment engine. For example, if the user is experiencing stress, a clearer and more concise report is provided. Specifically, graphs are created using Matplotlib and Seaborn, and reports are generated using a template engine such as Jinja2.

[0641] feedback

[0642] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then performs actual activities based on these results, records the activity results and emotional information on the terminal, and sends it to the server. This feedback data allows for a detailed understanding of which emotional states influence the activity results. Specifically, the terminal displays the results using HTML / CSS and sends the data to the server using AJAX.

[0643] Algorithm update

[0644] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes necessary adjustments to improve the model's accuracy. Including user sentiment data in the dataset allows for a more accurate consideration of the impact of emotions in subsequent analyses. TensorFlow and Keras are used for retraining.

[0645] Specific example

[0646] Example 1: Analysis of sales data

[0647] The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data. In the preprocessing step, the terminal fills in missing values ​​in the transaction history, corrects outliers, and merges duplicate customer information. The emotion engine analyzes the emotional data and incorporates it into the preprocessing. By applying machine learning algorithms, the server analyzes customer purchasing patterns and emotional data to generate a list of products to recommend next. An emotionally sensitive product recommendation list is provided to the sales team. Sales representatives present the recommended products to customers and record the results and emotional state. This feedback data is sent to the server and used to retrain the algorithm.

[0648] Example 2: Analysis of customer support data

[0649] Customer support inquiry data is received by the server, and the sentiment engine collects the emotions of the user being supported. The terminal removes noise from the inquiry content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into preprocessing. The server analyzes the inquiry history and sentiment data to determine the best course of action and generates sentiment-sensitive recommended responses. The support staff implement the recommended response and record the results and sentiment status. This feedback data is sent to the server and used to retrain the algorithm.

[0650] Example of a prompt

[0651] "Generate product suggestions based on customer purchasing patterns using sales data. These suggestions are suitable for situations where the user is experiencing stress."

[0652] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0653] Step 1:

[0654] The server receives and temporarily stores sales data and customer support data provided by users. Specifically, the server receives data using REST APIs or WebSockets. The input in this process is data from the user, and the output is data stored on the server.

[0655] Step 2:

[0656] The server uses an emotion engine to collect user interaction data and analyze emotional information. The input is user interaction data, and the output is the analyzed emotional data. Specifically, the emotion engine uses the Sentiment Analysis API and the Speech Emotion Recognition API to collect and analyze emotional data.

[0657] Step 3:

[0658] The terminal performs data cleaning on the data received from the server, checking for and imputing missing values. The input is the data obtained from the server, and the output is the imputed data. Specifically, the terminal uses the Python library Pandas to imputate missing values.

[0659] Step 4:

[0660] The terminal detects and corrects anomalies in the data. The input is the augmented data, and the output is the corrected data. Specifically, the terminal uses the Scikit-learn anomaly detection module.

[0661] Step 5:

[0662] The terminal removes duplicate data and creates a consistent dataset. The input is corrected data, and the output is formatted data. Specifically, the terminal uses SQL queries and Pandas' duplicate detection methods.

[0663] Step 6:

[0664] The emotion engine analyzes the collected emotion data and integrates it into the preprocessing results. The input consists of formatted data and emotion data, and the output is preprocessed data with integrated emotion data. Specifically, the emotion engine uses Natural Language Toolkit (NLTK) or SpaCy to integrate the emotion data.

[0665] Step 7:

[0666] The server applies machine learning algorithms to preprocessed data, extracts data features, and selects optimal parameters. The input is preprocessed data with integrated sentiment data, and the output is the analysis result. Specifically, the server uses TensorFlow or Scikit-learn to execute machine learning algorithms.

[0667] Step 8:

[0668] The server generates reports and graphs for the user based on the analysis results obtained from machine learning algorithms. The input is the analysis results, and the output is the user-facing reports and graphs. Specifically, the server uses Matplotlib or Seaborn to create graphs and a template engine such as Jinja2 to generate reports.

[0669] Step 9:

[0670] The terminal displays the generated analysis results to the user. Input is user-facing reports and graphs, while output is user feedback. Specifically, the terminal displays the results using HTML / CSS and sends user feedback to the server using AJAX.

[0671] Step 10:

[0672] The server retrains the machine learning algorithm based on feedback data. The input is the feedback data, and the output is the updated model. Specifically, the server uses TensorFlow or Keras to retrain the algorithm and improve the model's accuracy.

[0673] In this way, through a series of steps, the system can perform data preprocessing, analysis, result generation, user feedback, and model updates, enabling it to provide more accurate and effective results that take user sentiment into account.

[0674] (Application Example 2)

[0675] 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."

[0676] Traditional data analysis systems provided results based purely on data analysis, without considering user emotions. Therefore, if a user was experiencing a particular emotional state, the analysis results and recommendations may not be optimal, potentially significantly detracting from the user experience. Especially in e-commerce sites, product recommendations and customer support tailored to customer emotions are crucial, and a system capable of effectively achieving this was needed.

[0677] 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.

[0678] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, means including an emotion engine that recognizes user emotions, and means for customizing analysis results based on real-time emotion data. This makes it possible to provide analysis results and suggestions that take user emotions into consideration.

[0679] "Data collection means" refers to devices or mechanisms that collect and temporarily store user behavior data and real-time emotional data.

[0680] "Preprocessing means" refers to devices or mechanisms for preprocessing collected data, such as imputing missing values, correcting outliers, removing duplicate data, and preprocessing sentiment data.

[0681] A "machine learning algorithm application means" refers to a device or mechanism that applies a machine learning algorithm to preprocessed data, extracts data features, and optimizes parameters.

[0682] "Analysis result generation means" refers to a device or mechanism that generates reports and proposals for users based on analysis results obtained by machine learning algorithms.

[0683] A "feedback mechanism" is a device or mechanism that displays the generated analysis results and suggestions to the user, records the user's activity results and emotional information, and transmits it to the system.

[0684] An "algorithm update mechanism" is a device or mechanism that retrains a machine learning algorithm based on feedback data to improve the accuracy of the model.

[0685] An "emotion engine" is a device or mechanism that analyzes user input data and interactions to recognize the user's emotions.

[0686] "Means for customizing analysis results based on real-time emotional data" refers to devices or mechanisms that adjust analysis results and suggestions according to the user's emotional state, based on real-time recognized user emotional data.

[0687] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. By coordinating these means, it is possible to rapidly and accurately analyze large amounts of data and provide results that take user emotions into consideration.

[0688] System Overview

[0689] 1. Data collection:

[0690] The server collects and temporarily stores user behavior data (such as product browsing and purchase history) and real-time sentiment data. The sentiment engine uses the Emotion_recognition API to perform facial recognition and voice analysis of the user and obtain real-time sentiment data.

[0691] 2. Pre-processing:

[0692] The terminal preprocesses the data received from the server. Missing values ​​are imputed with the mean or median, and outliers are corrected using standard methods. Duplicate data is removed to create a consistent dataset. Sentiment data acquired by the sentiment engine is also incorporated into the preprocessing.

[0693] 3. Application of machine learning algorithms:

[0694] The server applies machine learning algorithms to the pre-processed data. It extracts data features using algorithms such as RandomForestClassifier, selects optimal parameters, and performs analysis. Sentiment data is also considered an important factor during this process.

[0695] 4. Generating analysis results:

[0696] The server generates product suggestions and reports based on analysis results obtained by machine learning algorithms. If a user is experiencing stress, concise and easy-to-understand suggestions are generated based on data from the emotion engine. For example, a prompt such as "Generate an easy-to-understand product description for a user experiencing stress" might be used.

[0697] 5. Feedback:

[0698] Users view and consider suggested products, providing feedback including their emotional state during the process. The device records this feedback data and sends it to the server. This data is then used for future analysis.

[0699] 6. Algorithm update:

[0700] The server retrains the machine learning algorithm based on feedback data to improve the model's accuracy. It compares the feedback results with the initial analysis results and makes necessary adjustments. Since data on user emotions is also included in the training dataset, the influence of emotions on subsequent analysis results can be more accurately understood.

[0701] In this way, analysis systems that incorporate emotion recognition are an effective means of improving the user experience. A concrete example is a system used on an e-commerce site that analyzes customer purchasing behavior and emotional data to provide optimal product recommendations.

[0702] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0703] Step 1:

[0704] Data collection

[0705] Input: User behavior data (product browsing and purchase history), sentiment data

[0706] Specific actions:

[0707] The server collects user behavior data. This behavior data includes information such as which pages the user viewed, which products they added to their cart, and which products they purchased. Furthermore, it uses an emotion engine to perform facial recognition and voice analysis of users via the Emotion_recognition API, collecting real-time emotion data.

[0708] Output: Collected behavioral and emotional data

[0709] Step 2:

[0710] Pre-treatment

[0711] Input: Collected behavioral data, emotional data

[0712] Specific actions:

[0713] The terminal performs missing value imputation, outlier correction, and duplicate data removal. Specifically, it imputates missing values ​​in behavioral data with the mean or median, and corrects outliers using standard methods. Duplicate data is removed to create a consistent dataset. Furthermore, sentiment data collected by the sentiment engine is also incorporated into the preprocessing.

[0714] Output: Consistent preprocessed data

[0715] Step 3:

[0716] Application of machine learning algorithms

[0717] Input: Preprocessed data (behavioral data, emotional data)

[0718] Specific actions:

[0719] The server applies machine learning algorithms to the preprocessed data. Specifically, it uses RandomForestClassifier to extract data features and selects optimal parameters. It also includes sentiment data in the analysis and generates an optimal model based on the characteristics of the data.

[0720] Output: Analysis results (e.g., a list of optimal product suggestions for each user)

[0721] Step 4:

[0722] Generation of analysis results

[0723] Input: Analysis results obtained from machine learning algorithms, real-time sentiment data

[0724] Specific actions:

[0725] The server generates reports and suggestions for the user based on analysis results obtained by machine learning algorithms. Based on real-time data from the emotion engine, it provides concise and easy-to-understand suggestions if the user is experiencing stress. For example, it might use a prompt such as, "Generate an easy-to-understand product description for a user who is experiencing stress."

[0726] Output: Customized proposals and reports

[0727] Step 5:

[0728] feedback

[0729] Input: Provided suggestions, user actions, sentiment data

[0730] Specific actions:

[0731] Users consider the suggested products, make purchases, or continue browsing. Their actions and emotional states are collected as feedback. The device records this feedback data and sends it to the server. This allows for the accumulation of important data for future analysis.

[0732] Output: User feedback data

[0733] Step 6:

[0734] Algorithm update

[0735] Input: Feedback data, analysis result data

[0736] Specific actions:

[0737] The server retrains its machine learning algorithm based on feedback data received from users. It compares the initial analysis results with the feedback results, makes necessary adjustments, and improves the algorithm's accuracy. By incorporating user sentiment data into the retraining process, it can provide more accurate results in subsequent analyses.

