system

Abstract

We provide the system. [Solution] Means of collecting industry-specific information, payment status, and company information, A means of evaluating a company's payment risk using the collected information, A means of generating an action plan suitable for each company based on the evaluation results, A means to automate notifications based on the generated action plan, A means of monitoring the progress of the action plan and setting follow-up actions, A system that includes this.

Description

Technical Field

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

Background Art

[0002] Patent Document 1 discloses a method for controlling a persona chatbot, which is performed by at least one processor, and includes steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a character of the chatbot, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In conventional debt collection operations, data collection and analysis are often performed manually, which may lead to delays and errors in response. Also, a large part of the judgment on what actions should be taken against which company depends on experience and intuition, making it difficult to conduct efficient collection operations. As a result, problems such as a decrease in the recovery rate and an increase in the workload have occurred. Furthermore, follow-up may not be properly carried out, resulting in a decrease in the effectiveness of debt collection.

Means for Solving the Problems

[0005] The system of the present invention includes means for automatically collecting industry-specific information, payment status, and company information, and for normalizing and centralizing the collected information. This enables efficient data collection and management. Furthermore, by including means for evaluating a company's payment risk using machine learning algorithms, objective and rapid risk assessment becomes possible. The system also includes means for automatically generating action plans tailored to each company based on the evaluation results, and means for automating notifications based on the generated action plans. In addition, by including means for monitoring the progress of the action plans and setting follow-up actions, timely responses become possible. This leads to increased efficiency in debt collection operations, improved collection rates, and reduced workload.

[0006] "Industry-specific information" refers to data about the general characteristics and economic trends of companies in a particular industry.

[0007] "Payment history" refers to information about how a company has fulfilled its debt obligations in the past, including data such as a history of late payments and defaults.

[0008] "Company information" refers to basic information about the company in question, such as company name, address, industry, sales figures, and transaction history.

[0009] "Normalization" is the process of unifying the format and units of collected data and converting it into a consistent data format.

[0010] "Unification" is the process of combining information collected from multiple different data sources into a single dataset.

[0011] A "machine learning algorithm" is a computational method for analyzing large amounts of data to find patterns and regularities, and is a technology that performs predictions and classifications based on these findings.

[0012] "Payment risk" is an assessment of the likelihood that a company may not be able to fulfill its debt obligations in a timely manner. It is an indicator calculated based on factors such as payment history and economic conditions.

[0013] An "action plan" is a plan that outlines the specific steps and measures to be taken in order to achieve a particular goal.

[0014] "Automating notifications" refers to the process of automatically sending communications or notifications based on a pre-set schedule or conditions.

[0015] "Monitoring progress" is the process of tracking the progress of an action plan in real time and recording and verifying the necessary information.

[0016] "Follow-up action" refers to additional actions taken after the initial response, including continuous progress checks and follow-up. [Brief explanation of the drawing]

[0017] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8]It is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] It shows an emotion map to which a plurality of emotions are mapped. [Figure 10] It shows an emotion map to which a plurality of emotions are mapped. [Figure 11] It is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] It is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] It is a sequence diagram showing the processing flow of the data processing system in Example 2 when an emotion engine is combined. [Figure 14] It is a sequence diagram showing the processing flow of the data processing system in Application Example 2 when an emotion engine is combined.

Embodiments for Carrying Out the Invention

[0018] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

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

[0020] In the following embodiments, a processor with a reference numeral (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of a plurality of arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of a plurality of types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0021] In the following embodiments, signed RAM (Random Access Memory) is a memory that temporarily stores information and is used as work memory by the processor.

[0022] In the following embodiments, the signed storage is one or more non-volatile storage devices that store various programs and various parameters. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes.

[0023] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0025] [First Embodiment]

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

[0027] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0028] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0030] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

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

[0034] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0037] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0038] System Overview

[0039] The debt collection system of the present invention aims to improve collection rates by evaluating a company's payment risk and generating and executing appropriate action plans. This system has the functionality to collect and analyze industry-specific information, payment status, company information, etc., and to automate follow-up actions.

