System

Abstract

A system is provided.SOLUTION: A system, comprising: means for obtaining merchant information from a database; means for preprocessing the obtained merchant information; means for splitting the preprocessed merchant information into training data and test data; means for training a machine-learning model using the training data; means for evaluating the model with the test data; and means for preprocessing new merchant information and calculating a credit score using the trained model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Conventional merchant scoring requires time and effort because the evaluation is performed manually, resulting in inconsistencies in the quality and speed of the evaluation. Furthermore, manual evaluation makes it difficult to process large amounts of data, reducing the efficiency of the risk assessment and overall screening process. Furthermore, manual evaluation is prone to human error, making it difficult to achieve consistent credit scoring. [Means for solving the problem]

[0005] The present invention provides a system including means for acquiring merchant information from a database, means for preprocessing the acquired merchant information, means for splitting the preprocessed merchant information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model using the test data, and means for preprocessing new merchant information and calculating a credit score using the trained model. This automates the manual evaluation process and enables fast and accurate calculation of credit scores. Furthermore, by increasing the consistency of the splitting of training data and test data, model training, and evaluation process, human error is reduced, resulting in more reliable credit scoring. Furthermore, the present invention uses a random forest classifier to prevent overfitting and enable more accurate predictions.

[0006] A "database" is a system for centrally managing multiple data sets, and has the ability to efficiently store, search, and retrieve large amounts of data, such as member store information.

[0007] "Affiliated store information" is detailed data about each affiliated store, specifically including attributes such as store size, annual sales, and industry.

[0008] "Preprocessing" is the process of converting acquired data into a format that can be used by a machine learning model, and includes removing missing values ​​and converting categorical data into numeric values.

[0009] "Training data" is the data set used to train a machine learning model; it is the input information from which the model learns patterns.

[0010] "Test data" is a dataset used to evaluate the performance of a trained machine learning model, and is validation data used to confirm the accuracy and validity of the model.

[0011] A "machine learning model" is a collection of algorithms that learn specific patterns and rules from data, and is a trained model that uses training data to make predictions and classifications.

[0012] A "random forest classifier" is an ensemble learning algorithm that uses multiple decision trees to prevent overfitting and make more accurate predictions.

[0013] A "credit score" is a numerical indicator of a merchant's credit risk, calculated based on the merchant's reliability and risk assessment. [Brief explanation of the drawings]

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

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

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

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

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

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

[0020] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0021] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0022] [First embodiment]

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

[0024] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

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

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

[0027] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0028] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0029] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0031] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

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

[0033] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

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

[0035] The system of the present invention improves the accuracy and efficiency of credit evaluation by automating credit scoring of merchants. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0036] The server first connects to the database and retrieves information about affiliated stores, including store size, annual sales, industry, etc. The retrieved data cannot be used directly in a machine learning model, so preprocessing is required.

[0037] During preprocessing, the server first removes missing values ​​from the dataset and then converts categorical data, such as industry, into numerical data using a method called one-hot encoding, which prepares the data for processing by machine learning models.

[0038] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0039] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. The calculated credit score is provided to the user.

[0040] Specific examples

[0041] For example, consider the following new merchant information being entered:

[0042] Store size: 100

[0043] Annual sales: 5000000

[0044] Industry: Food and beverage

[0045] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0046] Users can view the resulting credit scores and use them to improve their credit rating and underwriting processes, resulting in faster and more consistent credit scoring than traditional manual processes.

[0047] The above is an embodiment of the present invention. This system significantly improves the efficiency and accuracy of merchant scoring, and increases the reliability of risk assessment and screening processes.

[0048] The processing flow will be explained below.

[0049] Step 1:

[0050] The server connects to the database and retrieves merchant information. To retrieve merchant information from the database, it executes an SQL query and converts the results into a data frame. The retrieved merchant information includes attributes such as store size, annual sales, and industry.

[0051] Step 2:

[0052] The server preprocesses the acquired merchant information by removing missing values ​​in the dataset and converting categorical data (e.g., industry) into numerical data using one-hot encoding, which converts the categorical data into a format that can be processed by machine learning models.

[0053] Step 3:

[0054] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly. The training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0055] Step 4:

[0056] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees to learn patterns from the training data, which prevents the model from overfitting and gives it the ability to make accurate predictions.

[0057] Step 5:

[0058] The server evaluates the trained model with test data. It inputs the test data into the model and compares the credit scores predicted by the model with the actual credit scores. From this comparison, it calculates the accuracy of the model and evaluates the performance of the model.

[0059] Step 6:

[0060] The user enters information about the new franchisee into the system, such as store size, annual sales, and industry. This information is used to calculate the new franchisee's credit score.

[0061] Step 7:

[0062] The server pre-processes the new merchant information entered, converting categorical data into numerical data, and then inputs the pre-processed data into a trained random forest model.

[0063] Step 8:

[0064] The server uses the trained model to calculate a credit score for the new merchant, which serves as a benchmark for assessing the merchant's credit risk.

[0065] Step 9:

[0066] The server provides the calculated credit score to the user, which provides the user with information that can be used to assess the risk of new merchants and improve the screening process.

[0067] Example 1

[0068] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0069] Conventional credit scoring systems require a lot of time and effort for data preprocessing, model training, and evaluation, which can result in insufficient accuracy and efficiency in the assessment. Furthermore, many manual processes make it difficult to achieve consistent scoring. Furthermore, delays in processing new merchant information can lead to lost business opportunities.

[0070] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0071] In this invention, the server includes means for acquiring company information from a database, means for preprocessing the acquired company information, means for dividing the preprocessed company information into training data and test data, means for training an inference model using the training data, means for evaluating the model using the test data, means for preprocessing new company information and calculating a reliability using the trained model, and means for providing the results to the user. This allows for efficient credit scoring, improves the consistency and accuracy of evaluations, and enables quick scoring of new franchisees.

[0072] A "database" is a collection of information stored in a format that allows multiple data to be managed and searched efficiently.

[0073] "Company information" refers to various data relating to a company, specifically including information such as store size, annual sales, and industry.

[0074] "Preprocessing" refers to a series of operations that transform data into a format suitable for machine learning models, including removing missing values ​​and converting categorical data into numerical data.

[0075] "Training data" refers to a dataset used to train a machine learning model and improve the model's predictive capabilities.

[0076] "Test data" is a dataset used to evaluate the performance of a machine learning model and is used to confirm the model's generalization ability.

[0077] An "inference model" is a model that uses machine learning algorithms to learn patterns from data and make predictions about unknown data.

[0078] "Trustworthiness" is a numerical indicator that indicates the trustworthiness of a new company, and is calculated using a machine learning model.

[0079] "User" means a person or organization that uses this system to obtain credit scoring results and use them for evaluation or screening.

[0080] The system of the present invention automates the credit scoring of companies, improving the accuracy and efficiency of the evaluation. This system can calculate credit scores using machine learning models based on company information stored in a database.

[0081] The server first connects to the database to retrieve company information. This information includes store size, annual sales, industry, etc. Specific database connection libraries used include JDBC and PDO. The retrieved data is not suitable for machine learning models as is, so preprocessing is required.

[0082] In the preprocessing stage, the server first removes missing values ​​from the dataset and converts categorical data such as industry into numerical data using the dropna method from the pandas library and OneHotEncoder from the sklearn library, converting the data into a format that can be processed by the machine learning model.

[0083] After preprocessing, the data is then split into training data and test data. The server uses the sklearn train_test_split method to split the data, with 80% of the data being used for training and 20% for testing.

[0084] Next, the server trains a machine learning model using a random forest classifier, which uses multiple decision trees to prevent overfitting and make more accurate predictions. After training, the server evaluates the accuracy of the model using test data. sklearn's accuracy_score is used in the evaluation process.

[0085] To calculate a new business's credit score, the user enters information about the new business into the device. Specific input items include:

[0086] Store size: 100

[0087] Annual sales: 5000000

[0088] Industry: Food and beverage

[0089] The server first preprocesses this information. For example, it converts "Industry: Food & Beverage" into numerical data. Then, it inputs the preprocessed data into a trained random forest model to predict a credit score. Users can then view this credit score and use it to improve their rating and screening process.

[0090] Prompt Sentence Examples

[0091] "Calculate the credit score for a new franchisee whose store size is 100, annual sales are 5,000,000, and the industry is food and beverage."

[0092] This concludes the description of an embodiment of the present invention. This system significantly improves the efficiency and accuracy of company scoring, increasing the reliability of risk assessment and screening processes. It also enables faster and more consistent credit scoring than traditional manual processes.

[0093] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0094] Step 1:

[0095] The server connects to the database and retrieves company information. Specifically, it uses a database connection library such as JDBC or PDO. It executes an SQL query to retrieve a dataset containing information such as store size, annual sales, and industry. The input for this step is the database, and the output is the retrieved dataset of company information.

[0096] Step 2:

[0097] The server preprocesses the acquired company information. First, it uses the dropna method in the pandas library to remove missing values ​​in the dataset. Then it uses sklearn's OneHotEncoder to convert categorical data such as industry to numerical data. The input to this step is the acquired company information dataset, and the output is the preprocessed data.

[0098] Step 3:

[0099] The server splits the preprocessed data into training data and test data. It uses sklearn's train_test_split method, with 80% of the data for training and 20% for testing. The input to this step is the preprocessed data, and the output is the training data and test data.

[0100] Step 4:

[0101] The server uses the training data to train a machine learning model using a random forest classifier. The classifier uses sklearn's RandomForestClassifier. To train the model, the training data is used as input and a trained model is output.

[0102] Step 5:

[0103] The server uses the trained model to evaluate the model on the test data. The evaluation uses sklearn's accuracy_score method to calculate the model's predictive accuracy. The inputs to this step are the trained model and the test data, and the output is the model's accuracy score.

[0104] Step 6:

[0105] The user inputs information about the new company from the terminal. This information includes store size, annual sales, and industry. The input for this step is the new company's information, and the output is the input data.

[0106] Step 7:

[0107] The server preprocesses the new business information and calculates the credit score using the trained model. After preprocessing, the new business information is input into the trained model to predict the credit score. Preprocessing also includes converting categorical data to numerical data. The input of this step is the preprocessed new business information, and the output is the credit score.

[0108] Step 8:

[0109] The server provides the calculated credit score to the user, which is used to improve the rating and underwriting process and ultimately to make effective business decisions. The input to this step is the credit score, and the output is the user's confirmation of the credit score.

[0110] (Application example 1)

[0111] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0112] Conventional credit scoring systems require time and effort to obtain and preprocess merchant information and calculate credit scores, making it difficult to perform efficient and accurate scoring. In particular, virtual store operators lack a means to instantly check credit scores, making it difficult to make quick operational decisions.

[0113] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0114] In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, and means for displaying the credit score for the virtual store via the cloud server using the generative AI model, thereby enabling the virtual store operator to check the credit score in real time and make quick decisions.

[0115] A "database" is a system for efficiently storing and managing data such as affiliated store information.

[0116] "Affiliated store information" is detailed data about affiliated stores, such as store size, annual sales, and industry.

[0117] "Preprocessing" is the process of converting data into a format that can be used by a machine learning model, and includes removing missing values ​​and converting categorical data into numerical data.

[0118] "Training data" is a dataset used to train a machine learning model.

[0119] "Test data" is a dataset used to evaluate a trained machine learning model.

[0120] A "machine learning model" is an algorithm that learns knowledge and patterns from data and makes predictions and classifications.

[0121] A "random forest classifier" is a type of machine learning model that combines multiple decision trees to improve prediction accuracy.

[0122] A "generative AI model" is an artificial intelligence algorithm that automatically generates a specific output (such as a prompt) from data.

[0123] A "cloud server" is a remote server system that shares resources over the Internet and performs data storage and computation.

[0124] A "credit score" is a numerical representation of a merchant's creditworthiness and is used in risk assessment and screening processes.

[0125] The system of the present invention automates credit scoring for virtual stores, improving the accuracy and efficiency of evaluation. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0126] Main processing steps

[0127] Retrieving data from a database

[0128] The server first connects to the database and retrieves information about the affiliated store, including store size, annual sales, industry, etc. The server receives the retrieved data in a format that requires preprocessing.

[0129] Data Preprocessing

[0130] The acquired data cannot be used as is for machine learning models, so preprocessing is required. In preprocessing, the server first removes missing values ​​from the dataset. It then converts categorical data, such as industry, into numerical data. This conversion is performed using a method called one-hot encoding. This puts the data in a format that can be processed by machine learning models.

[0131] Splitting the data and training the model

[0132] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and enable more accurate predictions. After training is complete, the server evaluates the model with the test data and calculates its accuracy.

[0133] Calculating and displaying credit scores

[0134] When the system calculates a new merchant's credit score, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.) through the smart glasses. The server preprocesses this new merchant information and inputs it into the trained model to calculate the credit score. The calculated credit score is displayed in real time on the smart glasses' display via the cloud server, allowing the user to check it.

[0135] Specific examples

[0136] For example, consider the following new merchant information being entered:

[0137] Store size: 100

[0138] Annual sales: 5000000

[0139] Industry: Food and beverage

[0140] The server first preprocesses this information. Specifically, it converts "Industry: Food & Beverage" into numerical data. The preprocessed data is then input into a trained random forest model. The model predicts a credit score based on these features. Users can view the resulting credit score and use it to improve their evaluation and screening process.

[0141] The system enables faster and more consistent credit scoring than traditional manual processes, and allows virtual store operators to view credit scores in real time and make necessary decisions quickly.

[0142] Prompt Sentence Examples

[0143] Store size: 100

[0144] Annual sales: 5000000

[0145] Industry: Food and beverage

[0146] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0147] Step 1:

[0148] The server connects to the database and retrieves merchant information, including store size, annual sales, and industry. The input is a database access request, and the output is the retrieved merchant information. This operation includes, for example, issuing an SQL query to retrieve the data.

[0149] Step 2:

[0150] The server preprocesses the acquired data. Specifically, it removes missing values ​​in the merchant information and converts categorical data to numerical data. The input is the raw merchant information, and the output is the preprocessed data. This conversion is performed using the Python Pandas library for data cleansing.

[0151] Step 3:

[0152] The server splits the preprocessed data into training data and test data. As input, it has the preprocessed data, and as output, it has training data and test data. This operation is performed using the train_test_split function of the Scikit-learn library.

[0153] Step 4:

[0154] The server uses the training data to train a machine learning model. For this, it uses a random forest classifier. The training data is the input, and the trained model is the output. For this operation, the model is trained using Scikit-learn's RandomForestClassifier.

[0155] Step 5:

[0156] The server evaluates the trained model with test data. The input is the test data, and the output is the model's evaluation result or accuracy. This operation uses Scikit-learn's evaluation function.

[0157] Step 6:

[0158] The user inputs new franchisee information through the smart glasses, for example, using voice recognition or touch gestures. The input includes new franchisee information such as store size, annual sales, and industry, but there is no output.

[0159] Step 7:

[0160] The server preprocesses the new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information, and the output is the credit score. Specifically, the server repeats the preprocessing in step 2 and then inputs the data into the trained model.