[0738] Output: Updated machine learning model

[0739] 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.

[0740] 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.

[0741] 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.

[0742] [Third Embodiment]

[0743] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0744] 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.

[0745] 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).

[0746] 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.

[0747] 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.

[0748] 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).

[0749] 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.

[0750] 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.

[0751] 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.

[0752] 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.

[0753] 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.

[0754] 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".

[0755] The invention of this system involves implementing a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, and algorithm update means. Each means works in cooperation to rapidly and accurately analyze large amounts of data, improving user convenience.

[0756] System program processing overview

[0757] Data collection

[0758] The server receives data provided by users. This data comes in various forms, such as sales data and customer support data, and is stored temporarily.

[0759] Pre-treatment

[0760] The terminal performs preprocessing on the data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0761] Application of machine learning algorithms

[0762] The server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. The results of this analysis form the basis for providing specific results tailored to the user's needs.

[0763] Generation of analysis results

[0764] The server generates reports, graphs, and other data in a user-friendly format based on analysis results obtained by machine learning algorithms. These generated analysis results are provided to the sales team and customer support team for use in their subsequent activities.

[0765] feedback

[0766] The terminal records user activity results and sends them to the server. For example, it collects information such as whether a product was purchased as a result of sales activities or whether an inquiry was resolved as a result of customer support. This information is used for subsequent analysis.

[0767] Algorithm update

[0768] The server retrains its machine learning algorithm based on feedback data, improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, gradually increasing the accuracy of its analyses.

[0769] Specific example

[0770] Example 1: Analysis of sales data

[0771] 1. Data Collection: The server receives customer information and transaction history from the sales team.

[0772] 2. Preprocessing: The terminal imputes missing values ​​in the transaction history with the average value of past data, corrects outliers, and integrates duplicate customer information.

[0773] 3. Application of machine learning algorithms: The server analyzes the customer's purchase history to identify purchasing patterns and generates a list of products to suggest next.

[0774] 4. Generation of analysis results: The server generates a list of optimal product suggestions and provides them to the sales team.

[0775] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result on the terminal. The feedback data is sent to the server.

[0776] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0777] Example 2: Analysis of customer support data

[0778] 1. Data collection: The server receives inquiry data from customer support.

[0779] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes.

[0780] 3. Application of machine learning algorithms: The server analyzes the inquiry history and proposes the optimal response method.

[0781] 4. Generation of analysis results: The server generates recommended actions and provides them to the support staff.

[0782] 5. Feedback: The user (support staff) implements the suggested solution and records the results on their device. The feedback data is sent to the server.

[0783] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0784] Thus, the system of the present invention efficiently executes a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, providing valuable information to the user.

[0785] The following describes the processing flow.

[0786] Step 1:

[0787] The server receives data from users. This data can be in various formats, such as sales data or customer support data. The server temporarily stores the received data.

[0788] Step 2:

[0789] The terminal performs preprocessing on the data received from the server. The terminal checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data and creates a consistent dataset.

[0790] Step 3:

[0791] The server applies machine learning algorithms to extract features from preprocessed data. Specifically, it analyzes data patterns and automatically selects important features. During this process, various parameters are adjusted and optimized.

[0792] Step 4:

[0793] The server generates reports and graphs to provide to users based on the analysis results. For example, it provides the sales team with a list of products to propose next, and the customer support team with the best course of action.

[0794] Step 5:

[0795] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these analysis results to carry out actual activities. For example, they might propose products to customers based on the suggestion list, or implement the most appropriate countermeasures.

[0796] Step 6:

[0797] The device records user activity. For example, it collects information such as whether a product was purchased or whether an inquiry was resolved. This information is used for subsequent analysis.

[0798] Step 7:

[0799] The terminal sends the recorded feedback data to the server. The server analyzes the feedback data and retrains the algorithm. Specifically, it compares the initial analysis results with the feedback results and makes adjustments necessary to improve the accuracy of the model.

[0800] Step 8:

[0801] The server prepares to apply the updated algorithm to the next analysis. This allows the system to continuously improve itself, gradually increasing the accuracy of the analysis.

[0802] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately, provide useful information to the user, and continuously improve the accuracy of the algorithm.

[0803] (Example 1)

[0804] 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."

[0805] Conventional data analysis systems struggle to analyze large amounts of data quickly and accurately, and are hampered by insufficient data preprocessing and feedback on analysis results. Furthermore, they lack mechanisms to effectively incorporate user feedback and improve analysis accuracy.

[0806] 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.

[0807] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by the terminal, data analysis means by the server, and user feedback recording means. This makes it possible to analyze large amounts of data quickly and accurately, and to improve the accuracy of the analysis by incorporating feedback from the user.

[0808] A "data collection method" is a means by which a server receives and temporarily stores data provided by a user.

[0809] "Preprocessing means" are methods for improving data quality by imputing missing data values, correcting outliers, and eliminating duplicate data.

[0810] A "machine learning algorithm application method" is a method of performing analysis by applying a machine learning algorithm to pre-processed data, extracting data features, and optimizing parameters.

[0811] "Analysis result generation means" refers to a means of generating reports and graphs in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0812] A "feedback mechanism" is a means of recording the user's activity results and sending that data to a server.

[0813] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0814] "Data preprocessing means by the terminal" refers to means by which the terminal preprocesses data received from the server, specifically performing missing value imputation, outlier correction, and duplicate data removal.

[0815] A "server-based data analysis method" refers to a method in which a server receives pre-processed data and performs analysis by applying machine learning algorithms.

[0816] A "user feedback recording means" is a means for a user to record the results of their activities, input that feedback data into a terminal, and send it to a server.

[0817] This system invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by a terminal, data analysis means by a server, and user feedback recording means. The program processing of this system will be described below in natural language.

[0818] First, the server receives a wide variety of data provided by the user. This data, such as sales data and customer support data, is temporarily stored. Next, the server sends the stored data to the terminal for preprocessing.

[0819] The terminal performs preprocessing on the data received from the server. Preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis. Once preprocessing is complete, the data is sent back to the server.

[0820] Next, the server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features and selects optimal parameters. The analysis results form the basis for providing specific results tailored to the user's needs.

[0821] The server generates reports and graphs in an easy-to-understand format for users based on analysis results obtained from machine learning algorithms. The generated analysis results are shared with the sales team and customer support team and used for subsequent activities. Users perform activities based on the reports provided by the server and record the results on their devices. The devices send the activity results to the server as feedback data.

[0822] The feedback data is used on the server to retrain the machine learning algorithm, improving the accuracy of subsequent analyses. This process allows the system to continuously improve itself, analyze large amounts of data quickly and accurately, and incorporate user feedback to further enhance analysis accuracy.

[0823] As a concrete example, let's consider the analysis of sales data. When a user uploads sales data to the system, the server receives the data. The server sends the received data to a terminal for preprocessing, where the terminal performs tasks such as imputing missing values, correcting outliers, and merging duplicate data. The preprocessed data is sent back to the server, which performs analysis using a machine learning algorithm. The analysis results are generated as a report and provided to the user. The user acts based on this report, records the results on their terminal, and sends them back to the server as feedback data. The server uses this feedback data to retrain its algorithm and improve the accuracy of the next analysis.

[0824] Example of a prompt:

[0825] Collect sales data, apply an algorithm to analyze purchasing patterns based on customer information and transaction history, and generate a list of products to recommend next. Output the results as a report and send it to the user via email. Retrain the algorithm using user feedback data to improve its accuracy.

[0826] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0827] System program processing flow

[0828] Step 1: Data Collection

[0829] 1. Providing user data

[0830] Users input necessary information, such as sales data and customer support data, into the system.

[0831] Input: Sales data, customer support data

[0832] Specific action: The sales representative uploads customer transaction data as a text file.

[0833] 2. Receiving data from the server

[0834] The server receives the uploaded data and stores it temporarily.

[0835] Output: Temporarily saved data file

[0836] Specific action: Receive a file named 'customers.csv' and save it to a specific folder on the server.

[0837] Step 2: Data Preprocessing

[0838] 1. Server data transmission

[0839] The server sends the stored data to the terminal for preprocessing.

[0840] Input: Temporarily saved data file

[0841] Output: Data file sent to the terminal

[0842] Specific action: Send 'customers.csv' to terminal A.

[0843] 2. Data preprocessing at the terminal

[0844] The terminal preprocesses the received data.

[0845] Input: 'customers.csv'

[0846] Output: Preprocessed data file ('data_processed.csv')

[0847] Specific actions: Imput missing values ​​with the mean of past data, correct outliers, and remove duplicate data.

[0848] Step 3: Applying the machine learning algorithm

[0849] 1. Data transmission from the device

[0850] The terminal sends the pre-processed data to the server.

[0851] Input: Preprocessed data file ('data_processed.csv')

[0852] Output: Preprocessed data file sent to the server

[0853] Specific action: Send 'data_processed.csv' to the server.

[0854] 2. Server algorithm application

[0855] The server analyzes the data using machine learning algorithms.

[0856] Input: Preprocessed data file ('data_processed.csv')

[0857] Output: Analysis results (list of proposed products, corresponding methods, etc.)

[0858] Specific operation: Analyzes purchasing patterns using 'data_processed.csv' and generates an optimal product suggestion list.

[0859] Step 4: Generating analysis results

[0860] 1. Server report generation

[0861] The server generates reports and graphs based on the analysis results.

[0862] Input: Analysis results

[0863] Output: Reports and graphs (e.g., 'proposal_list_report.pdf')

[0864] Specific operation: Output the analysis results as a PDF report, including graphed data.

[0865] 2. Server data transmission

[0866] The server sends the generated reports and graphs to the user.

[0867] Input: Reports and graphs

[0868] Output: Reports and graphs sent to the user

[0869] Specific action: Send 'proposal_list_report.pdf' to the user's email address.

[0870] Step 5: Feedback

[0871] 1. User activity log

[0872] Users perform activities based on the generated reports and record the results on their devices.

[0873] Input: Activity results based on reports and graphs

[0874] Output: Recorded activity data (e.g., 'activity_log.xlsx')

[0875] Specific operation: A sales representative proposes a product to a customer and enters the result (whether or not a purchase was made) into the terminal.

[0876] 2. Sending feedback data from the device

[0877] The device sends feedback data to the server.

[0878] Input: Recorded activity data (e.g., 'activity_log.xlsx')

[0879] Output: Feedback data sent to the server

[0880] Specific action: Send 'activity_log.xlsx' to the server.

[0881] Step 6: Update the algorithm

[0882] 1. Server retraining

[0883] The server retrains the machine learning algorithm based on the feedback data, improving the model's accuracy.