[0040] Data collection

[0041] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0042] Data Analysis

[0043] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0044] Generating an action plan

[0045] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0046] Automated notifications

[0047] The terminal schedules notifications based on the generated action plan. For example, it can be configured to make weekly follow-up calls to high-risk companies and send overdue fee warning emails. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0048] Monitoring and follow-up on progress

[0049] The server monitors the progress of the action plan in real time and saves it to the database. The terminal displays the progress to the user and sets additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0050] Specific example

[0051] For example, if company A in the food and beverage industry is included in the list, the server will check company A's payment history and discover that it has a history of frequent payment delays. Based on this, the server will classify company A as "high risk" and generate an action plan that includes weekly follow-up calls and late fee warning emails. The terminal will automatically execute notifications based on this plan, and the server will track progress and provide reports to the user. If the payment situation does not improve, the server will readjust to increase the frequency of follow-ups.

[0052] In this way, the system of the present invention can efficiently manage a company's payment risk and increase the debt collection rate by taking appropriate action.

[0053] The following describes the processing flow.

[0054] Step 1:

[0055] The user logs into the system and uploads a list of companies to be included in the collection process.

[0056] Step 2:

[0057] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0058] Step 3:

[0059] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0060] Step 4:

[0061] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0062] Step 5:

[0063] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0064] Step 6:

[0065] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0066] Step 7:

[0067] The device notifies the user of the generated action plan and displays its contents.

[0068] Step 8:

[0069] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0070] Step 9:

[0071] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0072] Step 10:

[0073] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0074] Step 11:

[0075] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0076] Step 12:

[0077] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0078] In this way, by going through a series of steps, the efficiency and recovery rate of debt collection can be increased.

[0079] (Example 1)

[0080] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0081] Traditional debt collection systems have suffered from low collection rates due to a lack of means to accurately assess the payment risk of businesses and automatically execute appropriate collection actions. Furthermore, the process of prioritizing follow-up actions for each business is often done manually, resulting in inefficiencies.

[0082] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0083] In this invention, the server includes means for collecting industry-specific information, transaction status, and business information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of businesses using the normalized information; means for generating an action plan suitable for each business based on the evaluation results; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan and setting follow-up actions. This enables accurate evaluation of the payment risk of businesses, automation of the execution of optimal collection actions and follow-up of progress, and improves the efficiency of debt collection and the recovery rate.

[0084] "Industry-specific information" refers to information such as economic trends, market conditions, and performance indicators related to a specific industry.

[0085] "Transaction status" refers to the history of transactions a business has conducted in the past and the current status of those transactions, including payment history and delay information.

[0086] "Business information" refers to various data about a specific business, such as basic company information, financial status, and credit information.

[0087] "Normalization" is the process of maintaining data integrity and consistency by converting collected data into a unified format or standard.

[0088] "Centralization" is a process that improves data accessibility and management efficiency by integrating data collected from multiple sources and managing it centrally.

[0089] "Payment risk" is an assessment of the risk that a business will be able to pay its debts in time.

[0090] An "action plan" is a specific action plan customized for each business, including detailed measures such as payment reminders and follow-ups.

[0091] A "notification" is a message or warning that is automatically sent based on the generated action plan.

[0092] "Follow-up" is the process of monitoring the progress of an action plan that has already been implemented and taking additional actions as needed.

[0093] A "machine learning algorithm" is a mathematical model that automatically learns from data and is used to assess the payment risk of businesses.

[0094] The debt collection system of the present invention is a system that aims to improve the collection rate by collecting and analyzing industry-specific information, transaction status, and business information, assessing payment risk, and generating and executing appropriate action plans. This system includes the main elements of a server, terminals, and users.

[0095] Data collection

[0096] First, the user logs into the system and uploads a list of businesses to be included in the collection process. This list may include Excel or CSV files. The terminal temporarily saves this file and then transfers it to the server. Based on the uploaded list of companies, the server accesses external APIs and databases to collect necessary information. For example, it uses APIs from credit rating agencies to obtain credit information on businesses. It also collects industry economic trends and market conditions from other data sources. Next, the server normalizes and centralizes the collected data to maintain data integrity and consistency.

[0097] Data Analysis

[0098] The server runs a machine learning algorithm using normalized data. This algorithm assesses the payment risk of each business based on past payment history, industry information, and business information. The assessment results are stored as a payment risk score, and high-risk businesses are identified.