[0161] Step 8:

[0162] The server sends the calculated credit score to the smart glasses via the cloud server and displays it on the display. The credit score is the input, and the credit score displayed on the smart glasses is obtained as the output. This operation uses a communication protocol to send and receive data.

[0163] Specific examples

[0164] Input information: Store size 100, annual sales 5,000,000, industry Food and beverage

[0165] The server pre-processes this information and inputs it into a trained model.

[0166] The calculated credit score, for example "Credit score: 0.85", will be displayed on the smart glasses.

[0167] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0168] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. This system calculates credit scores using a machine learning model based on merchant information stored in a database, and uses the emotion engine to provide display content that corresponds to the user's emotions.

[0169] The server first connects to the database and retrieves member store information. The information retrieved from the database includes attributes such as store size, annual sales, and industry. The retrieved data cannot be used as is in a machine learning model, so preprocessing is required.

[0170] During preprocessing, the server first removes missing values ​​in the dataset and converts categorical data (such as industry) into numerical data using one-hot encoding, which prepares the data for processing by machine learning models.

[0171] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0172] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate a credit score. The calculated credit score is displayed by the emotion engine according to the user's emotions.

[0173] The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the content displayed will be more detailed and reassuring.

[0174] Specific examples

[0175] For example, consider the following new merchant information being entered:

[0176] Store size: 100

[0177] Annual sales: 5000000

[0178] Industry: Food and beverage

[0179] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0180] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0181] For example, if the user looks worried, we might show them a detailed explanation of how the credit score is calculated or some additional information to reassure them, while if the user looks interested, we might offer more advanced analysis and profit predictions.

[0182] In this way, the system of the present invention significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user sentiment, resulting in a more reliable risk assessment and screening process and an improved user experience.

[0183] The processing flow will be explained below.

[0184] Step 1:

[0185] The server connects to the database and retrieves merchant information, including attributes such as store size, annual sales, and industry. Data is retrieved from the database using an SQL query, and the results are converted into a data frame.

[0186] Step 2:

[0187] The server preprocesses the acquired merchant information. Specifically, it removes missing values ​​in the dataset and converts categorical data to numerical data using the following procedure. By converting categorical data (e.g., industry) to numerical data using one-hot encoding, it creates a format that can be processed by the machine learning model.

[0188] Step 3:

[0189] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly, and the training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0190] Step 4:

[0191] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees, each of which learns independently to prevent overfitting and enable more accurate predictions.

[0192] Step 5:

[0193] The server evaluates the trained model with test data. It inputs the test data into the model and compares the model's predictions with the actual credit scores. Based on the comparison results, it calculates the model's accuracy (e.g., accuracy rate) and evaluates its performance.

[0194] Step 6:

[0195] The user enters information about the new franchisee into the system, including the store size, annual sales, industry, etc. This information becomes the basic data for calculating the new franchisee's credit score.

[0196] Step 7:

[0197] The server preprocesses the input new merchant information, converting categorical data into numerical data, and inputs the preprocessed data into a trained random forest model. The preprocessing method is the same as in Step 2.

[0198] Step 8:

[0199] The server uses the trained model to calculate a credit score for the new merchant, and the model's predicted credit score serves as a benchmark for assessing the new merchant's credit risk.

[0200] Step 9:

[0201] The device's built-in camera and microphone analyze the user's behavior in real time and recognize emotions from the user's facial expressions and voice. Based on the analyzed emotional data, the emotion engine recognizes how the user is receiving information (e.g., relief, anxiety, doubt).

[0202] Step 10:

[0203] The server dynamically adjusts the calculated trust score according to the user's emotions and changes the presentation method. For example, if the user looks anxious, it will display detailed explanations and additional information that provides reassurance. On the other hand, if the user shows interest, it will be configured to provide more advanced analysis results and predictive data.

[0204] Step 11:

[0205] The server displays the final credit score and related information to the user in an appropriate format, providing information that can be used to improve risk assessment and screening processes for new merchants.

[0206] Example 2

[0207] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0208] Conventional merchant credit scoring systems require a lot of time and effort for data preprocessing and machine learning model training. Furthermore, they only provide uniform information to users and are unable to respond flexibly to user emotions, resulting in a lack of user experience. This leads to issues such as a decrease in the accuracy and reliability of credit scores and a decline in the efficiency of risk assessment and screening processes.

[0209] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0210] In this invention, the server includes means for acquiring merchant information from a database, means for preprocessing the acquired merchant information, means for dividing the preprocessed merchant information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new merchant information and calculating a credit score using the trained model, means for recognizing user emotions, and means for dynamically adjusting information display content based on user emotions. This significantly improves the efficiency and accuracy of merchant scoring and enables flexible responses according to user emotions.

[0211] A "database" is an information system that systematically organizes and stores information, allowing it to be searched and retrieved efficiently.

[0212] "Affiliated store information" is information about affiliated stores, and is data including attributes such as store size, annual sales, and industry type.

[0213] "Preprocessing" is the process of improving data quality, removing unnecessary information, and converting data into an appropriate format for use in data analysis and machine learning models.

[0214] A "machine learning model" is an algorithm or mathematical model that learns from data and makes decisions such as predictions and classifications.

[0215] A "random forest classifier" is a machine learning algorithm that combines and learns from multiple decision trees to improve the accuracy of classification and prediction.

[0216] The "emotion engine" is a system that analyzes data such as the user's facial expressions and voice to recognize their emotional state.

[0217] "Means for recognizing user emotions" refers to technology that uses a camera or microphone to capture the user's facial expressions and voice and analyze their emotions.

[0218] "Means for dynamically adjusting the content of displayed information" is a mechanism for changing the content and format of the information presented in real time based on the user's emotions.

[0219] A "credit score" is a numerical representation of the reliability and risk of a merchant, and is an indicator used to make decisions about transactions.

[0220] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. A specific implementation of this system is described below.

[0221] The server first connects to a database (for example, an SQL database) and retrieves member store information. This member store information includes attributes such as store size, annual sales, and industry. The information retrieved from the database cannot be used in its raw form for machine learning models, so preprocessing is required. The Python Pandas library is used for data preprocessing, and one-hot encoding is applied to convert categorical data into numerical values.

[0222] After preprocessing, the data is then split into training data and test data. This splitting is done using the Scikit-Learn library. The server uses the training data to train a machine learning model using a random forest classifier. The trained model is then evaluated on the test data to calculate its accuracy.

[0223] When calculating the credit score of a new merchant, the user enters information about the new merchant (e.g., store size, annual sales, industry, etc.) into a dedicated form on the terminal. This information is sent to the server via an HTTP request. The server similarly preprocesses the received information about the new merchant, inputs it into the trained model, and calculates the credit score.

[0224] The calculated credit score is displayed by an emotion engine according to the user's emotions. The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. Emotion recognition is performed using, for example, Microsoft Azure Emotion API. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the display content can be changed to something more detailed and reassuring.

[0225] Specific examples

[0226] For example, consider the following new merchant information being entered:

[0227] Store size: 100

[0228] Annual sales: 5000000

[0229] Industry: Food and beverage

[0230] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data, and then inputs the preprocessed data into a trained random forest model. The model predicts a credit score based on these features.

[0231] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0232] Prompt Sentence Examples

[0233] "Calculate a credit score for a new franchisee based on store size of 100, annual sales of 5,000,000, and industry of food and beverage. Also, explain in natural language the processing steps of the system that adjusts the way information is presented based on the user's emotions."

[0234] Using these prompts, the generative AI explains the overall system flow and the specific processing steps at each step. This system significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user emotions. This increases the reliability of the risk assessment and screening process and improves the user experience.

[0235] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0236] Step 1:

[0237] The server first connects to the database and retrieves member store information. This database connection retrieves attribute data such as "store size," "annual sales," and "industry type." The input is an SQL query, and the output is a dataset of member store information.

[0238] Step 2:

[0239] The server preprocesses the acquired merchant information. This preprocessing includes removing missing values ​​and converting categorical data to numeric data. Specifically, it uses Python's Pandas library to manipulate data frames. The input is the acquired merchant information dataset, and the output is the preprocessed dataset.

[0240] Step 3:

[0241] The server splits the preprocessed dataset into training data and test data. Here, we use the Scikit-Learn library to split the data. The input is the preprocessed dataset, and the output is the training data and test data.

[0242] Step 4:

[0243] The server uses the training data to train a machine learning model, using a random forest classifier, where the input is the training data and the output is the trained machine learning model.

[0244] Step 5:

[0245] The server uses the trained model to evaluate the test data. This evaluation uses various metrics to calculate the accuracy of the model. The inputs are the test data and the trained model, and the output is the accuracy score of the model.

[0246] Step 6:

[0247] The user enters the information of the new affiliated store into a dedicated form on the terminal. The entered information is sent to the server via an HTTP request. The input is the information of the new affiliated store, and the output is the HTTP request sent to the server.

[0248] Step 7:

[0249] The server preprocesses the received new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information and the trained model, and the output is the credit score.

[0250] Step 8:

[0251] The camera and microphone installed on the device analyze the user's facial expressions and voice to recognize emotions. This analysis is performed using an emotion recognition library (e.g., Microsoft Azure Emotion API). The input is the user's facial expressions and voice data, and the output is the recognized emotion data.

[0252] Step 9:

[0253] The server dynamically adjusts the displayed credit score based on the user's emotional data obtained from the emotion engine. Specifically, if the user looks anxious, the server takes action such as providing more information. The input is the credit score and the recognized emotional data, and the output is the adjusted display content.

[0254] (Application example 2)

[0255] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0256] Conventional merchant credit scoring systems take time to process and evaluate data, and lack flexibility in providing users with credit scores. Furthermore, they provide information uniformly without considering users' feelings, which results in a lack of improvement in the user experience. Furthermore, electronic payment services require support that allows users to use them safely and without anxiety. These issues need to be resolved.

[0257] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into learning data and test data, means for training a machine learning model using the learning data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, means for recognizing emotions from the user's facial expressions and voice using a camera and microphone installed in the terminal, and means for dynamically adjusting the presentation method based on the recognized emotion data. This makes it possible to not only score member stores' credit but also flexibly provide information according to the user's emotions, thereby realizing an environment in which users can use electronic payment services with peace of mind.

[0258] A "database" is a structured data storage system for efficiently managing, searching, and retrieving information.

[0259] "Affiliated store information" refers to detailed business data related to affiliated stores, such as store size, annual sales, and industry.

[0260] "Preprocessing" is the process of preparing data before it can be used in a machine learning model, such as removing missing values ​​or converting categorical data to numerical data.

[0261] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications.

[0262] A "credit score" is a numerical value calculated to assess the creditworthiness of a merchant and is used in risk assessment and screening processes.

[0263] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to operate the system.

[0264] A "camera" is an optical device for capturing images of a user's facial expressions and movements.

[0265] A "microphone" is a device for recording the user's voice.

[0266] "Emotion recognition" is a technology that analyzes a user's facial expressions and voice to identify the emotions the user is feeling.

[0267] "Dynamic adjustment" is the process of changing settings and display content in real time based on the situation.

[0268] System Overview

[0269] The system of the present invention automates credit scoring for merchants and dynamically adjusts information presentation methods by combining user emotion recognition. This system is composed of a server, terminals, a database, a machine learning model, an emotion recognition engine, etc.

[0270] Program processing overview

[0271] The server first connects to the database and retrieves information about the member stores. This information includes store size, annual sales, industry, etc. Next, the retrieved member store information is preprocessed. This preprocessing involves removing missing values ​​and converting categorical data into numerical data. This conversion uses techniques such as one-hot encoding.

[0272] After preprocessing, the data is split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The trained model is then evaluated on the test data to calculate its accuracy.

[0273] When calculating the credit score of a new merchant, the user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate the credit score.

[0274] Meanwhile, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize their emotions. The emotion recognition engine uses software such as OpenCV and EmotionRecognizer. This recognized emotion data is acquired by the server and used to dynamically adjust the way the credit score is presented.

[0275] Specific examples

[0276] Consider the case where the following data is entered as information for a new merchant:

[0277] Store size: 100

[0278] Annual sales: 5000000

[0279] Industry: Food and beverage

[0280] The server preprocesses this information, converting the industry into numerical data using one-hot encoding, and then feeds the preprocessed data into a trained random forest model to predict the credit score.

[0281] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how they feel about the information they receive. For example, if the user looks anxious, the system will display detailed explanations and additional information about how the credit score is provided, reassuring the user. On the other hand, if the user looks interested, the system will provide more advanced analysis results and profit predictions.

[0282] Prompt Sentence Examples

[0283] plaintext

[0284] Information about new franchisee "Restaurant A":

[0285] Store size: 100

[0286] Annual sales: 5000000

[0287] Industry: Food and beverage

[0288] Calculate a credit score and provide detailed information based on the user's "anxiety" emotion.

[0289] The system of the present invention can significantly improve the accuracy of the information provided and the user experience. By not only providing credit scores for member stores but also providing information flexibly according to the user's emotions, it provides an environment in which users can use electronic payment services with peace of mind.

[0290] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0291] Step 1:

[0292] Obtain affiliated store information from the database. The server connects to the database and obtains affiliated store information such as store size, annual sales, and industry. Specifically, it uses an SQL query to obtain all the information in one go. This requires database connection information as input, and the output is a dataset containing affiliated store information.

[0293] Step 2:

[0294] Preprocess the merchant information. The server removes missing values ​​from the acquired data and converts categorical data into numerical data. Specifically, categorical data such as industry type is converted using one-hot encoding. This input is the output data of step 1, and a preprocessed dataset is obtained as the output.

[0295] Step 3:

[0296] Split into training data and test data. The server splits the preprocessed data into training data and test data. Specifically, it randomly shuffles the dataset and splits it into two parts at a fixed ratio. This input is the output data of step 2, and the output is training data and test data.

[0297] Step 4:

[0298] Train the machine learning model. The server trains a random forest classifier using the training data. Specifically, it fits the model using the features and labels of the training data. This input is the training data from step 3, and the output is a trained model.

[0299] Step 5:

[0300] Evaluate the model with test data. The server evaluates the trained model with test data and calculates its accuracy. Specifically, it uses the test data to make predictions and compares them with the actual values. The input is the test data from step 3, and the output is the accuracy of the model.

[0301] Step 6:

[0302] Preprocessing new merchant information and calculating a credit score. The user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. Specifically, preprocessing is performed in the same way as in step 2, and the model is used in the same way as in step 4. The input is the new merchant information from the user, and the output is a credit score.

[0303] Step 7:

[0304] Recognizes user emotions. The device uses the on-board camera and microphone to recognize emotions from the user's facial expressions and voice. Specifically, OpenCV and EmotionRecognizer are used to perform real-time analysis. The input is video and audio data acquired from the device, and the output is recognized emotion data.

[0305] Step 8:

[0306] The server dynamically adjusts the presentation method based on the emotional data. Based on the recognized emotional data, the server dynamically adjusts the presentation method of the credit score. For example, if the user feels anxious, it will display additional detailed explanations or information that provides reassurance. Specifically, it executes logic that changes the display content based on the emotional data. The inputs are the credit score from step 6 and the emotional data from step 7, and the dynamically adjusted display content is obtained as the output.