[0884] Input: Feedback data (e.g., 'activity_log.xlsx')

[0885] Output: Updated machine learning model

[0886] Specific operation: Readjust the algorithm parameters using 'activity_log.xlsx' and generate a new model.

[0887] (Application Example 1)

[0888] 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."

[0889] Traditional e-commerce sites lacked adequate systems for recommending the most suitable products based on user purchasing behavior and browsing history, making it difficult to accurately recommend products that would be beneficial to the user. As a result, user convenience decreased, and purchasing intent tended to decline. Furthermore, there was a lack of mechanisms to update algorithms and improve recommendation accuracy based on feedback.

[0890] 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.

[0891] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means for recommending the optimal product based on the user's purchasing behavior and browsing history. This makes it possible to accurately recommend the optimal product based on the user's purchasing behavior and browsing history, improving user convenience and potentially increasing purchasing intent. Furthermore, the accuracy of the algorithm can be continuously improved using feedback data.

[0892] "Data collection methods" refer to means of collecting various forms of data, such as users' purchasing behavior and browsing history.

[0893] "Preprocessing means" are methods for improving the quality of collected data by imputing missing values, correcting outliers, and eliminating duplicate data.

[0894] A "machine learning algorithm application method" is a means of applying a machine learning algorithm to pre-processed data, extracting data features, selecting optimal parameters, and performing analysis.

[0895] "Analysis result generation means" refers to a means of generating reports, recommendation lists, and other similar documents in a user-friendly format based on analysis results obtained by machine learning algorithms.

[0896] A "feedback mechanism" is a means of improving the accuracy of analysis results by recording the user's activity results and sending those results to a server.

[0897] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[0898] A "recommendation method" is a means of recommending the most suitable products based on a user's purchasing behavior and browsing history.

[0899] A system for carrying out this invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means.

[0900] The server has data collection capabilities to gather various types of data, such as user purchasing behavior and browsing history. This data is temporarily stored and used for subsequent data analysis.

[0901] The terminal has a preprocessing mechanism that performs preprocessing on data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[0902] The server has a means of applying machine learning algorithms to pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. Specifically, it analyzes user interests and purchasing trends from purchase history and browsing history.

[0903] The server has a means of generating analysis results that, based on the analysis results obtained by machine learning algorithms, produce reports and recommendation lists in a user-friendly format. The generated analysis results are provided to the user for viewing and recommend products they should purchase next.

[0904] The feedback mechanism records user activity and sends it to the server. For example, it collects information such as whether the user actually purchased or viewed a recommended product. This information is used for subsequent analysis.

[0905] The algorithm update mechanism plays a role in retraining the machine learning algorithm based on feedback data, thereby improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, and the accuracy of the analysis gradually increases.

[0906] The hardware required includes a server for data collection, preprocessing, and training and application of machine learning models. Furthermore, smartphones are used to collect user purchase and browsing history and display the analysis results. The software used also includes the following:

[0907] Python: The main programming language used in programs.

[0908] pandas: A library for data collection and preprocessing.

[0909] scikit-learn: A library used to apply machine learning algorithms.

[0910] The following is a specific example. In this example, a system is designed to recommend the most suitable products based on the user's purchase and browsing history.

[0911] ---

[0912] Design a system that recommends the most suitable products based on a user's purchase and browsing history. This system will collect large amounts of data, perform preprocessing, apply machine learning algorithms, and provide valuable information to the user. As an example, show how to recommend new products based on the categories of products the user has previously purchased.

[0913] ---

[0914] By implementing such a system, it is possible to increase users' purchasing intent and improve the convenience of online shopping sites.

[0915] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0916] Step 1:

[0917] The server collects data. Specifically, the server collects and temporarily stores users' purchasing behavior and browsing history. This includes information about the user's past purchase history and the product pages they have viewed. The input data is user behavior data, and the output is the temporarily stored raw data.

[0918] Step 2:

[0919] The terminal performs preprocessing. Using the raw data received from the server, it imputes missing values, corrects outliers, and removes duplicate data. Specifically, missing values ​​are imputed with the mean or median of past data, and outliers are corrected based on predefined rules. The input data is temporarily stored raw data, and the output is preprocessed data.

[0920] Step 3:

[0921] The server applies machine learning algorithms to pre-processed data. Specifically, it extracts data features, selects optimal parameters, and performs analysis. Specific algorithms used include random forests and neural networks. The input data is pre-processed data, and the output is the analysis result.

[0922] Step 4:

[0923] The server generates reports and recommendation lists in an easy-to-understand format for the user based on the analysis results. Specifically, it lists products that the user might be interested in and displays them in graphs and tables. The input data is the analysis results, and the output is the reports and recommendation lists provided to the user.

[0924] Step 5:

[0925] Users review the provided recommendation lists and reports and then click on or purchase products. The user's actions are recorded by their device and sent to the server as feedback. The input data is the user's actions, and the output is feedback data.

[0926] Step 6:

[0927] The server retrains its machine learning algorithm based on the feedback data. Specifically, it uses the collected feedback data to update the algorithm's parameters and improve the accuracy of the next analysis. The input data is the feedback data, and the output is the updated machine learning model.

[0928] 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.

[0929] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes the user's emotions. By having each means and the emotion engine work together, it is possible to analyze large amounts of data quickly and accurately, and to provide results that take the user's emotions into consideration.

[0930] System program processing overview

[0931] Data collection

[0932] The server receives and temporarily stores data provided by users. This data comes in various forms, such as sales data and customer support data. The emotion engine also collects emotion information from user input data and real-time interactions.

[0933] Pre-treatment

[0934] The terminal preprocesses the data received from the server. It checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data to create a consistent dataset. Furthermore, the sentiment engine analyzes the collected sentiment data and includes it in the preprocessing results.

[0935] Application of machine learning algorithms

[0936] The server applies a machine learning algorithm to the pre-processed data. The algorithm extracts data features, selects optimal parameters, and performs analysis. During this process, sentiment data provided by the sentiment engine is also considered as an element of the analysis.

[0937] Generation of analysis results

[0938] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on the emotional data provided by the emotion engine. For example, if the user is experiencing stress, a clearer and more concise report will be provided.

[0939] feedback

[0940] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these results to carry out their actual activities. The terminal records the user's activity results and emotional information and sends it to the server. This allows for an understanding of what emotional states influence activity results.

[0941] Algorithm update

[0942] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. At this time, data on user emotions is also included in the training dataset, allowing for a more accurate consideration of the influence of emotions in subsequent analyses.

[0943] Specific example

[0944] Example 1: Analysis of sales data

[0945] 1. Data Collection: The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data.

[0946] 2. Preprocessing: The terminal fills in missing values ​​in the transaction history, corrects outliers, and consolidates duplicate customer information. The sentiment engine analyzes sentiment data and incorporates it into the preprocessing.

[0947] 3. Application of machine learning algorithms: The server analyzes customer purchasing patterns and user sentiment data to generate a list of products to suggest next.

[0948] 4. Generation of analysis results: The server generates a list of emotionally sensitive product suggestions and provides them to the sales team.

[0949] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result and emotional state on the device. This feedback data is sent to the server.

[0950] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[0951] Example 2: Analysis of customer support data

[0952] 1. Data Collection: The server receives inquiry data from customer support, and the sentiment engine collects the sentiment of the user being supported.

[0953] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into the preprocessing.

[0954] 3. Application of machine learning algorithms: The server analyzes the inquiry history and sentiment data to propose the optimal response method.

[0955] 4. Generation of analysis results: The server generates emotionally sensitive recommended responses and provides them to the support staff.

[0956] 5. Feedback: The support staff member implements the suggested solution and records the results and emotional state on their device. This feedback data is sent to the server.

[0957] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[0958] Thus, the system of the present invention takes user emotions into consideration throughout a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, to provide more appropriate and effective results.

[0959] The following describes the processing flow.

[0960] Step 1:

[0961] The server receives sales data and customer support data from users. This data, such as transaction history and customer information, is temporarily stored on the server. The emotion engine collects text and voice data entered by users and extracts emotional information.

[0962] Step 2:

[0963] The terminal begins preprocessing the data received from the server. Specifically, it checks for missing values ​​and imputes them with the mean or median of past data as needed. Next, it detects outliers in the data and corrects them to within the range of past normal values. Furthermore, it removes duplicate data and generates a consistent dataset.

[0964] Step 3:

[0965] The emotion engine analyzes the collected emotion data. The emotion engine extracts emotional states such as positive, negative, and neutral from the user's text and voice, and adds this as emotion data to the preprocessed results.

[0966] Step 4:

[0967] The server applies machine learning algorithms to the pre-processed data and sentiment data. Here, it extracts key features from the data, selects optimal parameters, and performs analysis. By incorporating sentiment data as a feature, it enables more accurate analysis based on user emotions.

[0968] Step 5:

[0969] The server generates reports and suggestion lists for the user based on analysis results obtained by machine learning algorithms. These results are then customized to reflect the user's emotions based on sentiment data from the emotion engine. For example, if the user is feeling stressed, the results are presented in a simple and intuitive interface.

[0970] Step 6:

[0971] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then begins their actual activities based on the displayed analysis results. For example, a sales representative might propose products to a customer based on a proposal list, and a support staff member might implement the recommended course of action.

[0972] Step 7:

[0973] The device meticulously records the user's activity results, including specific outcomes such as the success or failure of product proposals by sales representatives and the success rate of inquiry resolution by support staff. Additionally, the emotion engine monitors the user's emotions in real time during their activity and updates the emotion data as needed.

[0974] Step 8:

[0975] The device sends the recorded feedback data to the server. The server analyzes the transmitted feedback data and checks for any gaps with the initial analysis results. In particular, it analyzes how emotional data influenced the activity results.

[0976] Step 9:

[0977] The server uses feedback data to retrain the machine learning algorithm. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. By incorporating sentiment data into the retraining process, the next analysis will more accurately reflect the user's emotions.

[0978] Step 10:

[0979] The server prepares to apply the updated algorithm to the next analysis. This process allows the system to continuously improve itself, progressively increasing the accuracy of the analysis.

[0980] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately and provide optimal results that take into account the user's feelings.

[0981] (Example 2)

[0982] 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."

[0983] Traditionally, data analysis systems have involved data preprocessing and the application of machine learning algorithms, but they have faced the challenge of failing to provide analysis results that take user emotions into account. In particular, analysis that ignores the impact of user emotional states on activity outcomes is insufficient in providing results that are appropriate for actual activities.

[0984] 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.

[0985] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and emotion engine means for recognizing user emotions. This makes it possible to incorporate user emotions into the analysis process and provide emotion-sensitive analysis results. Furthermore, by using user emotions and feedback data to retrain the algorithm, the accuracy of the model can be improved, and more appropriate suggestions and solutions can be provided.