[0099] Generating an action plan

[0100] Based on the risk assessment results, the server automatically generates customized action plans for each business. For example, high-risk businesses will be assigned frequent contact and payment reminders, while low-risk businesses will be assigned flexible payment plans and notifications. The generated action plans include specific contact methods (email, phone) and contact frequency.

[0101] Automated notifications

[0102] The terminal schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to company A every Monday" or "send a late payment warning email." The server monitors and executes these notifications to ensure they are delivered appropriately.

[0103] Monitoring and follow-up on progress

[0104] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. The user can set additional follow-up actions as needed while viewing the progress. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Based on the progress, the server calculates the timing of the next action and sets the follow-up action.

[0105] Specific example

[0106] For example, if a restaurant business A is included in the list, the user uploads the list containing business A to the system. The server retrieves business A's payment history from a credit bureau's API and confirms that there have been frequent payment delays in the past. Based on this, the server classifies business A as "high risk." The server generates an action plan for business A, including weekly follow-up calls and emails warning about late fees. The terminal automatically schedules notifications based on this action plan, and the server tracks the progress and provides reports to the user.

[0107] Examples of prompts for generative AI models

[0108] The following prompt statements can be used as input to the generated AI model:

[0109] "Please generate a debt collection action plan based on Business A's past payment history. Business A belongs to the food and beverage industry and has experienced multiple payment delays in the past year."

[0110] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0111] Step 1:

[0112] The user logs into the system and enters their username and password. The server authenticates the entered information and allows the user to log in. In this process, the user's authentication information is provided as input, and the authentication result and user session are generated as output.

[0113] Step 2:

[0114] Users upload a list of businesses to be included in the collection process to the system. Common file formats include Excel and CSV. The terminal temporarily stores the uploaded file and then transfers it to the server. The uploaded file is provided as input, and that file is transferred to the server as output.

[0115] Step 3:

[0116] The server accesses external APIs and databases based on the uploaded list of businesses to collect necessary information. Specifically, when making API calls, it sends requests to the APIs of credit rating agencies to retrieve credit information, past transaction history, and other data. The input is the list of businesses, and the output is detailed information about each collected business.

[0117] Step 4:

[0118] The server normalizes and centralizes the collected data. Specifically, it converts data collected in different formats into a unified format. This process ensures data integrity and consistency. Raw data is taken as input, and normalized, centralized data is generated as output.

[0119] Step 5:

[0120] The server runs a machine learning algorithm using normalized data. This algorithm assesses payment risk based on past payment history, industry information, and business information. Specifically, data is input into the algorithm, and a risk score is output.

[0121] Step 6:

[0122] The server saves the results of machine learning algorithm execution to a database. The input is the risk assessment results, and the output is the risk score and a list of high-risk businesses.

[0123] Step 7:

[0124] The server automatically generates customized action plans for each business based on the risk assessment results. Specifically, it sets up frequent contact and payment reminders for high-risk businesses. The input is the risk assessment results, and the output is the action plan.

[0125] Step 8:

[0126] The device schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to business A every Monday" or "send a late payment warning email." The action plan is set as the input, and the notification schedule is set as the output.

[0127] Step 9:

[0128] The server executes scheduled notifications and monitors their execution status. Specifically, it automatically sends emails and phone reminders at scheduled times. The notification schedule is the input, and a notification execution log is generated as the output.

[0129] Step 10:

[0130] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. Progress data is generated as input, and progress reports for the user are generated as output.

[0131] Step 11:

[0132] Users monitor the progress and set additional follow-up actions as needed. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Progress is set as the input, and additional follow-up actions are set as the output.

[0133] Step 12:

[0134] The server calculates the timing of the next action based on the progress and sets up follow-up actions. Progress data is taken as input, and the next action plan is generated as output.

[0135] (Application Example 1)

[0136] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0137] Debt collection systems, designed to assess a company's payment risk and generate and implement effective action plans, currently suffer from problems such as insufficient real-time monitoring and follow-up, or limited automation, making efficient debt collection difficult. Furthermore, automatically generating and reliably executing notifications remains a challenge.