[0307] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0308] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0309] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0310] [Second embodiment]

[0311] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0312] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0313] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0314] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0315] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0316] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0317] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0318] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0319] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0320] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0321] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0322] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0323] The system of the present invention improves the accuracy and efficiency of credit evaluation by automating credit scoring of merchants. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0324] The server first connects to the database and retrieves information about affiliated stores, including store size, annual sales, industry, etc. The retrieved data cannot be used directly in a machine learning model, so preprocessing is required.

[0325] During preprocessing, the server first removes missing values ​​from the dataset and then converts categorical data, such as industry, into numerical data using a method called one-hot encoding, which prepares the data for processing by machine learning models.

[0326] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0327] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. The calculated credit score is provided to the user.

[0328] Specific examples

[0329] For example, consider the following new merchant information being entered:

[0330] Store size: 100

[0331] Annual sales: 5000000

[0332] Industry: Food and beverage

[0333] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0334] Users can view the resulting credit scores and use them to improve their credit rating and underwriting processes, resulting in faster and more consistent credit scoring than traditional manual processes.

[0335] The above is an embodiment of the present invention. This system significantly improves the efficiency and accuracy of merchant scoring, and increases the reliability of risk assessment and screening processes.

[0336] The processing flow will be explained below.

[0337] Step 1:

[0338] The server connects to the database and retrieves merchant information. To retrieve merchant information from the database, it executes an SQL query and converts the results into a data frame. The retrieved merchant information includes attributes such as store size, annual sales, and industry.

[0339] Step 2:

[0340] The server preprocesses the acquired merchant information by removing missing values ​​in the dataset and converting categorical data (e.g., industry) into numerical data using one-hot encoding, which converts the categorical data into a format that can be processed by machine learning models.

[0341] Step 3:

[0342] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly. The training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0343] Step 4:

[0344] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees to learn patterns from the training data, which prevents the model from overfitting and gives it the ability to make accurate predictions.

[0345] Step 5:

[0346] The server evaluates the trained model with test data. It inputs the test data into the model and compares the credit scores predicted by the model with the actual credit scores. From this comparison, it calculates the accuracy of the model and evaluates the performance of the model.

[0347] Step 6:

[0348] The user enters information about the new franchisee into the system, such as store size, annual sales, and industry. This information is used to calculate the new franchisee's credit score.

[0349] Step 7:

[0350] The server pre-processes the new merchant information entered, converting categorical data into numerical data, and then inputs the pre-processed data into a trained random forest model.

[0351] Step 8:

[0352] The server uses the trained model to calculate a credit score for the new merchant, which serves as a benchmark for assessing the merchant's credit risk.

[0353] Step 9:

[0354] The server provides the calculated credit score to the user, which provides the user with information that can be used to assess the risk of new merchants and improve the screening process.

[0355] Example 1

[0356] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0357] Conventional credit scoring systems require a lot of time and effort for data preprocessing, model training, and evaluation, which can result in insufficient accuracy and efficiency in the assessment. Furthermore, many manual processes make it difficult to achieve consistent scoring. Furthermore, delays in processing new merchant information can lead to lost business opportunities.

[0358] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0359] In this invention, the server includes means for acquiring company information from a database, means for preprocessing the acquired company information, means for dividing the preprocessed company information into training data and test data, means for training an inference model using the training data, means for evaluating the model using the test data, means for preprocessing new company information and calculating a reliability using the trained model, and means for providing the results to the user. This allows for efficient credit scoring, improves the consistency and accuracy of evaluations, and enables quick scoring of new franchisees.

[0360] A "database" is a collection of information stored in a format that allows multiple data to be managed and searched efficiently.

[0361] "Company information" refers to various data relating to a company, specifically including information such as store size, annual sales, and industry.

[0362] "Preprocessing" refers to a series of operations that transform data into a format suitable for machine learning models, including removing missing values ​​and converting categorical data into numerical data.

[0363] "Training data" refers to a dataset used to train a machine learning model and improve the model's predictive capabilities.

[0364] "Test data" is a dataset used to evaluate the performance of a machine learning model and is used to confirm the model's generalization ability.

[0365] An "inference model" is a model that uses machine learning algorithms to learn patterns from data and make predictions about unknown data.

[0366] "Trustworthiness" is a numerical indicator that indicates the trustworthiness of a new company, and is calculated using a machine learning model.

[0367] "User" means a person or organization that uses this system to obtain credit scoring results and use them for evaluation or screening.

[0368] The system of the present invention automates the credit scoring of companies, improving the accuracy and efficiency of the evaluation. This system can calculate credit scores using machine learning models based on company information stored in a database.

[0369] The server first connects to the database to retrieve company information. This information includes store size, annual sales, industry, etc. Specific database connection libraries used include JDBC and PDO. The retrieved data is not suitable for machine learning models as is, so preprocessing is required.

[0370] In the preprocessing stage, the server first removes missing values ​​from the dataset and converts categorical data such as industry into numerical data using the dropna method from the pandas library and OneHotEncoder from the sklearn library, converting the data into a format that can be processed by the machine learning model.

[0371] After preprocessing, the data is then split into training data and test data. The server uses the sklearn train_test_split method to split the data, with 80% of the data being used for training and 20% for testing.

[0372] Next, the server trains a machine learning model using a random forest classifier, which uses multiple decision trees to prevent overfitting and make more accurate predictions. After training, the server evaluates the accuracy of the model using test data. sklearn's accuracy_score is used in the evaluation process.

[0373] To calculate a new business's credit score, the user enters information about the new business into the device. Specific input items include:

[0374] Store size: 100

[0375] Annual sales: 5000000

[0376] Industry: Food and beverage

[0377] The server first preprocesses this information. For example, it converts "Industry: Food & Beverage" into numerical data. Then, it inputs the preprocessed data into a trained random forest model to predict a credit score. Users can then view this credit score and use it to improve their rating and screening process.

[0378] Prompt Sentence Examples

[0379] "Calculate the credit score for a new franchisee whose store size is 100, annual sales are 5,000,000, and the industry is food and beverage."

[0380] This concludes the description of an embodiment of the present invention. This system significantly improves the efficiency and accuracy of company scoring, increasing the reliability of risk assessment and screening processes. It also enables faster and more consistent credit scoring than traditional manual processes.

[0381] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0382] Step 1:

[0383] The server connects to the database and retrieves company information. Specifically, it uses a database connection library such as JDBC or PDO. It executes an SQL query to retrieve a dataset containing information such as store size, annual sales, and industry. The input for this step is the database, and the output is the retrieved dataset of company information.

[0384] Step 2:

[0385] The server preprocesses the acquired company information. First, it uses the dropna method in the pandas library to remove missing values ​​in the dataset. Then it uses sklearn's OneHotEncoder to convert categorical data such as industry to numerical data. The input to this step is the acquired company information dataset, and the output is the preprocessed data.

[0386] Step 3:

[0387] The server splits the preprocessed data into training data and test data. It uses sklearn's train_test_split method, with 80% of the data for training and 20% for testing. The input to this step is the preprocessed data, and the output is the training data and test data.

[0388] Step 4:

[0389] The server uses the training data to train a machine learning model using a random forest classifier. The classifier uses sklearn's RandomForestClassifier. To train the model, the training data is used as input and a trained model is output.

[0390] Step 5:

[0391] The server uses the trained model to evaluate the model on the test data. The evaluation uses sklearn's accuracy_score method to calculate the model's predictive accuracy. The inputs to this step are the trained model and the test data, and the output is the model's accuracy score.

[0392] Step 6:

[0393] The user inputs information about the new company from the terminal. This information includes store size, annual sales, and industry. The input for this step is the new company's information, and the output is the input data.

[0394] Step 7:

[0395] The server preprocesses the new business information and calculates the credit score using the trained model. After preprocessing, the new business information is input into the trained model to predict the credit score. Preprocessing also includes converting categorical data to numerical data. The input of this step is the preprocessed new business information, and the output is the credit score.

[0396] Step 8:

[0397] The server provides the calculated credit score to the user, which is used to improve the rating and underwriting process and ultimately to make effective business decisions. The input to this step is the credit score, and the output is the user's confirmation of the credit score.

[0398] (Application example 1)

[0399] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0400] Conventional credit scoring systems require time and effort to obtain and preprocess merchant information and calculate credit scores, making it difficult to perform efficient and accurate scoring. In particular, virtual store operators lack a means to instantly check credit scores, making it difficult to make quick operational decisions.

[0401] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0402] In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, and means for displaying the credit score for the virtual store via the cloud server using the generative AI model, thereby enabling the virtual store operator to check the credit score in real time and make quick decisions.

[0403] A "database" is a system for efficiently storing and managing data such as affiliated store information.

[0404] "Affiliated store information" is detailed data about affiliated stores, such as store size, annual sales, and industry.

[0405] "Preprocessing" is the process of converting data into a format that can be used by a machine learning model, and includes removing missing values ​​and converting categorical data into numerical data.

[0406] "Training data" is a dataset used to train a machine learning model.

[0407] "Test data" is a dataset used to evaluate a trained machine learning model.

[0408] A "machine learning model" is an algorithm that learns knowledge and patterns from data and makes predictions and classifications.

[0409] A "random forest classifier" is a type of machine learning model that combines multiple decision trees to improve prediction accuracy.

[0410] A "generative AI model" is an artificial intelligence algorithm that automatically generates a specific output (such as a prompt) from data.

[0411] A "cloud server" is a remote server system that shares resources over the Internet and performs data storage and computation.

[0412] A "credit score" is a numerical representation of a merchant's creditworthiness and is used in risk assessment and screening processes.

[0413] The system of the present invention automates credit scoring for virtual stores, improving the accuracy and efficiency of evaluation. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0414] Main processing steps

[0415] Retrieving data from a database

[0416] The server first connects to the database and retrieves information about the affiliated store, including store size, annual sales, industry, etc. The server receives the retrieved data in a format that requires preprocessing.

[0417] Data Preprocessing

[0418] The acquired data cannot be used as is for machine learning models, so preprocessing is required. In preprocessing, the server first removes missing values ​​from the dataset. It then converts categorical data, such as industry, into numerical data. This conversion is performed using a method called one-hot encoding. This puts the data in a format that can be processed by machine learning models.

[0419] Splitting the data and training the model

[0420] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and enable more accurate predictions. After training is complete, the server evaluates the model with the test data and calculates its accuracy.

[0421] Calculating and displaying credit scores

[0422] When the system calculates a new merchant's credit score, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.) through the smart glasses. The server preprocesses this new merchant information and inputs it into the trained model to calculate the credit score. The calculated credit score is displayed in real time on the smart glasses' display via the cloud server, allowing the user to check it.

[0423] Specific examples

[0424] For example, consider the following new merchant information being entered:

[0425] Store size: 100

[0426] Annual sales: 5000000

[0427] Industry: Food and beverage

[0428] The server first preprocesses this information. Specifically, it converts "Industry: Food & Beverage" into numerical data. The preprocessed data is then input into a trained random forest model. The model predicts a credit score based on these features. Users can view the resulting credit score and use it to improve their evaluation and screening process.

[0429] The system enables faster and more consistent credit scoring than traditional manual processes, and allows virtual store operators to view credit scores in real time and make necessary decisions quickly.

[0430] Prompt Sentence Examples

[0431] Store size: 100

[0432] Annual sales: 5000000

[0433] Industry: Food and beverage

[0434] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0435] Step 1:

[0436] The server connects to the database and retrieves merchant information, including store size, annual sales, and industry. The input is a database access request, and the output is the retrieved merchant information. This operation includes, for example, issuing an SQL query to retrieve the data.

[0437] Step 2:

[0438] The server preprocesses the acquired data. Specifically, it removes missing values ​​in the merchant information and converts categorical data to numerical data. The input is the raw merchant information, and the output is the preprocessed data. This conversion is performed using the Python Pandas library for data cleansing.

[0439] Step 3:

[0440] The server splits the preprocessed data into training data and test data. As input, it has the preprocessed data, and as output, it has training data and test data. This operation is performed using the train_test_split function of the Scikit-learn library.

[0441] Step 4:

[0442] The server uses the training data to train a machine learning model. For this, it uses a random forest classifier. The training data is the input, and the trained model is the output. For this operation, the model is trained using Scikit-learn's RandomForestClassifier.

[0443] Step 5:

[0444] The server evaluates the trained model with test data. The input is the test data, and the output is the model's evaluation result or accuracy. This operation uses Scikit-learn's evaluation function.

[0445] Step 6:

[0446] The user inputs new franchisee information through the smart glasses, for example, using voice recognition or touch gestures. The input includes new franchisee information such as store size, annual sales, and industry, but there is no output.

[0447] Step 7:

[0448] The server preprocesses the new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information, and the output is the credit score. Specifically, the server repeats the preprocessing in step 2 and then inputs the data into the trained model.

[0449] Step 8:

[0450] The server sends the calculated credit score to the smart glasses via the cloud server and displays it on the display. The credit score is the input, and the credit score displayed on the smart glasses is obtained as the output. This operation uses a communication protocol to send and receive data.

[0451] Specific examples

[0452] Input information: Store size 100, annual sales 5,000,000, industry Food and beverage

[0453] The server pre-processes this information and inputs it into a trained model.

[0454] The calculated credit score, for example "Credit score: 0.85", will be displayed on the smart glasses.

[0455] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0456] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. This system calculates credit scores using a machine learning model based on merchant information stored in a database, and uses the emotion engine to provide display content that corresponds to the user's emotions.

[0457] The server first connects to the database and retrieves member store information. The information retrieved from the database includes attributes such as store size, annual sales, and industry. The retrieved data cannot be used as is in a machine learning model, so preprocessing is required.

[0458] During preprocessing, the server first removes missing values ​​in the dataset and converts categorical data (such as industry) into numerical data using one-hot encoding, which prepares the data for processing by machine learning models.

[0459] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0460] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate a credit score. The calculated credit score is displayed by the emotion engine according to the user's emotions.

[0461] The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the content displayed will be more detailed and reassuring.

[0462] Specific examples

[0463] For example, consider the following new merchant information being entered:

[0464] Store size: 100

[0465] Annual sales: 5000000

[0466] Industry: Food and beverage

[0467] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0468] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0469] For example, if the user looks worried, we might show them a detailed explanation of how the credit score is calculated or some additional information to reassure them, while if the user looks interested, we might offer more advanced analysis and profit predictions.

[0470] In this way, the system of the present invention significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user sentiment, resulting in a more reliable risk assessment and screening process and an improved user experience.

[0471] The processing flow will be explained below.

[0472] Step 1:

[0473] The server connects to the database and retrieves merchant information, including attributes such as store size, annual sales, and industry. Data is retrieved from the database using an SQL query, and the results are converted into a data frame.

[0474] Step 2:

[0475] The server preprocesses the acquired merchant information. Specifically, it removes missing values ​​in the dataset and converts categorical data to numerical data using the following procedure. By converting categorical data (e.g., industry) to numerical data using one-hot encoding, it creates a format that can be processed by the machine learning model.

[0476] Step 3:

[0477] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly, and the training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0478] Step 4:

[0479] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees, each of which learns independently to prevent overfitting and enable more accurate predictions.