[0986] "Data collection means" refers to a means of receiving and temporarily storing data provided by users or data generated within a system.

[0987] "Preprocessing means" refers to means of performing the following actions on the received data: imputing missing values, correcting outliers, removing duplicate data, and integrating sentiment data.

[0988] "Method for applying machine learning algorithms" refers to a method for applying machine learning algorithms to preprocessed data, extracting data features, and optimizing parameters while taking sentiment data into consideration.

[0989] "Analysis result generation means" refers to a means of generating reports and graphs for users based on analysis results obtained by machine learning algorithms.

[0990] A "feedback mechanism" is a means of recording the user's activity results and emotional information and sending it to a server.

[0991] An "algorithm update method" is a means of improving the accuracy of a model by retraining a machine learning algorithm based on feedback data.

[0992] An "emotion engine" is a means of collecting emotional information from user input data and real-time interactions, and incorporating it into the analysis process.

[0993] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. The following describes specific embodiments for implementing this system.

[0994] Data collection

[0995] The server receives sales data, customer support data, and other data provided by users and temporarily stores this data. The server receives data using communication methods such as REST APIs and WebSockets. The emotion engine collects emotion information from user input data and real-time interactions. This emotion data is collected using speech emotion recognition APIs and Sentiment Analysis APIs.

[0996] Pre-treatment

[0997] The terminal performs data cleaning on the data received from the server, checking for missing values ​​and imputing them as needed. Python libraries such as Pandas are used for this process. Furthermore, the terminal uses Scikit-learn's anomaly detection module and SQL queries to detect and correct outliers and remove duplicate data. The sentiment engine also analyzes the collected sentiment data and integrates it into the preprocessing results. Natural Language Toolkit (NLTK) and SpaCy are used for this task.

[0998] Application of machine learning algorithms

[0999] The server applies a machine learning algorithm to the preprocessed data. The algorithm extracts features from the data and selects optimal parameters, taking sentiment data into consideration. In this case, TensorFlow or Scikit-learn are used.

[1000] Generation of analysis results

[1001] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on sentiment data provided by the sentiment engine. For example, if the user is experiencing stress, a clearer and more concise report is provided. Specifically, graphs are created using Matplotlib and Seaborn, and reports are generated using a template engine such as Jinja2.

[1002] feedback

[1003] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then performs actual activities based on these results, records the activity results and emotional information on the terminal, and sends it to the server. This feedback data allows for a detailed understanding of which emotional states influence the activity results. Specifically, the terminal displays the results using HTML / CSS and sends the data to the server using AJAX.

[1004] Algorithm update

[1005] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes necessary adjustments to improve the model's accuracy. Including user sentiment data in the dataset allows for a more accurate consideration of the impact of emotions in subsequent analyses. TensorFlow and Keras are used for retraining.

[1006] Specific example

[1007] Example 1: Analysis of sales data

[1008] The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data. In the preprocessing step, the terminal fills in missing values ​​in the transaction history, corrects outliers, and merges duplicate customer information. The emotion engine analyzes the emotional data and incorporates it into the preprocessing. By applying machine learning algorithms, the server analyzes customer purchasing patterns and emotional data to generate a list of products to recommend next. An emotionally sensitive product recommendation list is provided to the sales team. Sales representatives present the recommended products to customers and record the results and emotional state. This feedback data is sent to the server and used to retrain the algorithm.

[1009] Example 2: Analysis of customer support data

[1010] Customer support inquiry data is received by the server, and the sentiment engine collects the emotions of the user being supported. The terminal removes noise from the inquiry content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into preprocessing. The server analyzes the inquiry history and sentiment data to determine the best course of action and generates sentiment-sensitive recommended responses. The support staff implement the recommended response and record the results and sentiment status. This feedback data is sent to the server and used to retrain the algorithm.

[1011] Example of a prompt

[1012] "Generate product suggestions based on customer purchasing patterns using sales data. These suggestions are suitable for situations where the user is experiencing stress."

[1013] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1014] Step 1:

[1015] The server receives and temporarily stores sales data and customer support data provided by users. Specifically, the server receives data using REST APIs or WebSockets. The input in this process is data from the user, and the output is data stored on the server.

[1016] Step 2:

[1017] The server uses an emotion engine to collect user interaction data and analyze emotional information. The input is user interaction data, and the output is the analyzed emotional data. Specifically, the emotion engine uses the Sentiment Analysis API and the Speech Emotion Recognition API to collect and analyze emotional data.

[1018] Step 3:

[1019] The terminal performs data cleaning on the data received from the server, checking for and imputing missing values. The input is the data obtained from the server, and the output is the imputed data. Specifically, the terminal uses the Python library Pandas to imputate missing values.

[1020] Step 4:

[1021] The terminal detects and corrects anomalies in the data. The input is the augmented data, and the output is the corrected data. Specifically, the terminal uses the Scikit-learn anomaly detection module.

[1022] Step 5:

[1023] The terminal removes duplicate data and creates a consistent dataset. The input is corrected data, and the output is formatted data. Specifically, the terminal uses SQL queries and Pandas' duplicate detection methods.

[1024] Step 6:

[1025] The emotion engine analyzes the collected emotion data and integrates it into the preprocessing results. The input consists of formatted data and emotion data, and the output is preprocessed data with integrated emotion data. Specifically, the emotion engine uses Natural Language Toolkit (NLTK) or SpaCy to integrate the emotion data.

[1026] Step 7:

[1027] The server applies machine learning algorithms to preprocessed data, extracts data features, and selects optimal parameters. The input is preprocessed data with integrated sentiment data, and the output is the analysis result. Specifically, the server uses TensorFlow or Scikit-learn to execute machine learning algorithms.

[1028] Step 8:

[1029] The server generates reports and graphs for the user based on the analysis results obtained from machine learning algorithms. The input is the analysis results, and the output is the user-facing reports and graphs. Specifically, the server uses Matplotlib or Seaborn to create graphs and a template engine such as Jinja2 to generate reports.

[1030] Step 9:

[1031] The terminal displays the generated analysis results to the user. Input is user-facing reports and graphs, while output is user feedback. Specifically, the terminal displays the results using HTML / CSS and sends user feedback to the server using AJAX.

[1032] Step 10:

[1033] The server retrains the machine learning algorithm based on feedback data. The input is the feedback data, and the output is the updated model. Specifically, the server uses TensorFlow or Keras to retrain the algorithm and improve the model's accuracy.

[1034] In this way, through a series of steps, the system can perform data preprocessing, analysis, result generation, user feedback, and model updates, enabling it to provide more accurate and effective results that take user sentiment into account.

[1035] (Application Example 2)

[1036] 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."

[1037] Traditional data analysis systems provided results based purely on data analysis, without considering user emotions. Therefore, if a user was experiencing a particular emotional state, the analysis results and recommendations may not be optimal, potentially significantly detracting from the user experience. Especially in e-commerce sites, product recommendations and customer support tailored to customer emotions are crucial, and a system capable of effectively achieving this was needed.

[1038] 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.

[1039] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, means including an emotion engine that recognizes user emotions, and means for customizing analysis results based on real-time emotion data. This makes it possible to provide analysis results and suggestions that take user emotions into consideration.

[1040] "Data collection means" refers to devices or mechanisms that collect and temporarily store user behavior data and real-time emotional data.

[1041] "Preprocessing means" refers to devices or mechanisms for preprocessing collected data, such as imputing missing values, correcting outliers, removing duplicate data, and preprocessing sentiment data.

[1042] A "machine learning algorithm application means" refers to a device or mechanism that applies a machine learning algorithm to preprocessed data, extracts data features, and optimizes parameters.

[1043] "Analysis result generation means" refers to a device or mechanism that generates reports and proposals for users based on analysis results obtained by machine learning algorithms.

[1044] A "feedback mechanism" is a device or mechanism that displays the generated analysis results and suggestions to the user, records the user's activity results and emotional information, and transmits it to the system.

[1045] An "algorithm update mechanism" is a device or mechanism that retrains a machine learning algorithm based on feedback data to improve the accuracy of the model.

[1046] An "emotion engine" is a device or mechanism that analyzes user input data and interactions to recognize the user's emotions.

[1047] "Means for customizing analysis results based on real-time emotional data" refers to devices or mechanisms that adjust analysis results and suggestions according to the user's emotional state, based on real-time recognized user emotional data.

[1048] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. By coordinating these means, it is possible to rapidly and accurately analyze large amounts of data and provide results that take user emotions into consideration.

[1049] System Overview

[1050] 1. Data collection:

[1051] The server collects and temporarily stores user behavior data (such as product browsing and purchase history) and real-time sentiment data. The sentiment engine uses the Emotion_recognition API to perform facial recognition and voice analysis of the user and obtain real-time sentiment data.

[1052] 2. Pre-processing:

[1053] The terminal preprocesses the data received from the server. Missing values ​​are imputed with the mean or median, and outliers are corrected using standard methods. Duplicate data is removed to create a consistent dataset. Sentiment data acquired by the sentiment engine is also incorporated into the preprocessing.

[1054] 3. Application of machine learning algorithms:

[1055] The server applies machine learning algorithms to the pre-processed data. It extracts data features using algorithms such as RandomForestClassifier, selects optimal parameters, and performs analysis. Sentiment data is also considered an important factor during this process.

[1056] 4. Generating analysis results:

[1057] The server generates product suggestions and reports based on analysis results obtained by machine learning algorithms. If a user is experiencing stress, concise and easy-to-understand suggestions are generated based on data from the emotion engine. For example, a prompt such as "Generate an easy-to-understand product description for a user experiencing stress" might be used.

[1058] 5. Feedback:

[1059] Users view and consider suggested products, providing feedback including their emotional state during the process. The device records this feedback data and sends it to the server. This data is then used for future analysis.

[1060] 6. Algorithm update:

[1061] The server retrains the machine learning algorithm based on feedback data to improve the model's accuracy. It compares the feedback results with the initial analysis results and makes necessary adjustments. Since data on user emotions is also included in the training dataset, the influence of emotions on subsequent analysis results can be more accurately understood.

[1062] In this way, analysis systems that incorporate emotion recognition are an effective means of improving the user experience. A concrete example is a system used on an e-commerce site that analyzes customer purchasing behavior and emotional data to provide optimal product recommendations.

[1063] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1064] Step 1:

[1065] Data collection

[1066] Input: User behavior data (product browsing and purchase history), sentiment data

[1067] Specific actions:

[1068] The server collects user behavior data. This behavior data includes information such as which pages the user viewed, which products they added to their cart, and which products they purchased. Furthermore, it uses an emotion engine to perform facial recognition and voice analysis of users via the Emotion_recognition API, collecting real-time emotion data.