[0138] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0139] In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating a company's payment risk using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for acquiring company data using an external API; means for quantifying a company's payment risk using the collected data; and means for automatically generating the content and timing of notifications based on the risk assessment. This makes it possible to evaluate a company's payment risk in real time, generate an optimal action plan according to that risk, and reliably follow up.

[0140] "Industry-specific information" refers to data and statistics related to a specific industry.

[0141] "Payment status" refers to data that shows a company's past or present payment history and payment status.

[0142] "Company information" refers to detailed information about a company, including its basic data, financial status, and management status.

[0143] An "external API" refers to an interface for obtaining information in real time from external databases or services.

[0144] "Company data" refers to various types of data related to a company, such as basic company information, payment history, and credit information.

[0145] "Risk assessment" refers to the process of analyzing and evaluating a company's ability to pay its debts and its associated risks based on collected information.

[0146] An "action plan" refers to a plan that specifies the optimal response methods for each company based on the results of a risk assessment.

[0147] "Automated notifications" refers to a feature that automatically sends notifications based on an action plan generated by the system, without requiring manual action from the user.

[0148] "Follow-up" refers to the activity of monitoring the progress of an action plan after its implementation and taking additional actions as needed.

[0149] "Progress monitoring" refers to the process of tracking in real time whether the action plan is being implemented, and the subsequent responses and progress of the company.

[0150] "Quantification" refers to the process of converting evaluation results and data into specific numbers or scores.

[0151] "Notification content" refers to the main body of the warning or reminder message that is sent.

[0152] "Notification timing" refers to the timing and schedule of when notifications will be sent.

[0153] This invention relates to a system for evaluating a company's payment risk and generating and executing an appropriate action plan. To specifically implement this invention, various means of servers, terminals, and users work together.

[0154] System Overview

[0155] Data collection

[0156] The server receives user-uploaded lists and collects industry-specific information, payment status, and company information through external APIs and internal databases. External APIs retrieve information in real time, and the retrieved data is normalized and centralized within the server.

[0157] Data Analysis

[0158] The server runs machine learning algorithms using normalized data to assess each company's payment risk. The risk assessment is based on collected payment history, industry information, and company information. Specifically, the assessment process quantifies each company's payment risk and stores the risk assessment results in a database.

[0159] Generating an action plan

[0160] The server automatically generates customized action plans for each company based on the assessment results. These action plans include specific contact methods (e.g., email, phone) and contact frequency. Companies at high risk are configured to receive frequent contact and warning emails.

[0161] Automated notifications

[0162] The device schedules and automatically executes notifications based on the generated action plan. The server uses the Twilio API to send SMS and phone notifications, automatically generating the content and timing of the notifications. This ensures that the appropriate actions are taken.

[0163] Monitoring and follow-up on progress

[0164] The server monitors the progress of the action plan in real time and stores the data in a database. The terminal displays the progress to the user and allows them to set additional follow-up actions as needed. This ensures that appropriate tracking and response are always based on the latest information.

[0165] Specific example

[0166] For example, consider the case where the company list includes "CompanyA" and "CompanyB" as follows.

[0167] Python

[0168] company_list = ["CompanyA", "CompanyB"]

[0169] api_key = "your_api_key"

[0170] account_sid = "your_account_sid"

[0171] auth_token = "your_auth_token"

[0172] secu_collect = SecuCollect(api_key, account_sid, auth_token)

[0173] secu_collect.run(company_list)

[0174] Here, the server retrieves payment history and basic information for "CompanyA" and "CompanyB" and performs a risk assessment. Subsequently, it sends warning emails and SMS reminders to high-risk companies and monitors the progress of the action plan. This process is performed automatically by the system, and users can check the results in real time.

[0175] Example of a prompt

[0176] The following are examples of prompts used when utilizing generative AI models:

[0177] Develop an application that assesses a company's payment risks and automates appropriate response actions based on those assessments. This application should handle everything from data collection and risk assessment to action plan generation, automated notifications, and follow-up.

[0178] Based on this text, the system provides a set of functions to efficiently manage a company's payment risk and improve its debt collection rate.

[0179] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0180] Step 1:

[0181] The server receives input from users who log in to the system and upload a list of companies. Based on this list, the server accesses external APIs and internal databases to collect industry-specific information, payment status, and company information. This collected data is obtained in its raw, unprocessed state.