[0480] Step 5:

[0481] The server evaluates the trained model with test data. It inputs the test data into the model and compares the model's predictions with the actual credit scores. Based on the comparison results, it calculates the model's accuracy (e.g., accuracy rate) and evaluates its performance.

[0482] Step 6:

[0483] The user enters information about the new franchisee into the system, including the store size, annual sales, industry, etc. This information becomes the basic data for calculating the new franchisee's credit score.

[0484] Step 7:

[0485] The server preprocesses the input new merchant information, converting categorical data into numerical data, and inputs the preprocessed data into a trained random forest model. The preprocessing method is the same as in Step 2.

[0486] Step 8:

[0487] The server uses the trained model to calculate a credit score for the new merchant, and the model's predicted credit score serves as a benchmark for assessing the new merchant's credit risk.

[0488] Step 9:

[0489] The device's built-in camera and microphone analyze the user's behavior in real time and recognize emotions from the user's facial expressions and voice. Based on the analyzed emotional data, the emotion engine recognizes how the user is receiving information (e.g., relief, anxiety, doubt).

[0490] Step 10:

[0491] The server dynamically adjusts the calculated trust score according to the user's emotions and changes the presentation method. For example, if the user looks anxious, it will display detailed explanations and additional information that provides reassurance. On the other hand, if the user shows interest, it will be configured to provide more advanced analysis results and predictive data.

[0492] Step 11:

[0493] The server displays the final credit score and related information to the user in an appropriate format, providing information that can be used to improve risk assessment and screening processes for new merchants.

[0494] Example 2

[0495] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0496] Conventional merchant credit scoring systems require a lot of time and effort for data preprocessing and machine learning model training. Furthermore, they only provide uniform information to users and are unable to respond flexibly to user emotions, resulting in a lack of user experience. This leads to issues such as a decrease in the accuracy and reliability of credit scores and a decline in the efficiency of risk assessment and screening processes.

[0497] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0498] In this invention, the server includes means for acquiring merchant information from a database, means for preprocessing the acquired merchant information, means for dividing the preprocessed merchant information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new merchant information and calculating a credit score using the trained model, means for recognizing user emotions, and means for dynamically adjusting information display content based on user emotions. This significantly improves the efficiency and accuracy of merchant scoring and enables flexible responses according to user emotions.

[0499] A "database" is an information system that systematically organizes and stores information, allowing it to be searched and retrieved efficiently.

[0500] "Affiliated store information" is information about affiliated stores, and is data including attributes such as store size, annual sales, and industry type.

[0501] "Preprocessing" is the process of improving data quality, removing unnecessary information, and converting data into an appropriate format for use in data analysis and machine learning models.

[0502] A "machine learning model" is an algorithm or mathematical model that learns from data and makes decisions such as predictions and classifications.

[0503] A "random forest classifier" is a machine learning algorithm that combines and learns from multiple decision trees to improve the accuracy of classification and prediction.

[0504] The "emotion engine" is a system that analyzes data such as the user's facial expressions and voice to recognize their emotional state.

[0505] "Means for recognizing user emotions" refers to technology that uses a camera or microphone to capture the user's facial expressions and voice and analyze their emotions.

[0506] "Means for dynamically adjusting the content of displayed information" is a mechanism for changing the content and format of the information presented in real time based on the user's emotions.

[0507] A "credit score" is a numerical representation of the reliability and risk of a merchant, and is an indicator used to make decisions about transactions.

[0508] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. A specific implementation of this system is described below.

[0509] The server first connects to a database (for example, an SQL database) and retrieves member store information. This member store information includes attributes such as store size, annual sales, and industry. The information retrieved from the database cannot be used in its raw form for machine learning models, so preprocessing is required. The Python Pandas library is used for data preprocessing, and one-hot encoding is applied to convert categorical data into numerical values.

[0510] After preprocessing, the data is then split into training data and test data. This splitting is done using the Scikit-Learn library. The server uses the training data to train a machine learning model using a random forest classifier. The trained model is then evaluated on the test data to calculate its accuracy.

[0511] When calculating the credit score of a new merchant, the user enters information about the new merchant (e.g., store size, annual sales, industry, etc.) into a dedicated form on the terminal. This information is sent to the server via an HTTP request. The server similarly preprocesses the received information about the new merchant, inputs it into the trained model, and calculates the credit score.

[0512] The calculated credit score is displayed by an emotion engine according to the user's emotions. The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. Emotion recognition is performed using, for example, Microsoft Azure Emotion API. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the display content can be changed to something more detailed and reassuring.

[0513] Specific examples

[0514] For example, consider the following new merchant information being entered:

[0515] Store size: 100

[0516] Annual sales: 5000000

[0517] Industry: Food and beverage

[0518] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data, and then inputs the preprocessed data into a trained random forest model. The model predicts a credit score based on these features.

[0519] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0520] Prompt Sentence Examples

[0521] "Calculate a credit score for a new franchisee based on store size of 100, annual sales of 5,000,000, and industry of food and beverage. Also, explain in natural language the processing steps of the system that adjusts the way information is presented based on the user's emotions."

[0522] Using these prompts, the generative AI explains the overall system flow and the specific processing steps at each step. This system significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user emotions. This increases the reliability of the risk assessment and screening process and improves the user experience.

[0523] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0524] Step 1:

[0525] The server first connects to the database and retrieves member store information. This database connection retrieves attribute data such as "store size," "annual sales," and "industry type." The input is an SQL query, and the output is a dataset of member store information.

[0526] Step 2:

[0527] The server preprocesses the acquired merchant information. This preprocessing includes removing missing values ​​and converting categorical data to numeric data. Specifically, it uses Python's Pandas library to manipulate data frames. The input is the acquired merchant information dataset, and the output is the preprocessed dataset.

[0528] Step 3:

[0529] The server splits the preprocessed dataset into training data and test data. Here, we use the Scikit-Learn library to split the data. The input is the preprocessed dataset, and the output is the training data and test data.

[0530] Step 4:

[0531] The server uses the training data to train a machine learning model, using a random forest classifier, where the input is the training data and the output is the trained machine learning model.

[0532] Step 5:

[0533] The server uses the trained model to evaluate the test data. This evaluation uses various metrics to calculate the accuracy of the model. The inputs are the test data and the trained model, and the output is the accuracy score of the model.

[0534] Step 6:

[0535] The user enters the information of the new affiliated store into a dedicated form on the terminal. The entered information is sent to the server via an HTTP request. The input is the information of the new affiliated store, and the output is the HTTP request sent to the server.

[0536] Step 7:

[0537] The server preprocesses the received new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information and the trained model, and the output is the credit score.

[0538] Step 8:

[0539] The camera and microphone installed on the device analyze the user's facial expressions and voice to recognize emotions. This analysis is performed using an emotion recognition library (e.g., Microsoft Azure Emotion API). The input is the user's facial expressions and voice data, and the output is the recognized emotion data.

[0540] Step 9:

[0541] The server dynamically adjusts the displayed credit score based on the user's emotional data obtained from the emotion engine. Specifically, if the user looks anxious, the server takes action such as providing more information. The input is the credit score and the recognized emotional data, and the output is the adjusted display content.

[0542] (Application example 2)

[0543] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0544] Conventional merchant credit scoring systems take time to process and evaluate data, and lack flexibility in providing users with credit scores. Furthermore, they provide information uniformly without considering users' feelings, which results in a lack of improvement in the user experience. Furthermore, electronic payment services require support that allows users to use them safely and without anxiety. These issues need to be resolved.

[0545] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into learning data and test data, means for training a machine learning model using the learning data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, means for recognizing emotions from the user's facial expressions and voice using a camera and microphone installed in the terminal, and means for dynamically adjusting the presentation method based on the recognized emotion data. This makes it possible to not only score member stores' credit but also flexibly provide information according to the user's emotions, thereby realizing an environment in which users can use electronic payment services with peace of mind.

[0546] A "database" is a structured data storage system for efficiently managing, searching, and retrieving information.

[0547] "Affiliated store information" refers to detailed business data related to affiliated stores, such as store size, annual sales, and industry.

[0548] "Preprocessing" is the process of preparing data before it can be used in a machine learning model, such as removing missing values ​​or converting categorical data to numerical data.

[0549] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications.

[0550] A "credit score" is a numerical value calculated to assess the creditworthiness of a merchant and is used in risk assessment and screening processes.

[0551] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to operate the system.

[0552] A "camera" is an optical device for capturing images of a user's facial expressions and movements.

[0553] A "microphone" is a device for recording the user's voice.

[0554] "Emotion recognition" is a technology that analyzes a user's facial expressions and voice to identify the emotions the user is feeling.

[0555] "Dynamic adjustment" is the process of changing settings and display content in real time based on the situation.

[0556] System Overview

[0557] The system of the present invention automates credit scoring for merchants and dynamically adjusts information presentation methods by combining user emotion recognition. This system is composed of a server, terminals, a database, a machine learning model, an emotion recognition engine, etc.

[0558] Program processing overview

[0559] The server first connects to the database and retrieves information about the member stores. This information includes store size, annual sales, industry, etc. Next, the retrieved member store information is preprocessed. This preprocessing involves removing missing values ​​and converting categorical data into numerical data. This conversion uses techniques such as one-hot encoding.

[0560] After preprocessing, the data is split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The trained model is then evaluated on the test data to calculate its accuracy.

[0561] When calculating the credit score of a new merchant, the user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate the credit score.

[0562] Meanwhile, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize their emotions. The emotion recognition engine uses software such as OpenCV and EmotionRecognizer. This recognized emotion data is acquired by the server and used to dynamically adjust the way the credit score is presented.

[0563] Specific examples

[0564] Consider the case where the following data is entered as information for a new merchant:

[0565] Store size: 100

[0566] Annual sales: 5000000

[0567] Industry: Food and beverage

[0568] The server preprocesses this information, converting the industry into numerical data using one-hot encoding, and then feeds the preprocessed data into a trained random forest model to predict the credit score.

[0569] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how they feel about the information they receive. For example, if the user looks anxious, the system will display detailed explanations and additional information about how the credit score is provided, reassuring the user. On the other hand, if the user looks interested, the system will provide more advanced analysis results and profit predictions.

[0570] Prompt Sentence Examples

[0571] plaintext

[0572] Information about new franchisee "Restaurant A":

[0573] Store size: 100

[0574] Annual sales: 5000000

[0575] Industry: Food and beverage

[0576] Calculate a credit score and provide detailed information based on the user's "anxiety" emotion.

[0577] The system of the present invention can significantly improve the accuracy of the information provided and the user experience. By not only providing credit scores for member stores but also providing information flexibly according to the user's emotions, it provides an environment in which users can use electronic payment services with peace of mind.

[0578] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0579] Step 1:

[0580] Obtain affiliated store information from the database. The server connects to the database and obtains affiliated store information such as store size, annual sales, and industry. Specifically, it uses an SQL query to obtain all the information in one go. This requires database connection information as input, and the output is a dataset containing affiliated store information.

[0581] Step 2:

[0582] Preprocess the merchant information. The server removes missing values ​​from the acquired data and converts categorical data into numerical data. Specifically, categorical data such as industry type is converted using one-hot encoding. This input is the output data of step 1, and a preprocessed dataset is obtained as the output.

[0583] Step 3:

[0584] Split into training data and test data. The server splits the preprocessed data into training data and test data. Specifically, it randomly shuffles the dataset and splits it into two parts at a fixed ratio. This input is the output data of step 2, and the output is training data and test data.

[0585] Step 4:

[0586] Train the machine learning model. The server trains a random forest classifier using the training data. Specifically, it fits the model using the features and labels of the training data. This input is the training data from step 3, and the output is a trained model.

[0587] Step 5:

[0588] Evaluate the model with test data. The server evaluates the trained model with test data and calculates its accuracy. Specifically, it uses the test data to make predictions and compares them with the actual values. The input is the test data from step 3, and the output is the accuracy of the model.

[0589] Step 6:

[0590] Preprocessing new merchant information and calculating a credit score. The user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. Specifically, preprocessing is performed in the same way as in step 2, and the model is used in the same way as in step 4. The input is the new merchant information from the user, and the output is a credit score.

[0591] Step 7:

[0592] Recognizes user emotions. The device uses the on-board camera and microphone to recognize emotions from the user's facial expressions and voice. Specifically, OpenCV and EmotionRecognizer are used to perform real-time analysis. The input is video and audio data acquired from the device, and the output is recognized emotion data.

[0593] Step 8:

[0594] The server dynamically adjusts the presentation method based on the emotional data. Based on the recognized emotional data, the server dynamically adjusts the presentation method of the credit score. For example, if the user feels anxious, it will display additional detailed explanations or information that provides reassurance. Specifically, it executes logic that changes the display content based on the emotional data. The inputs are the credit score from step 6 and the emotional data from step 7, and the dynamically adjusted display content is obtained as the output.

[0595] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0596] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0597] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[0598] [Third embodiment]

[0599] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0600] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0601] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0602] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[0603] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0604] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0605] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0606] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0607] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0608] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0609] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0610] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[0611] The system of the present invention improves the accuracy and efficiency of credit evaluation by automating credit scoring of merchants. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0612] The server first connects to the database and retrieves information about affiliated stores, including store size, annual sales, industry, etc. The retrieved data cannot be used directly in a machine learning model, so preprocessing is required.

[0613] During preprocessing, the server first removes missing values ​​from the dataset and then converts categorical data, such as industry, into numerical data using a method called one-hot encoding, which prepares the data for processing by machine learning models.

[0614] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0615] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. The calculated credit score is provided to the user.

[0616] Specific examples

[0617] For example, consider the following new merchant information being entered:

[0618] Store size: 100

[0619] Annual sales: 5000000

[0620] Industry: Food and beverage

[0621] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0622] Users can view the resulting credit scores and use them to improve their credit rating and underwriting processes, resulting in faster and more consistent credit scoring than traditional manual processes.

[0623] The above is an embodiment of the present invention. This system significantly improves the efficiency and accuracy of merchant scoring, and increases the reliability of risk assessment and screening processes.

[0624] The processing flow will be explained below.

[0625] Step 1:

[0626] The server connects to the database and retrieves merchant information. To retrieve merchant information from the database, it executes an SQL query and converts the results into a data frame. The retrieved merchant information includes attributes such as store size, annual sales, and industry.

[0627] Step 2:

[0628] The server preprocesses the acquired merchant information by removing missing values ​​in the dataset and converting categorical data (e.g., industry) into numerical data using one-hot encoding, which converts the categorical data into a format that can be processed by machine learning models.

[0629] Step 3:

[0630] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly. The training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0631] Step 4:

[0632] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees to learn patterns from the training data, which prevents the model from overfitting and gives it the ability to make accurate predictions.

[0633] Step 5:

[0634] The server evaluates the trained model with test data. It inputs the test data into the model and compares the credit scores predicted by the model with the actual credit scores. From this comparison, it calculates the accuracy of the model and evaluates the performance of the model.

[0635] Step 6:

[0636] The user enters information about the new franchisee into the system, such as store size, annual sales, and industry. This information is used to calculate the new franchisee's credit score.