[1069] Output: Collected behavioral and emotional data

[1070] Step 2:

[1071] Pre-treatment

[1072] Input: Collected behavioral data, emotional data

[1073] Specific actions:

[1074] The terminal performs missing value imputation, outlier correction, and duplicate data removal. Specifically, it imputates missing values ​​in behavioral data with the mean or median, and corrects outliers using standard methods. Duplicate data is removed to create a consistent dataset. Furthermore, sentiment data collected by the sentiment engine is also incorporated into the preprocessing.

[1075] Output: Consistent preprocessed data

[1076] Step 3:

[1077] Application of machine learning algorithms

[1078] Input: Preprocessed data (behavioral data, emotional data)

[1079] Specific actions:

[1080] The server applies machine learning algorithms to the preprocessed data. Specifically, it uses RandomForestClassifier to extract data features and selects optimal parameters. It also includes sentiment data in the analysis and generates an optimal model based on the characteristics of the data.

[1081] Output: Analysis results (e.g., a list of optimal product suggestions for each user)

[1082] Step 4:

[1083] Generation of analysis results

[1084] Input: Analysis results obtained from machine learning algorithms, real-time sentiment data

[1085] Specific actions:

[1086] The server generates reports and suggestions for the user based on analysis results obtained by machine learning algorithms. Based on real-time data from the emotion engine, it provides concise and easy-to-understand suggestions if the user is experiencing stress. For example, it might use a prompt such as, "Generate an easy-to-understand product description for a user who is experiencing stress."

[1087] Output: Customized proposals and reports

[1088] Step 5:

[1089] feedback

[1090] Input: Provided suggestions, user actions, sentiment data

[1091] Specific actions:

[1092] Users consider the suggested products, make purchases, or continue browsing. Their actions and emotional states are collected as feedback. The device records this feedback data and sends it to the server. This allows for the accumulation of important data for future analysis.

[1093] Output: User feedback data

[1094] Step 6:

[1095] Algorithm update

[1096] Input: Feedback data, analysis result data

[1097] Specific actions:

[1098] The server retrains its machine learning algorithm based on feedback data received from users. It compares the initial analysis results with the feedback results, makes necessary adjustments, and improves the algorithm's accuracy. By incorporating user sentiment data into the retraining process, it can provide more accurate results in subsequent analyses.

[1099] Output: Updated machine learning model

[1100] 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.

[1101] 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.

[1102] 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.

[1103] [Fourth Embodiment]

[1104] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1105] 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.

[1106] 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).

[1107] 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.

[1108] 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.

[1109] 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).

[1110] 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.

[1111] 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.

[1112] 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.

[1113] 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.

[1114] 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.

[1115] 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.

[1116] 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".

[1117] The invention of this system involves implementing a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, and algorithm update means. Each means works in cooperation to rapidly and accurately analyze large amounts of data, improving user convenience.

[1118] System program processing overview

[1119] Data collection

[1120] The server receives data provided by users. This data comes in various forms, such as sales data and customer support data, and is stored temporarily.

[1121] Pre-treatment

[1122] The terminal performs preprocessing on the data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[1123] Application of machine learning algorithms

[1124] The server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. The results of this analysis form the basis for providing specific results tailored to the user's needs.

[1125] Generation of analysis results

[1126] The server generates reports, graphs, and other data in a user-friendly format based on analysis results obtained by machine learning algorithms. These generated analysis results are provided to the sales team and customer support team for use in their subsequent activities.

[1127] feedback

[1128] The terminal records user activity results and sends them to the server. For example, it collects information such as whether a product was purchased as a result of sales activities or whether an inquiry was resolved as a result of customer support. This information is used for subsequent analysis.

[1129] Algorithm update

[1130] The server retrains its machine learning algorithm based on feedback data, improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, gradually increasing the accuracy of its analyses.

[1131] Specific example

[1132] Example 1: Analysis of sales data

[1133] 1. Data Collection: The server receives customer information and transaction history from the sales team.

[1134] 2. Preprocessing: The terminal imputes missing values ​​in the transaction history with the average value of past data, corrects outliers, and integrates duplicate customer information.

[1135] 3. Application of machine learning algorithms: The server analyzes the customer's purchase history to identify purchasing patterns and generates a list of products to suggest next.

[1136] 4. Generation of analysis results: The server generates a list of optimal product suggestions and provides them to the sales team.

[1137] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result on the terminal. The feedback data is sent to the server.

[1138] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[1139] Example 2: Analysis of customer support data

[1140] 1. Data collection: The server receives inquiry data from customer support.

[1141] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes.

[1142] 3. Application of machine learning algorithms: The server analyzes the inquiry history and proposes the optimal response method.

[1143] 4. Generation of analysis results: The server generates recommended actions and provides them to the support staff.

[1144] 5. Feedback: The user (support staff) implements the suggested solution and records the results on their device. The feedback data is sent to the server.

[1145] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[1146] Thus, the system of the present invention efficiently executes a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, providing valuable information to the user.

[1147] The following describes the processing flow.

[1148] Step 1:

[1149] The server receives data from users. This data can be in various formats, such as sales data or customer support data. The server temporarily stores the received data.

[1150] Step 2:

[1151] The terminal performs preprocessing on the data received from the server. The terminal checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data and creates a consistent dataset.

[1152] Step 3:

[1153] The server applies machine learning algorithms to extract features from preprocessed data. Specifically, it analyzes data patterns and automatically selects important features. During this process, various parameters are adjusted and optimized.

[1154] Step 4:

[1155] The server generates reports and graphs to provide to users based on the analysis results. For example, it provides the sales team with a list of products to propose next, and the customer support team with the best course of action.

[1156] Step 5:

[1157] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these analysis results to carry out actual activities. For example, they might propose products to customers based on the suggestion list, or implement the most appropriate countermeasures.

[1158] Step 6:

[1159] The device records user activity. For example, it collects information such as whether a product was purchased or whether an inquiry was resolved. This information is used for subsequent analysis.

[1160] Step 7:

[1161] The terminal sends the recorded feedback data to the server. The server analyzes the feedback data and retrains the algorithm. Specifically, it compares the initial analysis results with the feedback results and makes adjustments necessary to improve the accuracy of the model.

[1162] Step 8:

[1163] The server prepares to apply the updated algorithm to the next analysis. This allows the system to continuously improve itself, gradually increasing the accuracy of the analysis.

[1164] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately, provide useful information to the user, and continuously improve the accuracy of the algorithm.

[1165] (Example 1)

[1166] 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".

[1167] Conventional data analysis systems struggle to analyze large amounts of data quickly and accurately, and are hampered by insufficient data preprocessing and feedback on analysis results. Furthermore, they lack mechanisms to effectively incorporate user feedback and improve analysis accuracy.

[1168] 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.

[1169] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by the terminal, data analysis means by the server, and user feedback recording means. This makes it possible to analyze large amounts of data quickly and accurately, and to improve the accuracy of the analysis by incorporating feedback from the user.

[1170] A "data collection method" is a means by which a server receives and temporarily stores data provided by a user.

[1171] "Preprocessing means" are methods for improving data quality by imputing missing data values, correcting outliers, and eliminating duplicate data.

[1172] A "machine learning algorithm application method" is a method of performing analysis by applying a machine learning algorithm to pre-processed data, extracting data features, and optimizing parameters.

[1173] "Analysis result generation means" refers to a means of generating reports and graphs in a user-friendly format based on analysis results obtained by machine learning algorithms.

[1174] A "feedback mechanism" is a means of recording the user's activity results and sending that data to a server.

[1175] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[1176] "Data preprocessing means by the terminal" refers to means by which the terminal preprocesses data received from the server, specifically performing missing value imputation, outlier correction, and duplicate data removal.

[1177] A "server-based data analysis method" refers to a method in which a server receives pre-processed data and performs analysis by applying machine learning algorithms.

[1178] A "user feedback recording means" is a means for a user to record the results of their activities, input that feedback data into a terminal, and send it to a server.

[1179] This system invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, data preprocessing means by a terminal, data analysis means by a server, and user feedback recording means. The program processing of this system will be described below in natural language.

[1180] First, the server receives a wide variety of data provided by the user. This data, such as sales data and customer support data, is temporarily stored. Next, the server sends the stored data to the terminal for preprocessing.

[1181] The terminal performs preprocessing on the data received from the server. Preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis. Once preprocessing is complete, the data is sent back to the server.

[1182] Next, the server applies a machine learning algorithm to the pre-processed data. This algorithm extracts data features and selects optimal parameters. The analysis results form the basis for providing specific results tailored to the user's needs.

[1183] The server generates reports and graphs in an easy-to-understand format for users based on analysis results obtained from machine learning algorithms. The generated analysis results are shared with the sales team and customer support team and used for subsequent activities. Users perform activities based on the reports provided by the server and record the results on their devices. The devices send the activity results to the server as feedback data.

[1184] The feedback data is used on the server to retrain the machine learning algorithm, improving the accuracy of subsequent analyses. This process allows the system to continuously improve itself, analyze large amounts of data quickly and accurately, and incorporate user feedback to further enhance analysis accuracy.

[1185] As a concrete example, let's consider the analysis of sales data. When a user uploads sales data to the system, the server receives the data. The server sends the received data to a terminal for preprocessing, where the terminal performs tasks such as imputing missing values, correcting outliers, and merging duplicate data. The preprocessed data is sent back to the server, which performs analysis using a machine learning algorithm. The analysis results are generated as a report and provided to the user. The user acts based on this report, records the results on their terminal, and sends them back to the server as feedback data. The server uses this feedback data to retrain its algorithm and improve the accuracy of the next analysis.

[1186] Example of a prompt:

[1187] Collect sales data, apply an algorithm to analyze purchasing patterns based on customer information and transaction history, and generate a list of products to recommend next. Output the results as a report and send it to the user via email. Retrain the algorithm using user feedback data to improve its accuracy.

[1188] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1189] System program processing flow

[1190] Step 1: Data Collection

[1191] 1. Providing user data

[1192] Users input necessary information, such as sales data and customer support data, into the system.

[1193] Input: Sales data, customer support data

[1194] Specific action: The sales representative uploads customer transaction data as a text file.

[1195] 2. Receiving data from the server

[1196] The server receives the uploaded data and stores it temporarily.

[1197] Output: Temporarily saved data file

[1198] Specific action: Receive a file named 'customers.csv' and save it to a specific folder on the server.

[1199] Step 2: Data Preprocessing

[1200] 1. Server data transmission

[1201] The server sends the stored data to the terminal for preprocessing.

[1202] Input: Temporarily saved data file

[1203] Output: Data file sent to the terminal

[1204] Specific action: Send 'customers.csv' to terminal A.

[1205] 2. Data preprocessing at the terminal

[1206] The terminal preprocesses the received data.