[0182] Step 2:

[0183] The server normalizes and centralizes the collected raw data. Specifically, it standardizes each data field according to the database schema and transforms data from multiple sources into an integrated dataset. This process formats the input data into a consistent format.

[0184] Step 3:

[0185] The server inputs normalized data into a machine learning algorithm to quantify payment risk. This risk score is calculated based on payment history, the company's financial status, industry trends, and other factors. The risk score is stored in the server's database and used to generate subsequent action plans.

[0186] Step 4:

[0187] The server generates customized action plans for each company based on the risk assessment results. For example, it creates plans that include frequent contact and warning emails for high-risk companies, and regular reminders for medium-risk companies. These action plans are automatically generated within the server.

[0188] Step 5:

[0189] The device receives an action plan from the server and schedules notifications. Scheduled notifications are delivered via the Twilio API as SMS or phone calls. The content and timing of each notification are automatically determined based on the action plan.

[0190] Step 6:

[0191] The server monitors the progress of the action plan in real time. It tracks data such as whether each notification was sent successfully and whether there was a response from the company, and stores this data in a database. This information is then used to inform the next follow-up actions.

[0192] Step 7:

[0193] The terminal displays the progress to the user and allows for additional follow-up actions as needed. Users can stay informed in real time through progress reports and take necessary corrections or additional actions. This information is fed back to the server and considered when generating the next action plan.

[0194] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0195] System Overview

[0196] The debt collection system of the present invention aims to improve the collection rate by evaluating a company's payment risk and generating and executing an appropriate action plan. It also incorporates an emotion engine to recognize the user's emotions and adaptively execute actions. This system collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[0197] Data collection

[0198] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0199] Data Analysis

[0200] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0201] Generating an action plan

[0202] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0203] Emotional engine integration

[0204] The server uses an emotion engine to recognize the user's emotions. For example, it detects the user's emotional state through text message and voice analysis. Based on the recognized emotions, the server adaptively changes its action plan. For example, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[0205] Automated notifications

[0206] The terminal schedules notifications based on the generated action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0207] Monitoring and follow-up on progress

[0208] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The terminal displays the progress to the user and provides an interface for setting additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0209] Specific example

[0210] For example, consider a case where company A in the food and beverage industry is included in the list. The server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of the person in charge at company A, and if it determines that they are under high stress, it modifies the follow-up content to be softer and more supportive.

[0211] In this way, the system of the present invention efficiently manages a company's payment risk, increases the debt collection rate by taking appropriate actions, and enables flexible responses that respond to the user's emotions.

[0212] The following describes the processing flow.

[0213] Step 1:

[0214] The user logs into the system and uploads a list of companies to be included in the collection process.

[0215] Step 2:

[0216] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0217] Step 3:

[0218] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0219] Step 4:

[0220] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0221] Step 5:

[0222] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0223] Step 6:

[0224] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0225] Step 7:

[0226] The server uses an emotion engine to recognize the emotions of company representatives and users from text messages and voice data.

[0227] Step 8:

[0228] The server adaptively modifies the action plan based on the recognized emotion data. For example, if the person in charge is experiencing stress, the communication method and message content will be changed to softer language.

[0229] Step 9:

[0230] The device notifies the user of the generated action plan and displays its contents.

[0231] Step 10:

[0232] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0233] Step 11:

[0234] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0235] Step 12:

[0236] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0237] Step 13:

[0238] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0239] Step 14:

[0240] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0241] By following this series of steps, it is possible not only to improve the efficiency and recovery rate of debt collection, but also to provide flexible responses that respond to the user's emotions.

[0242] (Example 2)

[0243] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0244] In modern businesses, efficient and effective debt collection is crucial. However, assessing payment risk and setting appropriate follow-up actions requires considerable effort and time. Furthermore, traditional systems struggle to respond flexibly to user emotions, often leading to stress. This results in challenges such as declining debt collection rates and increased burden on users.