[0637] Step 7:

[0638] The server pre-processes the new merchant information entered, converting categorical data into numerical data, and then inputs the pre-processed data into a trained random forest model.

[0639] Step 8:

[0640] The server uses the trained model to calculate a credit score for the new merchant, which serves as a benchmark for assessing the merchant's credit risk.

[0641] Step 9:

[0642] The server provides the calculated credit score to the user, which provides the user with information that can be used to assess the risk of new merchants and improve the screening process.

[0643] Example 1

[0644] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0645] Conventional credit scoring systems require a lot of time and effort for data preprocessing, model training, and evaluation, which can result in insufficient accuracy and efficiency in the assessment. Furthermore, many manual processes make it difficult to achieve consistent scoring. Furthermore, delays in processing new merchant information can lead to lost business opportunities.

[0646] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0647] In this invention, the server includes means for acquiring company information from a database, means for preprocessing the acquired company information, means for dividing the preprocessed company information into training data and test data, means for training an inference model using the training data, means for evaluating the model using the test data, means for preprocessing new company information and calculating a reliability using the trained model, and means for providing the results to the user. This allows for efficient credit scoring, improves the consistency and accuracy of evaluations, and enables quick scoring of new franchisees.

[0648] A "database" is a collection of information stored in a format that allows multiple data to be managed and searched efficiently.

[0649] "Company information" refers to various data relating to a company, specifically including information such as store size, annual sales, and industry.

[0650] "Preprocessing" refers to a series of operations that transform data into a format suitable for machine learning models, including removing missing values ​​and converting categorical data into numerical data.

[0651] "Training data" refers to a dataset used to train a machine learning model and improve the model's predictive capabilities.

[0652] "Test data" is a dataset used to evaluate the performance of a machine learning model and is used to confirm the model's generalization ability.

[0653] An "inference model" is a model that uses machine learning algorithms to learn patterns from data and make predictions about unknown data.

[0654] "Trustworthiness" is a numerical indicator that indicates the trustworthiness of a new company, and is calculated using a machine learning model.

[0655] "User" means a person or organization that uses this system to obtain credit scoring results and use them for evaluation or screening.

[0656] The system of the present invention automates the credit scoring of companies, improving the accuracy and efficiency of the evaluation. This system can calculate credit scores using machine learning models based on company information stored in a database.

[0657] The server first connects to the database to retrieve company information. This information includes store size, annual sales, industry, etc. Specific database connection libraries used include JDBC and PDO. The retrieved data is not suitable for machine learning models as is, so preprocessing is required.

[0658] In the preprocessing stage, the server first removes missing values ​​from the dataset and converts categorical data such as industry into numerical data using the dropna method from the pandas library and OneHotEncoder from the sklearn library, converting the data into a format that can be processed by the machine learning model.

[0659] After preprocessing, the data is then split into training data and test data. The server uses the sklearn train_test_split method to split the data, with 80% of the data being used for training and 20% for testing.

[0660] Next, the server trains a machine learning model using a random forest classifier, which uses multiple decision trees to prevent overfitting and make more accurate predictions. After training, the server evaluates the accuracy of the model using test data. sklearn's accuracy_score is used in the evaluation process.

[0661] To calculate a new business's credit score, the user enters information about the new business into the device. Specific input items include:

[0662] Store size: 100

[0663] Annual sales: 5000000

[0664] Industry: Food and beverage

[0665] The server first preprocesses this information. For example, it converts "Industry: Food & Beverage" into numerical data. Then, it inputs the preprocessed data into a trained random forest model to predict a credit score. Users can then view this credit score and use it to improve their rating and screening process.

[0666] Prompt Sentence Examples

[0667] "Calculate the credit score for a new franchisee whose store size is 100, annual sales are 5,000,000, and the industry is food and beverage."

[0668] This concludes the description of an embodiment of the present invention. This system significantly improves the efficiency and accuracy of company scoring, increasing the reliability of risk assessment and screening processes. It also enables faster and more consistent credit scoring than traditional manual processes.

[0669] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0670] Step 1:

[0671] The server connects to the database and retrieves company information. Specifically, it uses a database connection library such as JDBC or PDO. It executes an SQL query to retrieve a dataset containing information such as store size, annual sales, and industry. The input for this step is the database, and the output is the retrieved dataset of company information.

[0672] Step 2:

[0673] The server preprocesses the acquired company information. First, it uses the dropna method in the pandas library to remove missing values ​​in the dataset. Then it uses sklearn's OneHotEncoder to convert categorical data such as industry to numerical data. The input to this step is the acquired company information dataset, and the output is the preprocessed data.

[0674] Step 3:

[0675] The server splits the preprocessed data into training data and test data. It uses sklearn's train_test_split method, with 80% of the data for training and 20% for testing. The input to this step is the preprocessed data, and the output is the training data and test data.

[0676] Step 4:

[0677] The server uses the training data to train a machine learning model using a random forest classifier. The classifier uses sklearn's RandomForestClassifier. To train the model, the training data is used as input and a trained model is output.

[0678] Step 5:

[0679] The server uses the trained model to evaluate the model on the test data. The evaluation uses sklearn's accuracy_score method to calculate the model's predictive accuracy. The inputs to this step are the trained model and the test data, and the output is the model's accuracy score.

[0680] Step 6:

[0681] The user inputs information about the new company from the terminal. This information includes store size, annual sales, and industry. The input for this step is the new company's information, and the output is the input data.

[0682] Step 7:

[0683] The server preprocesses the new business information and calculates the credit score using the trained model. After preprocessing, the new business information is input into the trained model to predict the credit score. Preprocessing also includes converting categorical data to numerical data. The input of this step is the preprocessed new business information, and the output is the credit score.

[0684] Step 8:

[0685] The server provides the calculated credit score to the user, which is used to improve the rating and underwriting process and ultimately to make effective business decisions. The input to this step is the credit score, and the output is the user's confirmation of the credit score.

[0686] (Application example 1)

[0687] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0688] Conventional credit scoring systems require time and effort to obtain and preprocess merchant information and calculate credit scores, making it difficult to perform efficient and accurate scoring. In particular, virtual store operators lack a means to instantly check credit scores, making it difficult to make quick operational decisions.

[0689] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0690] In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, and means for displaying the credit score for the virtual store via the cloud server using the generative AI model, thereby enabling the virtual store operator to check the credit score in real time and make quick decisions.

[0691] A "database" is a system for efficiently storing and managing data such as affiliated store information.

[0692] "Affiliated store information" is detailed data about affiliated stores, such as store size, annual sales, and industry.

[0693] "Preprocessing" is the process of converting data into a format that can be used by a machine learning model, and includes removing missing values ​​and converting categorical data into numerical data.

[0694] "Training data" is a dataset used to train a machine learning model.

[0695] "Test data" is a dataset used to evaluate a trained machine learning model.

[0696] A "machine learning model" is an algorithm that learns knowledge and patterns from data and makes predictions and classifications.

[0697] A "random forest classifier" is a type of machine learning model that combines multiple decision trees to improve prediction accuracy.

[0698] A "generative AI model" is an artificial intelligence algorithm that automatically generates a specific output (such as a prompt) from data.

[0699] A "cloud server" is a remote server system that shares resources over the Internet and performs data storage and computation.

[0700] A "credit score" is a numerical representation of a merchant's creditworthiness and is used in risk assessment and screening processes.

[0701] The system of the present invention automates credit scoring for virtual stores, improving the accuracy and efficiency of evaluation. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0702] Main processing steps

[0703] Retrieving data from a database

[0704] The server first connects to the database and retrieves information about the affiliated store, including store size, annual sales, industry, etc. The server receives the retrieved data in a format that requires preprocessing.

[0705] Data Preprocessing

[0706] The acquired data cannot be used as is for machine learning models, so preprocessing is required. In preprocessing, the server first removes missing values ​​from the dataset. It then converts categorical data, such as industry, into numerical data. This conversion is performed using a method called one-hot encoding. This puts the data in a format that can be processed by machine learning models.

[0707] Splitting the data and training the model

[0708] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and enable more accurate predictions. After training is complete, the server evaluates the model with the test data and calculates its accuracy.

[0709] Calculating and displaying credit scores

[0710] When the system calculates a new merchant's credit score, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.) through the smart glasses. The server preprocesses this new merchant information and inputs it into the trained model to calculate the credit score. The calculated credit score is displayed in real time on the smart glasses' display via the cloud server, allowing the user to check it.

[0711] Specific examples

[0712] For example, consider the following new merchant information being entered:

[0713] Store size: 100

[0714] Annual sales: 5000000

[0715] Industry: Food and beverage

[0716] The server first preprocesses this information. Specifically, it converts "Industry: Food & Beverage" into numerical data. The preprocessed data is then input into a trained random forest model. The model predicts a credit score based on these features. Users can view the resulting credit score and use it to improve their evaluation and screening process.

[0717] The system enables faster and more consistent credit scoring than traditional manual processes, and allows virtual store operators to view credit scores in real time and make necessary decisions quickly.

[0718] Prompt Sentence Examples

[0719] Store size: 100

[0720] Annual sales: 5000000

[0721] Industry: Food and beverage

[0722] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0723] Step 1:

[0724] The server connects to the database and retrieves merchant information, including store size, annual sales, and industry. The input is a database access request, and the output is the retrieved merchant information. This operation includes, for example, issuing an SQL query to retrieve the data.

[0725] Step 2:

[0726] The server preprocesses the acquired data. Specifically, it removes missing values ​​in the merchant information and converts categorical data to numerical data. The input is the raw merchant information, and the output is the preprocessed data. This conversion is performed using the Python Pandas library for data cleansing.

[0727] Step 3:

[0728] The server splits the preprocessed data into training data and test data. As input, it has the preprocessed data, and as output, it has training data and test data. This operation is performed using the train_test_split function of the Scikit-learn library.

[0729] Step 4:

[0730] The server uses the training data to train a machine learning model. For this, it uses a random forest classifier. The training data is the input, and the trained model is the output. For this operation, the model is trained using Scikit-learn's RandomForestClassifier.

[0731] Step 5:

[0732] The server evaluates the trained model with test data. The input is the test data, and the output is the model's evaluation result or accuracy. This operation uses Scikit-learn's evaluation function.

[0733] Step 6:

[0734] The user inputs new franchisee information through the smart glasses, for example, using voice recognition or touch gestures. The input includes new franchisee information such as store size, annual sales, and industry, but there is no output.

[0735] Step 7:

[0736] The server preprocesses the new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information, and the output is the credit score. Specifically, the server repeats the preprocessing in step 2 and then inputs the data into the trained model.

[0737] Step 8:

[0738] The server sends the calculated credit score to the smart glasses via the cloud server and displays it on the display. The credit score is the input, and the credit score displayed on the smart glasses is obtained as the output. This operation uses a communication protocol to send and receive data.

[0739] Specific examples

[0740] Input information: Store size 100, annual sales 5,000,000, industry Food and beverage

[0741] The server pre-processes this information and inputs it into a trained model.

[0742] The calculated credit score, for example "Credit score: 0.85", will be displayed on the smart glasses.

[0743] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0744] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. This system calculates credit scores using a machine learning model based on merchant information stored in a database, and uses the emotion engine to provide display content that corresponds to the user's emotions.

[0745] The server first connects to the database and retrieves member store information. The information retrieved from the database includes attributes such as store size, annual sales, and industry. The retrieved data cannot be used as is in a machine learning model, so preprocessing is required.

[0746] During preprocessing, the server first removes missing values ​​in the dataset and converts categorical data (such as industry) into numerical data using one-hot encoding, which prepares the data for processing by machine learning models.

[0747] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0748] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate a credit score. The calculated credit score is displayed by the emotion engine according to the user's emotions.

[0749] The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the content displayed will be more detailed and reassuring.

[0750] Specific examples

[0751] For example, consider the following new merchant information being entered:

[0752] Store size: 100

[0753] Annual sales: 5000000

[0754] Industry: Food and beverage

[0755] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0756] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0757] For example, if the user looks worried, we might show them a detailed explanation of how the credit score is calculated or some additional information to reassure them, while if the user looks interested, we might offer more advanced analysis and profit predictions.

[0758] In this way, the system of the present invention significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user sentiment, resulting in a more reliable risk assessment and screening process and an improved user experience.

[0759] The processing flow will be explained below.

[0760] Step 1:

[0761] The server connects to the database and retrieves merchant information, including attributes such as store size, annual sales, and industry. Data is retrieved from the database using an SQL query, and the results are converted into a data frame.

[0762] Step 2:

[0763] The server preprocesses the acquired merchant information. Specifically, it removes missing values ​​in the dataset and converts categorical data to numerical data using the following procedure. By converting categorical data (e.g., industry) to numerical data using one-hot encoding, it creates a format that can be processed by the machine learning model.

[0764] Step 3:

[0765] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly, and the training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0766] Step 4:

[0767] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees, each of which learns independently to prevent overfitting and enable more accurate predictions.

[0768] Step 5:

[0769] The server evaluates the trained model with test data. It inputs the test data into the model and compares the model's predictions with the actual credit scores. Based on the comparison results, it calculates the model's accuracy (e.g., accuracy rate) and evaluates its performance.

[0770] Step 6:

[0771] The user enters information about the new franchisee into the system, including the store size, annual sales, industry, etc. This information becomes the basic data for calculating the new franchisee's credit score.

[0772] Step 7:

[0773] The server preprocesses the input new merchant information, converting categorical data into numerical data, and inputs the preprocessed data into a trained random forest model. The preprocessing method is the same as in Step 2.

[0774] Step 8:

[0775] The server uses the trained model to calculate a credit score for the new merchant, and the model's predicted credit score serves as a benchmark for assessing the new merchant's credit risk.

[0776] Step 9:

[0777] The device's built-in camera and microphone analyze the user's behavior in real time and recognize emotions from the user's facial expressions and voice. Based on the analyzed emotional data, the emotion engine recognizes how the user is receiving information (e.g., relief, anxiety, doubt).

[0778] Step 10:

[0779] The server dynamically adjusts the calculated trust score according to the user's emotions and changes the presentation method. For example, if the user looks anxious, it will display detailed explanations and additional information that provides reassurance. On the other hand, if the user shows interest, it will be configured to provide more advanced analysis results and predictive data.

[0780] Step 11:

[0781] The server displays the final credit score and related information to the user in an appropriate format, providing information that can be used to improve risk assessment and screening processes for new merchants.

[0782] Example 2

[0783] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0784] Conventional merchant credit scoring systems require a lot of time and effort for data preprocessing and machine learning model training. Furthermore, they only provide uniform information to users and are unable to respond flexibly to user emotions, resulting in a lack of user experience. This leads to issues such as a decrease in the accuracy and reliability of credit scores and a decline in the efficiency of risk assessment and screening processes.

[0785] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0786] In this invention, the server includes means for acquiring merchant information from a database, means for preprocessing the acquired merchant information, means for dividing the preprocessed merchant information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new merchant information and calculating a credit score using the trained model, means for recognizing user emotions, and means for dynamically adjusting information display content based on user emotions. This significantly improves the efficiency and accuracy of merchant scoring and enables flexible responses according to user emotions.