[1207] Input: 'customers.csv'

[1208] Output: Preprocessed data file ('data_processed.csv')

[1209] Specific actions: Imput missing values ​​with the mean of past data, correct outliers, and remove duplicate data.

[1210] Step 3: Applying the machine learning algorithm

[1211] 1. Data transmission from the device

[1212] The terminal sends the pre-processed data to the server.

[1213] Input: Preprocessed data file ('data_processed.csv')

[1214] Output: Preprocessed data file sent to the server

[1215] Specific action: Send 'data_processed.csv' to the server.

[1216] 2. Server algorithm application

[1217] The server analyzes the data using machine learning algorithms.

[1218] Input: Preprocessed data file ('data_processed.csv')

[1219] Output: Analysis results (list of proposed products, corresponding methods, etc.)

[1220] Specific operation: Analyzes purchasing patterns using 'data_processed.csv' and generates an optimal product suggestion list.

[1221] Step 4: Generating analysis results

[1222] 1. Server report generation

[1223] The server generates reports and graphs based on the analysis results.

[1224] Input: Analysis results

[1225] Output: Reports and graphs (e.g., 'proposal_list_report.pdf')

[1226] Specific operation: Output the analysis results as a PDF report, including graphed data.

[1227] 2. Server data transmission

[1228] The server sends the generated reports and graphs to the user.

[1229] Input: Reports and graphs

[1230] Output: Reports and graphs sent to the user

[1231] Specific action: Send 'proposal_list_report.pdf' to the user's email address.

[1232] Step 5: Feedback

[1233] 1. User activity log

[1234] Users perform activities based on the generated reports and record the results on their devices.

[1235] Input: Activity results based on reports and graphs

[1236] Output: Recorded activity data (e.g., 'activity_log.xlsx')

[1237] Specific operation: A sales representative proposes a product to a customer and enters the result (whether or not a purchase was made) into the terminal.

[1238] 2. Sending feedback data from the device

[1239] The device sends feedback data to the server.

[1240] Input: Recorded activity data (e.g., 'activity_log.xlsx')

[1241] Output: Feedback data sent to the server

[1242] Specific action: Send 'activity_log.xlsx' to the server.

[1243] Step 6: Update the algorithm

[1244] 1. Server retraining

[1245] The server retrains the machine learning algorithm based on the feedback data, improving the model's accuracy.

[1246] Input: Feedback data (e.g., 'activity_log.xlsx')

[1247] Output: Updated machine learning model

[1248] Specific operation: Readjust the algorithm parameters using 'activity_log.xlsx' and generate a new model.

[1249] (Application Example 1)

[1250] 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".

[1251] Traditional e-commerce sites lacked adequate systems for recommending the most suitable products based on user purchasing behavior and browsing history, making it difficult to accurately recommend products that would be beneficial to the user. As a result, user convenience decreased, and purchasing intent tended to decline. Furthermore, there was a lack of mechanisms to update algorithms and improve recommendation accuracy based on feedback.

[1252] 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.

[1253] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means for recommending the optimal product based on the user's purchasing behavior and browsing history. This makes it possible to accurately recommend the optimal product based on the user's purchasing behavior and browsing history, improving user convenience and potentially increasing purchasing intent. Furthermore, the accuracy of the algorithm can be continuously improved using feedback data.

[1254] "Data collection methods" refer to means of collecting various forms of data, such as users' purchasing behavior and browsing history.

[1255] "Preprocessing means" are methods for improving the quality of collected data by imputing missing values, correcting outliers, and eliminating duplicate data.

[1256] A "machine learning algorithm application method" is a means of applying a machine learning algorithm to pre-processed data, extracting data features, selecting optimal parameters, and performing analysis.

[1257] "Analysis result generation means" refers to a means of generating reports, recommendation lists, and other similar documents in a user-friendly format based on analysis results obtained by machine learning algorithms.

[1258] A "feedback mechanism" is a means of improving the accuracy of analysis results by recording the user's activity results and sending those results to a server.

[1259] An "algorithm update method" is a means of retraining a machine learning algorithm based on feedback data to improve the accuracy of the analysis.

[1260] A "recommendation method" is a means of recommending the most suitable products based on a user's purchasing behavior and browsing history.

[1261] A system for carrying out this invention includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and recommendation means.

[1262] The server has data collection capabilities to gather various types of data, such as user purchasing behavior and browsing history. This data is temporarily stored and used for subsequent data analysis.

[1263] The terminal has a preprocessing mechanism that performs preprocessing on data received from the server. This preprocessing involves imputing missing values, correcting outliers, and removing duplicate data. This improves data quality and increases the accuracy of the analysis.

[1264] The server has a means of applying machine learning algorithms to pre-processed data. This algorithm extracts data features, selects optimal parameters, and performs analysis. Specifically, it analyzes user interests and purchasing trends from purchase history and browsing history.

[1265] The server has a means of generating analysis results that, based on the analysis results obtained by machine learning algorithms, produce reports and recommendation lists in a user-friendly format. The generated analysis results are provided to the user for viewing and recommend products they should purchase next.

[1266] The feedback mechanism records user activity and sends it to the server. For example, it collects information such as whether the user actually purchased or viewed a recommended product. This information is used for subsequent analysis.

[1267] The algorithm update mechanism plays a role in retraining the machine learning algorithm based on feedback data, thereby improving the accuracy of subsequent analyses. Through this process, the system continuously improves itself, and the accuracy of the analysis gradually increases.

[1268] The hardware required includes a server for data collection, preprocessing, and training and application of machine learning models. Furthermore, smartphones are used to collect user purchase and browsing history and display the analysis results. The software used also includes the following:

[1269] Python: The main programming language used in programs.

[1270] pandas: A library for data collection and preprocessing.

[1271] scikit-learn: A library used to apply machine learning algorithms.

[1272] The following is a specific example. In this example, a system is designed to recommend the most suitable products based on the user's purchase and browsing history.

[1273] ---

[1274] Design a system that recommends the most suitable products based on a user's purchase and browsing history. This system will collect large amounts of data, perform preprocessing, apply machine learning algorithms, and provide valuable information to the user. As an example, show how to recommend new products based on the categories of products the user has previously purchased.

[1275] ---

[1276] By implementing such a system, it is possible to increase users' purchasing intent and improve the convenience of online shopping sites.

[1277] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1278] Step 1:

[1279] The server collects data. Specifically, the server collects and temporarily stores users' purchasing behavior and browsing history. This includes information about the user's past purchase history and the product pages they have viewed. The input data is user behavior data, and the output is the temporarily stored raw data.

[1280] Step 2:

[1281] The terminal performs preprocessing. Using the raw data received from the server, it imputes missing values, corrects outliers, and removes duplicate data. Specifically, missing values ​​are imputed with the mean or median of past data, and outliers are corrected based on predefined rules. The input data is temporarily stored raw data, and the output is preprocessed data.

[1282] Step 3:

[1283] The server applies machine learning algorithms to pre-processed data. Specifically, it extracts data features, selects optimal parameters, and performs analysis. Specific algorithms used include random forests and neural networks. The input data is pre-processed data, and the output is the analysis result.

[1284] Step 4:

[1285] The server generates reports and recommendation lists in an easy-to-understand format for the user based on the analysis results. Specifically, it lists products that the user might be interested in and displays them in graphs and tables. The input data is the analysis results, and the output is the reports and recommendation lists provided to the user.

[1286] Step 5:

[1287] Users review the provided recommendation lists and reports and then click on or purchase products. The user's actions are recorded by their device and sent to the server as feedback. The input data is the user's actions, and the output is feedback data.

[1288] Step 6:

[1289] The server retrains its machine learning algorithm based on the feedback data. Specifically, it uses the collected feedback data to update the algorithm's parameters and improve the accuracy of the next analysis. The input data is the feedback data, and the output is the updated machine learning model.

[1290] 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.

[1291] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes the user's emotions. By having each means and the emotion engine work together, it is possible to analyze large amounts of data quickly and accurately, and to provide results that take the user's emotions into consideration.

[1292] System program processing overview

[1293] Data collection

[1294] The server receives and temporarily stores data provided by users. This data comes in various forms, such as sales data and customer support data. The emotion engine also collects emotion information from user input data and real-time interactions.

[1295] Pre-treatment

[1296] The terminal preprocesses the data received from the server. It checks for missing values ​​and imputes them with the mean or median of past data. It detects outliers and corrects them to reasonable values. It also removes duplicate data to create a consistent dataset. Furthermore, the sentiment engine analyzes the collected sentiment data and includes it in the preprocessing results.

[1297] Application of machine learning algorithms

[1298] The server applies a machine learning algorithm to the pre-processed data. The algorithm extracts data features, selects optimal parameters, and performs analysis. During this process, sentiment data provided by the sentiment engine is also considered as an element of the analysis.

[1299] Generation of analysis results

[1300] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on the emotional data provided by the emotion engine. For example, if the user is experiencing stress, a clearer and more concise report will be provided.

[1301] feedback

[1302] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then uses these results to carry out their actual activities. The terminal records the user's activity results and emotional information and sends it to the server. This allows for an understanding of what emotional states influence activity results.

[1303] Algorithm update

[1304] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. At this time, data on user emotions is also included in the training dataset, allowing for a more accurate consideration of the influence of emotions in subsequent analyses.

[1305] Specific example

[1306] Example 1: Analysis of sales data

[1307] 1. Data Collection: The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data.

[1308] 2. Preprocessing: The terminal fills in missing values ​​in the transaction history, corrects outliers, and consolidates duplicate customer information. The sentiment engine analyzes sentiment data and incorporates it into the preprocessing.

[1309] 3. Application of machine learning algorithms: The server analyzes customer purchasing patterns and user sentiment data to generate a list of products to suggest next.

[1310] 4. Generation of analysis results: The server generates a list of emotionally sensitive product suggestions and provides them to the sales team.

[1311] 5. Feedback: The user (sales representative) presents the proposed product to the customer and records the result and emotional state on the device. This feedback data is sent to the server.

[1312] 6. Algorithm Update: The server uses feedback data to retrain the algorithm and improve the accuracy of the next proposal.

[1313] Example 2: Analysis of customer support data

[1314] 1. Data Collection: The server receives inquiry data from customer support, and the sentiment engine collects the sentiment of the user being supported.

[1315] 2. Preprocessing: The terminal removes noise from the query content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into the preprocessing.

[1316] 3. Application of machine learning algorithms: The server analyzes the inquiry history and sentiment data to propose the optimal response method.

[1317] 4. Generation of analysis results: The server generates emotionally sensitive recommended responses and provides them to the support staff.

[1318] 5. Feedback: The support staff member implements the suggested solution and records the results and emotional state on their device. This feedback data is sent to the server.

[1319] 6. Algorithm Update: The server retrains its algorithm based on feedback data to improve the accuracy of its next response.