[0787] A "database" is an information system that systematically organizes and stores information, allowing it to be searched and retrieved efficiently.

[0788] "Affiliated store information" is information about affiliated stores, and is data including attributes such as store size, annual sales, and industry type.

[0789] "Preprocessing" is the process of improving data quality, removing unnecessary information, and converting data into an appropriate format for use in data analysis and machine learning models.

[0790] A "machine learning model" is an algorithm or mathematical model that learns from data and makes decisions such as predictions and classifications.

[0791] A "random forest classifier" is a machine learning algorithm that combines and learns from multiple decision trees to improve the accuracy of classification and prediction.

[0792] The "emotion engine" is a system that analyzes data such as the user's facial expressions and voice to recognize their emotional state.

[0793] "Means for recognizing user emotions" refers to technology that uses a camera or microphone to capture the user's facial expressions and voice and analyze their emotions.

[0794] "Means for dynamically adjusting the content of displayed information" is a mechanism for changing the content and format of the information presented in real time based on the user's emotions.

[0795] A "credit score" is a numerical representation of the reliability and risk of a merchant, and is an indicator used to make decisions about transactions.

[0796] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. A specific implementation of this system is described below.

[0797] The server first connects to a database (for example, an SQL database) and retrieves member store information. This member store information includes attributes such as store size, annual sales, and industry. The information retrieved from the database cannot be used in its raw form for machine learning models, so preprocessing is required. The Python Pandas library is used for data preprocessing, and one-hot encoding is applied to convert categorical data into numerical values.

[0798] After preprocessing, the data is then split into training data and test data. This splitting is done using the Scikit-Learn library. The server uses the training data to train a machine learning model using a random forest classifier. The trained model is then evaluated on the test data to calculate its accuracy.

[0799] When calculating the credit score of a new merchant, the user enters information about the new merchant (e.g., store size, annual sales, industry, etc.) into a dedicated form on the terminal. This information is sent to the server via an HTTP request. The server similarly preprocesses the received information about the new merchant, inputs it into the trained model, and calculates the credit score.

[0800] The calculated credit score is displayed by an emotion engine according to the user's emotions. The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. Emotion recognition is performed using, for example, Microsoft Azure Emotion API. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the display content can be changed to something more detailed and reassuring.

[0801] Specific examples

[0802] For example, consider the following new merchant information being entered:

[0803] Store size: 100

[0804] Annual sales: 5000000

[0805] Industry: Food and beverage

[0806] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data, and then inputs the preprocessed data into a trained random forest model. The model predicts a credit score based on these features.

[0807] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[0808] Prompt Sentence Examples

[0809] "Calculate a credit score for a new franchisee based on store size of 100, annual sales of 5,000,000, and industry of food and beverage. Also, explain in natural language the processing steps of the system that adjusts the way information is presented based on the user's emotions."

[0810] Using these prompts, the generative AI explains the overall system flow and the specific processing steps at each step. This system significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user emotions. This increases the reliability of the risk assessment and screening process and improves the user experience.

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

[0812] Step 1:

[0813] The server first connects to the database and retrieves member store information. This database connection retrieves attribute data such as "store size," "annual sales," and "industry type." The input is an SQL query, and the output is a dataset of member store information.

[0814] Step 2:

[0815] The server preprocesses the acquired merchant information. This preprocessing includes removing missing values ​​and converting categorical data to numeric data. Specifically, it uses Python's Pandas library to manipulate data frames. The input is the acquired merchant information dataset, and the output is the preprocessed dataset.

[0816] Step 3:

[0817] The server splits the preprocessed dataset into training data and test data. Here, we use the Scikit-Learn library to split the data. The input is the preprocessed dataset, and the output is the training data and test data.

[0818] Step 4:

[0819] The server uses the training data to train a machine learning model, using a random forest classifier, where the input is the training data and the output is the trained machine learning model.

[0820] Step 5:

[0821] The server uses the trained model to evaluate the test data. This evaluation uses various metrics to calculate the accuracy of the model. The inputs are the test data and the trained model, and the output is the accuracy score of the model.

[0822] Step 6:

[0823] The user enters the information of the new affiliated store into a dedicated form on the terminal. The entered information is sent to the server via an HTTP request. The input is the information of the new affiliated store, and the output is the HTTP request sent to the server.

[0824] Step 7:

[0825] The server preprocesses the received new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information and the trained model, and the output is the credit score.

[0826] Step 8:

[0827] The camera and microphone installed on the device analyze the user's facial expressions and voice to recognize emotions. This analysis is performed using an emotion recognition library (e.g., Microsoft Azure Emotion API). The input is the user's facial expressions and voice data, and the output is the recognized emotion data.

[0828] Step 9:

[0829] The server dynamically adjusts the displayed credit score based on the user's emotional data obtained from the emotion engine. Specifically, if the user looks anxious, the server takes action such as providing more information. The input is the credit score and the recognized emotional data, and the output is the adjusted display content.

[0830] (Application example 2)

[0831] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[0832] Conventional merchant credit scoring systems take time to process and evaluate data, and lack flexibility in providing users with credit scores. Furthermore, they provide information uniformly without considering users' feelings, which results in a lack of improvement in the user experience. Furthermore, electronic payment services require support that allows users to use them safely and without anxiety. These issues need to be resolved.

[0833] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into learning data and test data, means for training a machine learning model using the learning data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, means for recognizing emotions from the user's facial expressions and voice using a camera and microphone installed in the terminal, and means for dynamically adjusting the presentation method based on the recognized emotion data. This makes it possible to not only score member stores' credit but also flexibly provide information according to the user's emotions, thereby realizing an environment in which users can use electronic payment services with peace of mind.

[0834] A "database" is a structured data storage system for efficiently managing, searching, and retrieving information.

[0835] "Affiliated store information" refers to detailed business data related to affiliated stores, such as store size, annual sales, and industry.

[0836] "Preprocessing" is the process of preparing data before it can be used in a machine learning model, such as removing missing values ​​or converting categorical data to numerical data.

[0837] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications.

[0838] A "credit score" is a numerical value calculated to assess the creditworthiness of a merchant and is used in risk assessment and screening processes.

[0839] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to operate the system.

[0840] A "camera" is an optical device for capturing images of a user's facial expressions and movements.

[0841] A "microphone" is a device for recording the user's voice.

[0842] "Emotion recognition" is a technology that analyzes a user's facial expressions and voice to identify the emotions the user is feeling.

[0843] "Dynamic adjustment" is the process of changing settings and display content in real time based on the situation.

[0844] System Overview

[0845] The system of the present invention automates credit scoring for merchants and dynamically adjusts information presentation methods by combining user emotion recognition. This system is composed of a server, terminals, a database, a machine learning model, an emotion recognition engine, etc.

[0846] Program processing overview

[0847] The server first connects to the database and retrieves information about the member stores. This information includes store size, annual sales, industry, etc. Next, the retrieved member store information is preprocessed. This preprocessing involves removing missing values ​​and converting categorical data into numerical data. This conversion uses techniques such as one-hot encoding.

[0848] After preprocessing, the data is split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The trained model is then evaluated on the test data to calculate its accuracy.

[0849] When calculating the credit score of a new merchant, the user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate the credit score.

[0850] Meanwhile, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize their emotions. The emotion recognition engine uses software such as OpenCV and EmotionRecognizer. This recognized emotion data is acquired by the server and used to dynamically adjust the way the credit score is presented.

[0851] Specific examples

[0852] Consider the case where the following data is entered as information for a new merchant:

[0853] Store size: 100

[0854] Annual sales: 5000000

[0855] Industry: Food and beverage

[0856] The server preprocesses this information, converting the industry into numerical data using one-hot encoding, and then feeds the preprocessed data into a trained random forest model to predict the credit score.

[0857] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how they feel about the information they receive. For example, if the user looks anxious, the system will display detailed explanations and additional information about how the credit score is provided, reassuring the user. On the other hand, if the user looks interested, the system will provide more advanced analysis results and profit predictions.

[0858] Prompt Sentence Examples

[0859] plaintext

[0860] Information about new franchisee "Restaurant A":

[0861] Store size: 100

[0862] Annual sales: 5000000

[0863] Industry: Food and beverage

[0864] Calculate a credit score and provide detailed information based on the user's "anxiety" emotion.

[0865] The system of the present invention can significantly improve the accuracy of the information provided and the user experience. By not only providing credit scores for member stores but also providing information flexibly according to the user's emotions, it provides an environment in which users can use electronic payment services with peace of mind.

[0866] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0867] Step 1:

[0868] Obtain affiliated store information from the database. The server connects to the database and obtains affiliated store information such as store size, annual sales, and industry. Specifically, it uses an SQL query to obtain all the information in one go. This requires database connection information as input, and the output is a dataset containing affiliated store information.

[0869] Step 2:

[0870] Preprocess the merchant information. The server removes missing values ​​from the acquired data and converts categorical data into numerical data. Specifically, categorical data such as industry type is converted using one-hot encoding. This input is the output data of step 1, and a preprocessed dataset is obtained as the output.

[0871] Step 3:

[0872] Split into training data and test data. The server splits the preprocessed data into training data and test data. Specifically, it randomly shuffles the dataset and splits it into two parts at a fixed ratio. This input is the output data of step 2, and the output is training data and test data.

[0873] Step 4:

[0874] Train the machine learning model. The server trains a random forest classifier using the training data. Specifically, it fits the model using the features and labels of the training data. This input is the training data from step 3, and the output is a trained model.

[0875] Step 5:

[0876] Evaluate the model with test data. The server evaluates the trained model with test data and calculates its accuracy. Specifically, it uses the test data to make predictions and compares them with the actual values. The input is the test data from step 3, and the output is the accuracy of the model.

[0877] Step 6:

[0878] Preprocessing new merchant information and calculating a credit score. The user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. Specifically, preprocessing is performed in the same way as in step 2, and the model is used in the same way as in step 4. The input is the new merchant information from the user, and the output is a credit score.

[0879] Step 7:

[0880] Recognizes user emotions. The device uses the on-board camera and microphone to recognize emotions from the user's facial expressions and voice. Specifically, OpenCV and EmotionRecognizer are used to perform real-time analysis. The input is video and audio data acquired from the device, and the output is recognized emotion data.

[0881] Step 8:

[0882] The server dynamically adjusts the presentation method based on the emotional data. Based on the recognized emotional data, the server dynamically adjusts the presentation method of the credit score. For example, if the user feels anxious, it will display additional detailed explanations or information that provides reassurance. Specifically, it executes logic that changes the display content based on the emotional data. The inputs are the credit score from step 6 and the emotional data from step 7, and the dynamically adjusted display content is obtained as the output.

[0883] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0884] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0885] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[0886] [Fourth embodiment]

[0887] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0888] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0889] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0890] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[0891] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0892] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0893] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0894] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[0895] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0896] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0897] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0898] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0899] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0900] The system of the present invention improves the accuracy and efficiency of credit evaluation by automating credit scoring of merchants. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0901] The server first connects to the database and retrieves information about affiliated stores, including store size, annual sales, industry, etc. The retrieved data cannot be used directly in a machine learning model, so preprocessing is required.

[0902] During preprocessing, the server first removes missing values ​​from the dataset and then converts categorical data, such as industry, into numerical data using a method called one-hot encoding, which prepares the data for processing by machine learning models.

[0903] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[0904] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. The calculated credit score is provided to the user.

[0905] Specific examples

[0906] For example, consider the following new merchant information being entered:

[0907] Store size: 100

[0908] Annual sales: 5000000

[0909] Industry: Food and beverage

[0910] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[0911] Users can view the resulting credit scores and use them to improve their credit rating and underwriting processes, resulting in faster and more consistent credit scoring than traditional manual processes.

[0912] The above is an embodiment of the present invention. This system significantly improves the efficiency and accuracy of merchant scoring, and increases the reliability of risk assessment and screening processes.

[0913] The processing flow will be explained below.

[0914] Step 1:

[0915] The server connects to the database and retrieves merchant information. To retrieve merchant information from the database, it executes an SQL query and converts the results into a data frame. The retrieved merchant information includes attributes such as store size, annual sales, and industry.

[0916] Step 2:

[0917] The server preprocesses the acquired merchant information by removing missing values ​​in the dataset and converting categorical data (e.g., industry) into numerical data using one-hot encoding, which converts the categorical data into a format that can be processed by machine learning models.

[0918] Step 3:

[0919] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly. The training data is used to train the machine learning model, and the test data is used to evaluate the model.

[0920] Step 4:

[0921] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees to learn patterns from the training data, which prevents the model from overfitting and gives it the ability to make accurate predictions.

[0922] Step 5:

[0923] The server evaluates the trained model with test data. It inputs the test data into the model and compares the credit scores predicted by the model with the actual credit scores. From this comparison, it calculates the accuracy of the model and evaluates the performance of the model.

[0924] Step 6:

[0925] The user enters information about the new franchisee into the system, such as store size, annual sales, and industry. This information is used to calculate the new franchisee's credit score.

[0926] Step 7:

[0927] The server pre-processes the new merchant information entered, converting categorical data into numerical data, and then inputs the pre-processed data into a trained random forest model.

[0928] Step 8:

[0929] The server uses the trained model to calculate a credit score for the new merchant, which serves as a benchmark for assessing the merchant's credit risk.

[0930] Step 9:

[0931] The server provides the calculated credit score to the user, which provides the user with information that can be used to assess the risk of new merchants and improve the screening process.

[0932] Example 1

[0933] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0934] Conventional credit scoring systems require a lot of time and effort for data preprocessing, model training, and evaluation, which can result in insufficient accuracy and efficiency in the assessment. Furthermore, many manual processes make it difficult to achieve consistent scoring. Furthermore, delays in processing new merchant information can lead to lost business opportunities.

[0935] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0936] In this invention, the server includes means for acquiring company information from a database, means for preprocessing the acquired company information, means for dividing the preprocessed company information into training data and test data, means for training an inference model using the training data, means for evaluating the model using the test data, means for preprocessing new company information and calculating a reliability using the trained model, and means for providing the results to the user. This allows for efficient credit scoring, improves the consistency and accuracy of evaluations, and enables quick scoring of new franchisees.

[0937] A "database" is a collection of information stored in a format that allows multiple data to be managed and searched efficiently.

[0938] "Company information" refers to various data relating to a company, specifically including information such as store size, annual sales, and industry.

[0939] "Preprocessing" refers to a series of operations that transform data into a format suitable for machine learning models, including removing missing values ​​and converting categorical data into numerical data.

[0940] "Training data" refers to a dataset used to train a machine learning model and improve the model's predictive capabilities.

[0941] "Test data" is a dataset used to evaluate the performance of a machine learning model and is used to confirm the model's generalization ability.

[0942] An "inference model" is a model that uses machine learning algorithms to learn patterns from data and make predictions about unknown data.

[0943] "Trustworthiness" is a numerical indicator that indicates the trustworthiness of a new company, and is calculated using a machine learning model.

[0944] "User" means a person or organization that uses this system to obtain credit scoring results and use them for evaluation or screening.

[0945] The system of the present invention automates the credit scoring of companies, improving the accuracy and efficiency of the evaluation. This system can calculate credit scores using machine learning models based on company information stored in a database.

[0946] The server first connects to the database to retrieve company information. This information includes store size, annual sales, industry, etc. Specific database connection libraries used include JDBC and PDO. The retrieved data is not suitable for machine learning models as is, so preprocessing is required.

[0947] In the preprocessing stage, the server first removes missing values ​​from the dataset and converts categorical data such as industry into numerical data using the dropna method from the pandas library and OneHotEncoder from the sklearn library, converting the data into a format that can be processed by the machine learning model.

[0948] After preprocessing, the data is then split into training data and test data. The server uses the sklearn train_test_split method to split the data, with 80% of the data being used for training and 20% for testing.

[0949] Next, the server trains a machine learning model using a random forest classifier, which uses multiple decision trees to prevent overfitting and make more accurate predictions. After training, the server evaluates the accuracy of the model using test data. sklearn's accuracy_score is used in the evaluation process.

[0950] To calculate a new business's credit score, the user enters information about the new business into the device. Specific input items include:

[0951] Store size: 100

[0952] Annual sales: 5000000

[0953] Industry: Food and beverage

[0954] The server first preprocesses this information. For example, it converts "Industry: Food & Beverage" into numerical data. Then, it inputs the preprocessed data into a trained random forest model to predict a credit score. Users can then view this credit score and use it to improve their rating and screening process.

[0955] Prompt Sentence Examples

[0956] "Calculate the credit score for a new franchisee whose store size is 100, annual sales are 5,000,000, and the industry is food and beverage."

[0957] This concludes the description of an embodiment of the present invention. This system significantly improves the efficiency and accuracy of company scoring, increasing the reliability of risk assessment and screening processes. It also enables faster and more consistent credit scoring than traditional manual processes.

[0958] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0959] Step 1:

[0960] The server connects to the database and retrieves company information. Specifically, it uses a database connection library such as JDBC or PDO. It executes an SQL query to retrieve a dataset containing information such as store size, annual sales, and industry. The input for this step is the database, and the output is the retrieved dataset of company information.

[0961] Step 2:

[0962] The server preprocesses the acquired company information. First, it uses the dropna method in the pandas library to remove missing values ​​in the dataset. Then it uses sklearn's OneHotEncoder to convert categorical data such as industry to numerical data. The input to this step is the acquired company information dataset, and the output is the preprocessed data.

[0963] Step 3:

[0964] The server splits the preprocessed data into training data and test data. It uses sklearn's train_test_split method, with 80% of the data for training and 20% for testing. The input to this step is the preprocessed data, and the output is the training data and test data.

[0965] Step 4:

[0966] The server uses the training data to train a machine learning model using a random forest classifier. The classifier uses sklearn's RandomForestClassifier. To train the model, the training data is used as input and a trained model is output.

[0967] Step 5:

[0968] The server uses the trained model to evaluate the model on the test data. The evaluation uses sklearn's accuracy_score method to calculate the model's predictive accuracy. The inputs to this step are the trained model and the test data, and the output is the model's accuracy score.

[0969] Step 6:

[0970] The user inputs information about the new company from the terminal. This information includes store size, annual sales, and industry. The input for this step is the new company's information, and the output is the input data.

[0971] Step 7:

[0972] The server preprocesses the new business information and calculates the credit score using the trained model. After preprocessing, the new business information is input into the trained model to predict the credit score. Preprocessing also includes converting categorical data to numerical data. The input of this step is the preprocessed new business information, and the output is the credit score.

[0973] Step 8:

[0974] The server provides the calculated credit score to the user, which is used to improve the rating and underwriting process and ultimately to make effective business decisions. The input to this step is the credit score, and the output is the user's confirmation of the credit score.

[0975] (Application example 1)

[0976] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[0977] Conventional credit scoring systems require time and effort to obtain and preprocess merchant information and calculate credit scores, making it difficult to perform efficient and accurate scoring. In particular, virtual store operators lack a means to instantly check credit scores, making it difficult to make quick operational decisions.

[0978] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0979] In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, and means for displaying the credit score for the virtual store via the cloud server using the generative AI model, thereby enabling the virtual store operator to check the credit score in real time and make quick decisions.

[0980] A "database" is a system for efficiently storing and managing data such as affiliated store information.

[0981] "Affiliated store information" is detailed data about affiliated stores, such as store size, annual sales, and industry.

[0982] "Preprocessing" is the process of converting data into a format that can be used by a machine learning model, and includes removing missing values ​​and converting categorical data into numerical data.

[0983] "Training data" is a dataset used to train a machine learning model.

[0984] "Test data" is a dataset used to evaluate a trained machine learning model.

[0985] A "machine learning model" is an algorithm that learns knowledge and patterns from data and makes predictions and classifications.

[0986] A "random forest classifier" is a type of machine learning model that combines multiple decision trees to improve prediction accuracy.

[0987] A "generative AI model" is an artificial intelligence algorithm that automatically generates a specific output (such as a prompt) from data.

[0988] A "cloud server" is a remote server system that shares resources over the Internet and performs data storage and computation.

[0989] A "credit score" is a numerical representation of a merchant's creditworthiness and is used in risk assessment and screening processes.

[0990] The system of the present invention automates credit scoring for virtual stores, improving the accuracy and efficiency of evaluation. This system calculates credit scores using machine learning models based on merchant information stored in a database.

[0991] Main processing steps

[0992] Retrieving data from a database

[0993] The server first connects to the database and retrieves information about the affiliated store, including store size, annual sales, industry, etc. The server receives the retrieved data in a format that requires preprocessing.

[0994] Data Preprocessing

[0995] The acquired data cannot be used as is for machine learning models, so preprocessing is required. In preprocessing, the server first removes missing values ​​from the dataset. It then converts categorical data, such as industry, into numerical data. This conversion is performed using a method called one-hot encoding. This puts the data in a format that can be processed by machine learning models.

[0996] Splitting the data and training the model

[0997] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and enable more accurate predictions. After training is complete, the server evaluates the model with the test data and calculates its accuracy.

[0998] Calculating and displaying credit scores

[0999] When the system calculates a new merchant's credit score, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.) through the smart glasses. The server preprocesses this new merchant information and inputs it into the trained model to calculate the credit score. The calculated credit score is displayed in real time on the smart glasses' display via the cloud server, allowing the user to check it.

[1000] Specific examples

[1001] For example, consider the following new merchant information being entered:

[1002] Store size: 100

[1003] Annual sales: 5000000

[1004] Industry: Food and beverage

[1005] The server first preprocesses this information. Specifically, it converts "Industry: Food & Beverage" into numerical data. The preprocessed data is then input into a trained random forest model. The model predicts a credit score based on these features. Users can view the resulting credit score and use it to improve their evaluation and screening process.

[1006] The system enables faster and more consistent credit scoring than traditional manual processes, and allows virtual store operators to view credit scores in real time and make necessary decisions quickly.

[1007] Prompt Sentence Examples

[1008] Store size: 100

[1009] Annual sales: 5000000

[1010] Industry: Food and beverage

[1011] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1012] Step 1:

[1013] The server connects to the database and retrieves merchant information, including store size, annual sales, and industry. The input is a database access request, and the output is the retrieved merchant information. This operation includes, for example, issuing an SQL query to retrieve the data.

[1014] Step 2:

[1015] The server preprocesses the acquired data. Specifically, it removes missing values ​​in the merchant information and converts categorical data to numerical data. The input is the raw merchant information, and the output is the preprocessed data. This conversion is performed using the Python Pandas library for data cleansing.

[1016] Step 3:

[1017] The server splits the preprocessed data into training data and test data. As input, it has the preprocessed data, and as output, it has training data and test data. This operation is performed using the train_test_split function of the Scikit-learn library.

[1018] Step 4:

[1019] The server uses the training data to train a machine learning model. For this, it uses a random forest classifier. The training data is the input, and the trained model is the output. For this operation, the model is trained using Scikit-learn's RandomForestClassifier.

[1020] Step 5:

[1021] The server evaluates the trained model with test data. The input is the test data, and the output is the model's evaluation result or accuracy. This operation uses Scikit-learn's evaluation function.

[1022] Step 6:

[1023] The user inputs new franchisee information through the smart glasses, for example, using voice recognition or touch gestures. The input includes new franchisee information such as store size, annual sales, and industry, but there is no output.

[1024] Step 7:

[1025] The server preprocesses the new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information, and the output is the credit score. Specifically, the server repeats the preprocessing in step 2 and then inputs the data into the trained model.

[1026] Step 8:

[1027] The server sends the calculated credit score to the smart glasses via the cloud server and displays it on the display. The credit score is the input, and the credit score displayed on the smart glasses is obtained as the output. This operation uses a communication protocol to send and receive data.

[1028] Specific examples

[1029] Input information: Store size 100, annual sales 5,000,000, industry Food and beverage

[1030] The server pre-processes this information and inputs it into a trained model.

[1031] The calculated credit score, for example "Credit score: 0.85", will be displayed on the smart glasses.

[1032] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1033] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. This system calculates credit scores using a machine learning model based on merchant information stored in a database, and uses the emotion engine to provide display content that corresponds to the user's emotions.

[1034] The server first connects to the database and retrieves member store information. The information retrieved from the database includes attributes such as store size, annual sales, and industry. The retrieved data cannot be used as is in a machine learning model, so preprocessing is required.

[1035] During preprocessing, the server first removes missing values ​​in the dataset and converts categorical data (such as industry) into numerical data using one-hot encoding, which prepares the data for processing by machine learning models.

[1036] After preprocessing, the data is then split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The random forest classifier uses multiple decision trees to prevent overfitting and make more accurate predictions. The server evaluates the trained model with the test data and calculates its accuracy.

[1037] When the system calculates a credit score for a new merchant, the user inputs the merchant's information (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate a credit score. The calculated credit score is displayed by the emotion engine according to the user's emotions.

[1038] The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the content displayed will be more detailed and reassuring.

[1039] Specific examples

[1040] For example, consider the following new merchant information being entered:

[1041] Store size: 100

[1042] Annual sales: 5000000

[1043] Industry: Food and beverage

[1044] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data. The preprocessed data is then fed into a trained random forest model, which predicts a credit score based on these features.

[1045] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[1046] For example, if the user looks worried, we might show them a detailed explanation of how the credit score is calculated or some additional information to reassure them, while if the user looks interested, we might offer more advanced analysis and profit predictions.

[1047] In this way, the system of the present invention significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user sentiment, resulting in a more reliable risk assessment and screening process and an improved user experience.

[1048] The processing flow will be explained below.

[1049] Step 1:

[1050] The server connects to the database and retrieves merchant information, including attributes such as store size, annual sales, and industry. Data is retrieved from the database using an SQL query, and the results are converted into a data frame.

[1051] Step 2:

[1052] The server preprocesses the acquired merchant information. Specifically, it removes missing values ​​in the dataset and converts categorical data to numerical data using the following procedure. By converting categorical data (e.g., industry) to numerical data using one-hot encoding, it creates a format that can be processed by the machine learning model.

[1053] Step 3:

[1054] The server splits the preprocessed data into training data and test data. 80% of the data is used as training data, and the remaining 20% ​​is used as test data. This split is done randomly, and the training data is used to train the machine learning model, and the test data is used to evaluate the model.

[1055] Step 4:

[1056] The server uses the training data to train a random forest classifier, an ensemble learning algorithm that uses multiple decision trees, each of which learns independently to prevent overfitting and enable more accurate predictions.

[1057] Step 5:

[1058] The server evaluates the trained model with test data. It inputs the test data into the model and compares the model's predictions with the actual credit scores. Based on the comparison results, it calculates the model's accuracy (e.g., accuracy rate) and evaluates its performance.

[1059] Step 6:

[1060] The user enters information about the new franchisee into the system, including the store size, annual sales, industry, etc. This information becomes the basic data for calculating the new franchisee's credit score.

[1061] Step 7:

[1062] The server preprocesses the input new merchant information, converting categorical data into numerical data, and inputs the preprocessed data into a trained random forest model. The preprocessing method is the same as in Step 2.

[1063] Step 8:

[1064] The server uses the trained model to calculate a credit score for the new merchant, and the model's predicted credit score serves as a benchmark for assessing the new merchant's credit risk.

[1065] Step 9:

[1066] The device's built-in camera and microphone analyze the user's behavior in real time and recognize emotions from the user's facial expressions and voice. Based on the analyzed emotional data, the emotion engine recognizes how the user is receiving information (e.g., relief, anxiety, doubt).

[1067] Step 10:

[1068] The server dynamically adjusts the calculated trust score according to the user's emotions and changes the presentation method. For example, if the user looks anxious, it will display detailed explanations and additional information that provides reassurance. On the other hand, if the user shows interest, it will be configured to provide more advanced analysis results and predictive data.

[1069] Step 11:

[1070] The server displays the final credit score and related information to the user in an appropriate format, providing information that can be used to improve risk assessment and screening processes for new merchants.

[1071] Example 2

[1072] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1073] Conventional merchant credit scoring systems require a lot of time and effort for data preprocessing and machine learning model training. Furthermore, they only provide uniform information to users and are unable to respond flexibly to user emotions, resulting in a lack of user experience. This leads to issues such as a decrease in the accuracy and reliability of credit scores and a decline in the efficiency of risk assessment and screening processes.

[1074] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1075] In this invention, the server includes means for acquiring merchant information from a database, means for preprocessing the acquired merchant information, means for dividing the preprocessed merchant information into training data and test data, means for training a machine learning model using the training data, means for evaluating the model with the test data, means for preprocessing new merchant information and calculating a credit score using the trained model, means for recognizing user emotions, and means for dynamically adjusting information display content based on user emotions. This significantly improves the efficiency and accuracy of merchant scoring and enables flexible responses according to user emotions.

[1076] A "database" is an information system that systematically organizes and stores information, allowing it to be searched and retrieved efficiently.

[1077] "Affiliated store information" is information about affiliated stores, and is data including attributes such as store size, annual sales, and industry type.

[1078] "Preprocessing" is the process of improving data quality, removing unnecessary information, and converting data into an appropriate format for use in data analysis and machine learning models.

[1079] A "machine learning model" is an algorithm or mathematical model that learns from data and makes decisions such as predictions and classifications.

[1080] A "random forest classifier" is a machine learning algorithm that combines and learns from multiple decision trees to improve the accuracy of classification and prediction.

[1081] The "emotion engine" is a system that analyzes data such as the user's facial expressions and voice to recognize their emotional state.

[1082] "Means for recognizing user emotions" refers to technology that uses a camera or microphone to capture the user's facial expressions and voice and analyze their emotions.

[1083] "Means for dynamically adjusting the content of displayed information" is a mechanism for changing the content and format of the information presented in real time based on the user's emotions.

[1084] A "credit score" is a numerical representation of the reliability and risk of a merchant, and is an indicator used to make decisions about transactions.

[1085] The system of the present invention automates credit scoring for merchants and combines it with an emotion engine that recognizes user emotions to improve the accuracy and efficiency of evaluation. A specific implementation of this system is described below.