[1320] Thus, the system of the present invention takes user emotions into consideration throughout a series of processes, from data preprocessing to the application of machine learning algorithms, generation of analysis results, feedback, and algorithm updates, to provide more appropriate and effective results.

[1321] The following describes the processing flow.

[1322] Step 1:

[1323] The server receives sales data and customer support data from users. This data, such as transaction history and customer information, is temporarily stored on the server. The emotion engine collects text and voice data entered by users and extracts emotional information.

[1324] Step 2:

[1325] The terminal begins preprocessing the data received from the server. Specifically, it checks for missing values ​​and imputes them with the mean or median of past data as needed. Next, it detects outliers in the data and corrects them to within the range of past normal values. Furthermore, it removes duplicate data and generates a consistent dataset.

[1326] Step 3:

[1327] The emotion engine analyzes the collected emotion data. The emotion engine extracts emotional states such as positive, negative, and neutral from the user's text and voice, and adds this as emotion data to the preprocessed results.

[1328] Step 4:

[1329] The server applies machine learning algorithms to the pre-processed data and sentiment data. Here, it extracts key features from the data, selects optimal parameters, and performs analysis. By incorporating sentiment data as a feature, it enables more accurate analysis based on user emotions.

[1330] Step 5:

[1331] The server generates reports and suggestion lists for the user based on analysis results obtained by machine learning algorithms. These results are then customized to reflect the user's emotions based on sentiment data from the emotion engine. For example, if the user is feeling stressed, the results are presented in a simple and intuitive interface.

[1332] Step 6:

[1333] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then begins their actual activities based on the displayed analysis results. For example, a sales representative might propose products to a customer based on a proposal list, and a support staff member might implement the recommended course of action.

[1334] Step 7:

[1335] The device meticulously records the user's activity results, including specific outcomes such as the success or failure of product proposals by sales representatives and the success rate of inquiry resolution by support staff. Additionally, the emotion engine monitors the user's emotions in real time during their activity and updates the emotion data as needed.

[1336] Step 8:

[1337] The device sends the recorded feedback data to the server. The server analyzes the transmitted feedback data and checks for any gaps with the initial analysis results. In particular, it analyzes how emotional data influenced the activity results.

[1338] Step 9:

[1339] The server uses feedback data to retrain the machine learning algorithm. It compares the initial analysis results with the feedback results and makes adjustments as needed to improve the model's accuracy. By incorporating sentiment data into the retraining process, the next analysis will more accurately reflect the user's emotions.

[1340] Step 10:

[1341] The server prepares to apply the updated algorithm to the next analysis. This process allows the system to continuously improve itself, progressively increasing the accuracy of the analysis.

[1342] Through the processing steps described above, the system of the present invention can analyze data quickly and accurately and provide optimal results that take into account the user's feelings.

[1343] (Example 2)

[1344] 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".

[1345] Traditionally, data analysis systems have involved data preprocessing and the application of machine learning algorithms, but they have faced the challenge of failing to provide analysis results that take user emotions into account. In particular, analysis that ignores the impact of user emotional states on activity outcomes is insufficient in providing results that are appropriate for actual activities.

[1346] 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.

[1347] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and emotion engine means for recognizing user emotions. This makes it possible to incorporate user emotions into the analysis process and provide emotion-sensitive analysis results. Furthermore, by using user emotions and feedback data to retrain the algorithm, the accuracy of the model can be improved, and more appropriate suggestions and solutions can be provided.

[1348] "Data collection means" refers to a means of receiving and temporarily storing data provided by users or data generated within a system.

[1349] "Preprocessing means" refers to means of performing the following actions on the received data: imputing missing values, correcting outliers, removing duplicate data, and integrating sentiment data.

[1350] "Method for applying machine learning algorithms" refers to a method for applying machine learning algorithms to preprocessed data, extracting data features, and optimizing parameters while taking sentiment data into consideration.

[1351] "Analysis result generation means" refers to a means of generating reports and graphs for users based on analysis results obtained by machine learning algorithms.

[1352] A "feedback mechanism" is a means of recording the user's activity results and emotional information and sending it to a server.

[1353] An "algorithm update method" is a means of improving the accuracy of a model by retraining a machine learning algorithm based on feedback data.

[1354] An "emotion engine" is a means of collecting emotional information from user input data and real-time interactions, and incorporating it into the analysis process.

[1355] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. The following describes specific embodiments for implementing this system.

[1356] Data collection

[1357] The server receives sales data, customer support data, and other data provided by users and temporarily stores this data. The server receives data using communication methods such as REST APIs and WebSockets. The emotion engine collects emotion information from user input data and real-time interactions. This emotion data is collected using speech emotion recognition APIs and Sentiment Analysis APIs.

[1358] Pre-treatment

[1359] The terminal performs data cleaning on the data received from the server, checking for missing values ​​and imputing them as needed. Python libraries such as Pandas are used for this process. Furthermore, the terminal uses Scikit-learn's anomaly detection module and SQL queries to detect and correct outliers and remove duplicate data. The sentiment engine also analyzes the collected sentiment data and integrates it into the preprocessing results. Natural Language Toolkit (NLTK) and SpaCy are used for this task.

[1360] Application of machine learning algorithms

[1361] The server applies a machine learning algorithm to the preprocessed data. The algorithm extracts features from the data and selects optimal parameters, taking sentiment data into consideration. In this case, TensorFlow or Scikit-learn are used.

[1362] Generation of analysis results

[1363] The server generates reports and graphs for the user based on analysis results obtained by machine learning algorithms. The reports and suggestions are customized based on sentiment data provided by the sentiment engine. For example, if the user is experiencing stress, a clearer and more concise report is provided. Specifically, graphs are created using Matplotlib and Seaborn, and reports are generated using a template engine such as Jinja2.

[1364] feedback

[1365] The terminal displays the generated analysis results to the user. The user (sales representative or support staff) then performs actual activities based on these results, records the activity results and emotional information on the terminal, and sends it to the server. This feedback data allows for a detailed understanding of which emotional states influence the activity results. Specifically, the terminal displays the results using HTML / CSS and sends the data to the server using AJAX.

[1366] Algorithm update

[1367] The server retrains the machine learning algorithm based on the feedback data. It compares the initial analysis results with the feedback results and makes necessary adjustments to improve the model's accuracy. Including user sentiment data in the dataset allows for a more accurate consideration of the impact of emotions in subsequent analyses. TensorFlow and Keras are used for retraining.

[1368] Specific example

[1369] Example 1: Analysis of sales data

[1370] The server receives customer information and transaction history from the sales team, and the emotion engine collects user emotions from interaction data. In the preprocessing step, the terminal fills in missing values ​​in the transaction history, corrects outliers, and merges duplicate customer information. The emotion engine analyzes the emotional data and incorporates it into the preprocessing. By applying machine learning algorithms, the server analyzes customer purchasing patterns and emotional data to generate a list of products to recommend next. An emotionally sensitive product recommendation list is provided to the sales team. Sales representatives present the recommended products to customers and record the results and emotional state. This feedback data is sent to the server and used to retrain the algorithm.

[1371] Example 2: Analysis of customer support data

[1372] Customer support inquiry data is received by the server, and the sentiment engine collects the emotions of the user being supported. The terminal removes noise from the inquiry content and extracts important attributes. The sentiment engine analyzes the collected sentiment data and incorporates it into preprocessing. The server analyzes the inquiry history and sentiment data to determine the best course of action and generates sentiment-sensitive recommended responses. The support staff implement the recommended response and record the results and sentiment status. This feedback data is sent to the server and used to retrain the algorithm.

[1373] Example of a prompt

[1374] "Generate product suggestions based on customer purchasing patterns using sales data. These suggestions are suitable for situations where the user is experiencing stress."

[1375] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1376] Step 1:

[1377] The server receives and temporarily stores sales data and customer support data provided by users. Specifically, the server receives data using REST APIs or WebSockets. The input in this process is data from the user, and the output is data stored on the server.

[1378] Step 2:

[1379] The server uses an emotion engine to collect user interaction data and analyze emotional information. The input is user interaction data, and the output is the analyzed emotional data. Specifically, the emotion engine uses the Sentiment Analysis API and the Speech Emotion Recognition API to collect and analyze emotional data.

[1380] Step 3:

[1381] The terminal performs data cleaning on the data received from the server, checking for and imputing missing values. The input is the data obtained from the server, and the output is the imputed data. Specifically, the terminal uses the Python library Pandas to imputate missing values.

[1382] Step 4:

[1383] The terminal detects and corrects anomalies in the data. The input is the augmented data, and the output is the corrected data. Specifically, the terminal uses the Scikit-learn anomaly detection module.

[1384] Step 5:

[1385] The terminal removes duplicate data and creates a consistent dataset. The input is corrected data, and the output is formatted data. Specifically, the terminal uses SQL queries and Pandas' duplicate detection methods.

[1386] Step 6:

[1387] The emotion engine analyzes the collected emotion data and integrates it into the preprocessing results. The input consists of formatted data and emotion data, and the output is preprocessed data with integrated emotion data. Specifically, the emotion engine uses Natural Language Toolkit (NLTK) or SpaCy to integrate the emotion data.

[1388] Step 7:

[1389] The server applies machine learning algorithms to preprocessed data, extracts data features, and selects optimal parameters. The input is preprocessed data with integrated sentiment data, and the output is the analysis result. Specifically, the server uses TensorFlow or Scikit-learn to execute machine learning algorithms.

[1390] Step 8:

[1391] The server generates reports and graphs for the user based on the analysis results obtained from machine learning algorithms. The input is the analysis results, and the output is the user-facing reports and graphs. Specifically, the server uses Matplotlib or Seaborn to create graphs and a template engine such as Jinja2 to generate reports.

[1392] Step 9:

[1393] The terminal displays the generated analysis results to the user. Input is user-facing reports and graphs, while output is user feedback. Specifically, the terminal displays the results using HTML / CSS and sends user feedback to the server using AJAX.

[1394] Step 10:

[1395] The server retrains the machine learning algorithm based on feedback data. The input is the feedback data, and the output is the updated model. Specifically, the server uses TensorFlow or Keras to retrain the algorithm and improve the model's accuracy.

[1396] In this way, through a series of steps, the system can perform data preprocessing, analysis, result generation, user feedback, and model updates, enabling it to provide more accurate and effective results that take user sentiment into account.

[1397] (Application Example 2)

[1398] 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".

[1399] Traditional data analysis systems provided results based purely on data analysis, without considering user emotions. Therefore, if a user was experiencing a particular emotional state, the analysis results and recommendations may not be optimal, potentially significantly detracting from the user experience. Especially in e-commerce sites, product recommendations and customer support tailored to customer emotions are crucial, and a system capable of effectively achieving this was needed.

[1400] 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.