[1086] The server first connects to a database (for example, an SQL database) and retrieves member store information. This member store information includes attributes such as store size, annual sales, and industry. The information retrieved from the database cannot be used in its raw form for machine learning models, so preprocessing is required. The Python Pandas library is used for data preprocessing, and one-hot encoding is applied to convert categorical data into numerical values.

[1087] After preprocessing, the data is then split into training data and test data. This splitting is done using the Scikit-Learn library. The server uses the training data to train a machine learning model using a random forest classifier. The trained model is then evaluated on the test data to calculate its accuracy.

[1088] When calculating the credit score of a new merchant, the user enters information about the new merchant (e.g., store size, annual sales, industry, etc.) into a dedicated form on the terminal. This information is sent to the server via an HTTP request. The server similarly preprocesses the received information about the new merchant, inputs it into the trained model, and calculates the credit score.

[1089] The calculated credit score is displayed by an emotion engine according to the user's emotions. The emotion engine uses the device's built-in camera and microphone to recognize emotions from the user's facial expressions and voice. Emotion recognition is performed using, for example, Microsoft Azure Emotion API. This recognized emotion data is used to dynamically adjust the way the credit score is presented. For example, if the user looks anxious, the display content can be changed to something more detailed and reassuring.

[1090] Specific examples

[1091] For example, consider the following new merchant information being entered:

[1092] Store size: 100

[1093] Annual sales: 5000000

[1094] Industry: Food and beverage

[1095] The server first preprocesses this information, specifically converting "Industry: Food & Beverage" into numerical data, and then inputs the preprocessed data into a trained random forest model. The model predicts a credit score based on these features.

[1096] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how the user is receiving the information (e.g., relief, anxiety, doubt). Once the trust score is calculated, the server dynamically adjusts the information content presented to the user based on feedback from the emotion engine.

[1097] Prompt Sentence Examples

[1098] "Calculate a credit score for a new franchisee based on store size of 100, annual sales of 5,000,000, and industry of food and beverage. Also, explain in natural language the processing steps of the system that adjusts the way information is presented based on the user's emotions."

[1099] Using these prompts, the generative AI explains the overall system flow and the specific processing steps at each step. This system significantly improves the efficiency and accuracy of merchant scoring while also enabling flexible responses based on user emotions. This increases the reliability of the risk assessment and screening process and improves the user experience.

[1100] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1101] Step 1:

[1102] The server first connects to the database and retrieves member store information. This database connection retrieves attribute data such as "store size," "annual sales," and "industry type." The input is an SQL query, and the output is a dataset of member store information.

[1103] Step 2:

[1104] The server preprocesses the acquired merchant information. This preprocessing includes removing missing values ​​and converting categorical data to numeric data. Specifically, it uses Python's Pandas library to manipulate data frames. The input is the acquired merchant information dataset, and the output is the preprocessed dataset.

[1105] Step 3:

[1106] The server splits the preprocessed dataset into training data and test data. Here, we use the Scikit-Learn library to split the data. The input is the preprocessed dataset, and the output is the training data and test data.

[1107] Step 4:

[1108] The server uses the training data to train a machine learning model, using a random forest classifier, where the input is the training data and the output is the trained machine learning model.

[1109] Step 5:

[1110] The server uses the trained model to evaluate the test data. This evaluation uses various metrics to calculate the accuracy of the model. The inputs are the test data and the trained model, and the output is the accuracy score of the model.

[1111] Step 6:

[1112] The user enters the information of the new affiliated store into a dedicated form on the terminal. The entered information is sent to the server via an HTTP request. The input is the information of the new affiliated store, and the output is the HTTP request sent to the server.

[1113] Step 7:

[1114] The server preprocesses the received new merchant information and inputs it into the trained model to calculate the credit score. The input is the new merchant information and the trained model, and the output is the credit score.

[1115] Step 8:

[1116] The camera and microphone installed on the device analyze the user's facial expressions and voice to recognize emotions. This analysis is performed using an emotion recognition library (e.g., Microsoft Azure Emotion API). The input is the user's facial expressions and voice data, and the output is the recognized emotion data.

[1117] Step 9:

[1118] The server dynamically adjusts the displayed credit score based on the user's emotional data obtained from the emotion engine. Specifically, if the user looks anxious, the server takes action such as providing more information. The input is the credit score and the recognized emotional data, and the output is the adjusted display content.

[1119] (Application example 2)

[1120] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1121] Conventional merchant credit scoring systems take time to process and evaluate data, and lack flexibility in providing users with credit scores. Furthermore, they provide information uniformly without considering users' feelings, which results in a lack of improvement in the user experience. Furthermore, electronic payment services require support that allows users to use them safely and without anxiety. These issues need to be resolved.

[1122] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for acquiring member store information from a database, means for preprocessing the member store information, means for dividing the preprocessed member store information into learning data and test data, means for training a machine learning model using the learning data, means for evaluating the model with the test data, means for preprocessing new member store information and calculating a credit score using the trained model, means for recognizing emotions from the user's facial expressions and voice using a camera and microphone installed in the terminal, and means for dynamically adjusting the presentation method based on the recognized emotion data. This makes it possible to not only score member stores' credit but also flexibly provide information according to the user's emotions, thereby realizing an environment in which users can use electronic payment services with peace of mind.

[1123] A "database" is a structured data storage system for efficiently managing, searching, and retrieving information.

[1124] "Affiliated store information" refers to detailed business data related to affiliated stores, such as store size, annual sales, and industry.

[1125] "Preprocessing" is the process of preparing data before it can be used in a machine learning model, such as removing missing values ​​or converting categorical data to numerical data.

[1126] A "machine learning model" is a collection of algorithms that learn patterns from data and make predictions or classifications.

[1127] A "credit score" is a numerical value calculated to assess the creditworthiness of a merchant and is used in risk assessment and screening processes.

[1128] "Terminal" refers to a device, such as a computer or smartphone, that a user uses to operate the system.

[1129] A "camera" is an optical device for capturing images of a user's facial expressions and movements.

[1130] A "microphone" is a device for recording the user's voice.

[1131] "Emotion recognition" is a technology that analyzes a user's facial expressions and voice to identify the emotions the user is feeling.

[1132] "Dynamic adjustment" is the process of changing settings and display content in real time based on the situation.

[1133] System Overview

[1134] The system of the present invention automates credit scoring for merchants and dynamically adjusts information presentation methods by combining user emotion recognition. This system is composed of a server, terminals, a database, a machine learning model, an emotion recognition engine, etc.

[1135] Program processing overview

[1136] The server first connects to the database and retrieves information about the member stores. This information includes store size, annual sales, industry, etc. Next, the retrieved member store information is preprocessed. This preprocessing involves removing missing values ​​and converting categorical data into numerical data. This conversion uses techniques such as one-hot encoding.

[1137] After preprocessing, the data is split into training data and test data. The server uses the training data to train a machine learning model. A random forest classifier is used for this training. The trained model is then evaluated on the test data to calculate its accuracy.

[1138] When calculating the credit score of a new merchant, the user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into a trained model to calculate the credit score.

[1139] Meanwhile, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize their emotions. The emotion recognition engine uses software such as OpenCV and EmotionRecognizer. This recognized emotion data is acquired by the server and used to dynamically adjust the way the credit score is presented.

[1140] Specific examples

[1141] Consider the case where the following data is entered as information for a new merchant:

[1142] Store size: 100

[1143] Annual sales: 5000000

[1144] Industry: Food and beverage

[1145] The server preprocesses this information, converting the industry into numerical data using one-hot encoding, and then feeds the preprocessed data into a trained random forest model to predict the credit score.

[1146] In parallel, the device's built-in camera and microphone analyze the user's facial expressions and voice to recognize how they feel about the information they receive. For example, if the user looks anxious, the system will display detailed explanations and additional information about how the credit score is provided, reassuring the user. On the other hand, if the user looks interested, the system will provide more advanced analysis results and profit predictions.

[1147] Prompt Sentence Examples

[1148] plaintext

[1149] Information about new franchisee "Restaurant A":

[1150] Store size: 100

[1151] Annual sales: 5000000

[1152] Industry: Food and beverage

[1153] Calculate a credit score and provide detailed information based on the user's "anxiety" emotion.

[1154] The system of the present invention can significantly improve the accuracy of the information provided and the user experience. By not only providing credit scores for member stores but also providing information flexibly according to the user's emotions, it provides an environment in which users can use electronic payment services with peace of mind.

[1155] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1156] Step 1:

[1157] Obtain affiliated store information from the database. The server connects to the database and obtains affiliated store information such as store size, annual sales, and industry. Specifically, it uses an SQL query to obtain all the information in one go. This requires database connection information as input, and the output is a dataset containing affiliated store information.

[1158] Step 2:

[1159] Preprocess the merchant information. The server removes missing values ​​from the acquired data and converts categorical data into numerical data. Specifically, categorical data such as industry type is converted using one-hot encoding. This input is the output data of step 1, and a preprocessed dataset is obtained as the output.

[1160] Step 3:

[1161] Split into training data and test data. The server splits the preprocessed data into training data and test data. Specifically, it randomly shuffles the dataset and splits it into two parts at a fixed ratio. This input is the output data of step 2, and the output is training data and test data.

[1162] Step 4:

[1163] Train the machine learning model. The server trains a random forest classifier using the training data. Specifically, it fits the model using the features and labels of the training data. This input is the training data from step 3, and the output is a trained model.

[1164] Step 5:

[1165] Evaluate the model with test data. The server evaluates the trained model with test data and calculates its accuracy. Specifically, it uses the test data to make predictions and compares them with the actual values. The input is the test data from step 3, and the output is the accuracy of the model.

[1166] Step 6:

[1167] Preprocessing new merchant information and calculating a credit score. The user inputs information about the new merchant (e.g., store size, annual sales, industry, etc.). The server preprocesses this information and inputs it into the trained model to calculate a credit score. Specifically, preprocessing is performed in the same way as in step 2, and the model is used in the same way as in step 4. The input is the new merchant information from the user, and the output is a credit score.

[1168] Step 7:

[1169] Recognizes user emotions. The device uses the on-board camera and microphone to recognize emotions from the user's facial expressions and voice. Specifically, OpenCV and EmotionRecognizer are used to perform real-time analysis. The input is video and audio data acquired from the device, and the output is recognized emotion data.

[1170] Step 8:

[1171] The server dynamically adjusts the presentation method based on the emotional data. Based on the recognized emotional data, the server dynamically adjusts the presentation method of the credit score. For example, if the user feels anxious, it will display additional detailed explanations or information that provides reassurance. Specifically, it executes logic that changes the display content based on the emotional data. The inputs are the credit score from step 6 and the emotional data from step 7, and the dynamically adjusted display content is obtained as the output.

[1172] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[1173] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1174] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1175] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1176] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1177] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1178] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1179] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1180] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1181] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1182] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1183] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1184] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1185] 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.

[1186] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1187] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1188] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1189] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1190] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1191] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1192] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1193] The following is further disclosed regarding the above embodiment.

[1194] (Claim 1)

[1195] A means for obtaining merchant information from the database;

[1196] A means for pre-processing the acquired member store information;

[1197] A means for dividing the preprocessed merchant information into training data and test data;

[1198] means for training a machine learning model using the training data;

[1199] A means of evaluating the model on test data;

[1200] a means for preprocessing new merchant information and calculating a credit score using the trained model;

[1201] A system including:

[1202] (Claim 2)

[1203] The system according to claim 1, wherein the system deletes missing values ​​from the acquired member store information and converts categorical data into numerical data.

[1204] (Claim 3)

[1205] 10. The system of claim 1, wherein the training data is trained using a random forest classifier.

[1206] "Example 1"

[1207] (Claim 1)

[1208] a means for retrieving company information from a database;

[1209] A means for preprocessing the acquired company information;

[1210] A means for dividing the preprocessed company information into training data and test data;

[1211] means for training an inference model using the training data;

[1212] A means of evaluating the model on test data;

[1213] A means for preprocessing new company information and calculating reliability using a trained model;

[1214] a means for providing the results to users;

[1215] A system including:

[1216] (Claim 2)

[1217] The system according to claim 1, wherein the system deletes missing values ​​from the acquired company information and converts categorical data into numerical data.

[1218] (Claim 3)

[1219] 10. The system of claim 1, wherein multiple learners are used to train the training data.

[1220] "Application Example 1"

[1221] (Claim 1)

[1222] A means for obtaining merchant information from the database;

[1223] A means for pre-processing the acquired member store information;

[1224] A means for dividing the preprocessed merchant information into training data and test data;

[1225] means for training a machine learning model using the training data;

[1226] A means of evaluating the model on test data;

[1227] a means for preprocessing new merchant information and calculating a credit score using the trained model;

[1228] A means of displaying credit scores for virtual stores via a cloud server using a generative AI model;

[1229] A system including:

[1230] (Claim 2)

[1231] The system according to claim 1, wherein the system deletes missing values ​​from the acquired member store information and converts categorical data into numerical data.

[1232] (Claim 3)

[1233] 10. The system of claim 1, wherein the training data is trained using a random forest classifier.

[1234] "Example 2: Combining Emotion Engines"

[1235] (Claim 1)

[1236] A means for obtaining merchant information from the database;

[1237] A means for pre-processing the acquired member store information;

[1238] A means for dividing the preprocessed merchant information into training data and test data;

[1239] means for training a machine learning model using the training data;

[1240] A means of evaluating the model on test data;

[1241] a means for preprocessing new merchant information and calculating a credit score using the trained model;

[1242] means for recognizing a user's emotion;

[1243] means for dynamically adjusting information display content based on user emotions;

[1244] A system including:

[1245] (Claim 2)

[1246] The system according to claim 1, wherein the system deletes missing values ​​from the acquired member store information and converts categorical data into numerical data.

[1247] (Claim 3)

[1248] 10. The system of claim 1, wherein the training data is trained using a random forest classifier.

[1249] "Application example 2 when combining emotion engines"

[1250] (Claim 1)

[1251] A means for obtaining merchant information from the database;

[1252] A means for pre-processing the acquired member store information;

[1253] A means for dividing the preprocessed merchant information into training data and test data;

[1254] means for training a machine learning model using the training data;

[1255] A means of evaluating the model on test data;

[1256] a means for preprocessing new merchant information and calculating a credit score using the trained model;

[1257] A means for recognizing emotions from the user's facial expressions and voice using a camera and microphone installed on the device;

[1258] means for dynamically adjusting the presentation method based on the recognized emotion data;

[1259] A system including:

[1260] (Claim 2)

[1261] The system according to claim 1, wherein the system deletes missing values ​​from the acquired member store information and converts categorical data into numerical data.

[1262] (Claim 3)

[1263] 10. The system of claim 1, wherein the training data is trained using a random forest classifier. [Explanation of symbols]

[1264] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means for obtaining merchant information from the database; A means for pre-processing the acquired member store information; A means for dividing the preprocessed merchant information into training data and test data; means for training a machine learning model using the training data; A means of evaluating the model on test data; a means for preprocessing new merchant information and calculating a credit score using the trained model; A system including:

2. The system according to claim 1 , wherein missing values ​​of the acquired member store information are deleted and categorical data is converted into numerical data.

3. The system of claim 1 , wherein the training data is trained using a random forest classifier.

Images