[1401] In this invention, the server includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, means including an emotion engine that recognizes user emotions, and means for customizing analysis results based on real-time emotion data. This makes it possible to provide analysis results and suggestions that take user emotions into consideration.

[1402] "Data collection means" refers to devices or mechanisms that collect and temporarily store user behavior data and real-time emotional data.

[1403] "Preprocessing means" refers to devices or mechanisms for preprocessing collected data, such as imputing missing values, correcting outliers, removing duplicate data, and preprocessing sentiment data.

[1404] A "machine learning algorithm application means" refers to a device or mechanism that applies a machine learning algorithm to preprocessed data, extracts data features, and optimizes parameters.

[1405] "Analysis result generation means" refers to a device or mechanism that generates reports and proposals for users based on analysis results obtained by machine learning algorithms.

[1406] A "feedback mechanism" is a device or mechanism that displays the generated analysis results and suggestions to the user, records the user's activity results and emotional information, and transmits it to the system.

[1407] An "algorithm update mechanism" is a device or mechanism that retrains a machine learning algorithm based on feedback data to improve the accuracy of the model.

[1408] An "emotion engine" is a device or mechanism that analyzes user input data and interactions to recognize the user's emotions.

[1409] "Means for customizing analysis results based on real-time emotional data" refers to devices or mechanisms that adjust analysis results and suggestions according to the user's emotional state, based on real-time recognized user emotional data.

[1410] This invention is a system that includes data collection means, preprocessing means, machine learning algorithm application means, analysis result generation means, feedback means, algorithm update means, and an emotion engine that recognizes user emotions. By coordinating these means, it is possible to rapidly and accurately analyze large amounts of data and provide results that take user emotions into consideration.

[1411] System Overview

[1412] 1. Data collection:

[1413] The server collects and temporarily stores user behavior data (such as product browsing and purchase history) and real-time sentiment data. The sentiment engine uses the Emotion_recognition API to perform facial recognition and voice analysis of the user and obtain real-time sentiment data.

[1414] 2. Pre-processing:

[1415] The terminal preprocesses the data received from the server. Missing values ​​are imputed with the mean or median, and outliers are corrected using standard methods. Duplicate data is removed to create a consistent dataset. Sentiment data acquired by the sentiment engine is also incorporated into the preprocessing.

[1416] 3. Application of machine learning algorithms:

[1417] The server applies machine learning algorithms to the pre-processed data. It extracts data features using algorithms such as RandomForestClassifier, selects optimal parameters, and performs analysis. Sentiment data is also considered an important factor during this process.

[1418] 4. Generating analysis results:

[1419] The server generates product suggestions and reports based on analysis results obtained by machine learning algorithms. If a user is experiencing stress, concise and easy-to-understand suggestions are generated based on data from the emotion engine. For example, a prompt such as "Generate an easy-to-understand product description for a user experiencing stress" might be used.

[1420] 5. Feedback:

[1421] Users view and consider suggested products, providing feedback including their emotional state during the process. The device records this feedback data and sends it to the server. This data is then used for future analysis.

[1422] 6. Algorithm update:

[1423] The server retrains the machine learning algorithm based on feedback data to improve the model's accuracy. It compares the feedback results with the initial analysis results and makes necessary adjustments. Since data on user emotions is also included in the training dataset, the influence of emotions on subsequent analysis results can be more accurately understood.

[1424] In this way, analysis systems that incorporate emotion recognition are an effective means of improving the user experience. A concrete example is a system used on an e-commerce site that analyzes customer purchasing behavior and emotional data to provide optimal product recommendations.

[1425] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1426] Step 1:

[1427] Data collection

[1428] Input: User behavior data (product browsing and purchase history), sentiment data

[1429] Specific actions:

[1430] The server collects user behavior data. This behavior data includes information such as which pages the user viewed, which products they added to their cart, and which products they purchased. Furthermore, it uses an emotion engine to perform facial recognition and voice analysis of users via the Emotion_recognition API, collecting real-time emotion data.

[1431] Output: Collected behavioral and emotional data

[1432] Step 2:

[1433] Pre-treatment

[1434] Input: Collected behavioral data, emotional data

[1435] Specific actions:

[1436] The terminal performs missing value imputation, outlier correction, and duplicate data removal. Specifically, it imputates missing values ​​in behavioral data with the mean or median, and corrects outliers using standard methods. Duplicate data is removed to create a consistent dataset. Furthermore, sentiment data collected by the sentiment engine is also incorporated into the preprocessing.

[1437] Output: Consistent preprocessed data

[1438] Step 3:

[1439] Application of machine learning algorithms

[1440] Input: Preprocessed data (behavioral data, emotional data)

[1441] Specific actions:

[1442] The server applies machine learning algorithms to the preprocessed data. Specifically, it uses RandomForestClassifier to extract data features and selects optimal parameters. It also includes sentiment data in the analysis and generates an optimal model based on the characteristics of the data.

[1443] Output: Analysis results (e.g., a list of optimal product suggestions for each user)

[1444] Step 4:

[1445] Generation of analysis results

[1446] Input: Analysis results obtained from machine learning algorithms, real-time sentiment data

[1447] Specific actions:

[1448] The server generates reports and suggestions for the user based on analysis results obtained by machine learning algorithms. Based on real-time data from the emotion engine, it provides concise and easy-to-understand suggestions if the user is experiencing stress. For example, it might use a prompt such as, "Generate an easy-to-understand product description for a user who is experiencing stress."

[1449] Output: Customized proposals and reports

[1450] Step 5:

[1451] feedback

[1452] Input: Provided suggestions, user actions, sentiment data

[1453] Specific actions:

[1454] Users consider the suggested products, make purchases, or continue browsing. Their actions and emotional states are collected as feedback. The device records this feedback data and sends it to the server. This allows for the accumulation of important data for future analysis.

[1455] Output: User feedback data

[1456] Step 6:

[1457] Algorithm update

[1458] Input: Feedback data, analysis result data

[1459] Specific actions:

[1460] The server retrains its machine learning algorithm based on feedback data received from users. It compares the initial analysis results with the feedback results, makes necessary adjustments, and improves the algorithm's accuracy. By incorporating user sentiment data into the retraining process, it can provide more accurate results in subsequent analyses.

[1461] Output: Updated machine learning model

[1462] 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.

[1463] 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.

[1464] 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.

[1465] 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.

[1466] 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.

[1467] 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.

[1468] 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.

[1469] 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.

[1470] 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."

[1471] 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.

[1472] 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.

[1473] 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.

[1474] 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.

[1475] 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.

[1476] 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.

[1477] 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.

[1478] 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.

[1479] 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.

[1480] 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.

[1481] 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.

[1482] 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.

[1483] The following is further disclosed regarding the embodiments described above.

[1484] (Claim 1)

[1485] Data collection means,

[1486] Pre-treatment means,

[1487] A means of applying machine learning algorithms,

[1488] means for generating analysis results,

[1489] Feedback methods,

[1490] Algorithm update method,

[1491] A system that includes this.

[1492] (Claim 2)

[1493] The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting outliers, and means for eliminating duplicate data.

[1494] (Claim 3)

[1495] The system according to claim 1, wherein the machine learning algorithm application means includes means for extracting data features and optimizing parameters.

[1496] "Example 1"

[1497] (Claim 1)

[1498] Data collection means,

[1499] Pre-treatment means,

[1500] A means of applying machine learning algorithms,

[1501] means for generating analysis results,

[1502] Feedback methods,

[1503] Algorithm update method,

[1504] Data preprocessing means by terminal,

[1505] A data analysis method using a server,

[1506] User feedback recording means,

[1507] A system that includes this.

[1508] (Claim 2)

[1509] The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting abnormal values, and means for eliminating duplicate data, and the server includes means for transmitting the stored data to a terminal.

[1510] (Claim 3)

[1511] The system according to claim 1, wherein the machine learning algorithm application means includes means for extracting data features and optimizing parameters, and means for generating reports and graphs as analysis results and sending them to the user.

[1512] "Application Example 1"

[1513] (Claim 1)

[1514] Data collection means,

[1515] Pre-treatment means,

[1516] A means of applying machine learning algorithms,

[1517] means for generating analysis results,

[1518] Feedback methods,

[1519] Algorithm update method,

[1520] A recommendation system that suggests the most suitable products based on the user's purchasing behavior and browsing history,

[1521] A system that includes this.

[1522] (Claim 2)

[1523] The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting outliers, and means for removing duplicate data, and the recommendation means further includes means for applying a machine learning algorithm based on the preprocessed data.

[1524] (Claim 3)

[1525] The system according to claim 1, wherein the machine learning algorithm application means includes means for extracting data features and optimizing parameters, and the recommendation means further includes means for recommending products based on the analysis results.

[1526] "Example 2 of combining an emotion engine"

[1527] (Claim 1)

[1528] Data collection means,

[1529] Pre-treatment means,

[1530] A means of applying machine learning algorithms,

[1531] means for generating analysis results,

[1532] Feedback methods,

[1533] Algorithm update method,

[1534] An emotion engine that recognizes the user's emotions,

[1535] A system that includes this.

[1536] (Claim 2)

[1537] The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting outliers, means for eliminating duplicate data, and further means for analyzing sentiment data and integrating it into the preprocessing results.

[1538] (Claim 3)

[1539] The system according to claim 1, wherein the machine learning algorithm application means includes means for extracting data features and optimizing parameters by taking sentiment data into consideration.

[1540] "Application example 2 when combining with an emotional engine"

[1541] (Claim 1)

[1542] Data collection means,

[1543] Pre-treatment means,

[1544] A means of applying machine learning algorithms,

[1545] means for generating analysis results,

[1546] Feedback methods,

[1547] Algorithm update method,

[1548] A means including an emotion engine that recognizes the user's emotions,

[1549] A means of customizing analysis results based on real-time sentiment data,

[1550] A system that includes this.

[1551] (Claim 2)

[1552] The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting outliers, means for removing duplicate data, and means for including sentiment data in the preprocessing.

[1553] (Claim 3)

[1554] The system according to claim 1, wherein the machine learning algorithm application means includes means for extracting data features and optimizing parameters, and means for including sentiment data in the analysis. [Explanation of Symbols]

[1555] 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. A data collection method that receives and temporarily stores data provided by users, A preprocessing means for the received data, which includes imputing missing values, correcting outliers, and removing duplicate data. A means for applying a machine learning algorithm that extracts features from preprocessed data, selects optimal parameters, and performs analysis, An analysis result generation means that outputs the analysis results obtained by the algorithm in a format that is easy for the user to understand, A feedback mechanism that records user activity results and sends them to a server, An algorithm update method that retrains the algorithm based on feedback data, A system that includes this.

2. The system according to claim 1, wherein the preprocessing means includes means for imputing missing values, means for correcting outliers, and means for eliminating duplicate data.

Images