[0245] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of companies using a machine learning algorithm with the collected information; means for generating customized action plans suitable for each company based on the evaluation results; means for recognizing the user's emotions using an emotion engine and adaptively changing the action plan; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan, saving the history of executed actions, and setting follow-up actions. This enables efficient and effective debt collection and allows for flexible responses that respond to the user's emotions.

[0246] "Industry-specific information" refers to information such as economic conditions, market trends, and industry-specific management indicators related to a particular industry.

[0247] "Payment status" refers to information about a company's past and present payment history, payment patterns, and whether or not there have been any payment delays.

[0248] "Company information" refers to basic company data such as name, address, date of establishment, financial status, and number of employees.

[0249] "Normalization" refers to the process of converting collected data into a consistent format and supplementing any missing fields.

[0250] "Unification" refers to the process of integrating information collected from different data sources and consolidating it into a single database.

[0251] A "machine learning algorithm" refers to an algorithm used to analyze data, identify patterns, and perform predictions and classifications.

[0252] "Payment risk" refers to the possibility that a particular company may delay or default on payments in the future.

[0253] A "customized action plan" refers to a plan that includes specific countermeasures, developed based on each company's payment risk assessment.

[0254] An "emotion engine" refers to a technology that analyzes a user's text and voice to identify their emotional state.

[0255] "Automated notifications" refers to the process of automatically sending notifications to users or companies based on a pre-configured action plan.

[0256] "Progress monitoring" refers to the process of tracking and evaluating in real time how well the set action plan is being implemented.

[0257] "Follow-up action" refers to taking additional steps or making contact based on the established action plan.

[0258] The debt collection system of the present invention is implemented with the following configuration. This system evaluates a company's payment risk, generates and executes an appropriate action plan, aims to improve the collection rate, and also has the function of recognizing the user's emotions by combining it with an emotion engine and performing actions adaptively.

[0259] Software and hardware to be used

[0260] This system is implemented using the following software and hardware:

[0261] 1. Server: Data collection, data normalization, execution of machine learning algorithms, generation of action plans, sentiment recognition by sentiment engine, monitoring and execution of notifications.

[0262] 2. Terminal: Provides the user interface, schedules and executes notifications, and displays progress.

[0263] 3. Software used:

[0264] External APIs for data collection: Examples include "Dun & Bradstreet API" and "Clearbit API".

[0265] Data analysis and machine learning: Python libraries "Pandas," "Scikit-learn," and "TENSORFLOW®"

[0266] Sentiment recognition: Microsoft® Azure® Text Analytics and IBM Watson® Tone Analyzer.

[0267] Databases: PostgreSQL and MongoDB for centralized data storage.

[0268] System operation

[0269] To assess a company's payment risk, the server first requires the user to log in and upload a list of companies to be collected from. Based on the uploaded list, the server accesses external APIs and databases to collect necessary data such as basic company information, payment history, and industry trends.

[0270] The collected data is normalized and centralized on the server using the "Pandas" library, and then stored in "PostgreSQL" or "MongoDB". The server then uses "Scikit-learn" or "TensorFlow" to run machine learning algorithms to assess a company's payment risk. The assessment criteria include historical payment history, industry economic trends, and company information.

[0271] Based on risk assessment results from machine learning algorithms, the server generates customized action plans for each company. Specifically, it sets up frequent contact and payment reminders for high-risk companies and proposes flexible payment plans for low-risk companies. The generated action plans include contact methods (email, phone) and contact frequency.

[0272] Furthermore, the server uses emotion engines such as "Text Analytics" and "Tone Analyzer" to recognize the user's emotions and adaptively change the action plan. For example, if it detects that the user is stressed, it softens the notification content and emphasizes support.

[0273] Based on the generated action plan, the device schedules and executes notifications. Specifically, it sets up weekly follow-up calls and reminders for high-risk companies and executes them automatically. The server monitors the execution status of notifications and stores progress data in a database. The device displays the progress to the user and sets additional follow-up actions as needed.

[0274] Specific example

[0275] For example, if company A in the food and beverage industry is included in the list, the server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of company A's representative, and if it determines that the representative is stressed, it changes the content of the follow-up to be softer and more supportive.

[0276] Prompts for Generative AI Models

[0277] The following are specific examples of prompt statements to input into a generative AI model:

[0278] Prompt message 1:

[0279] "Assess the payment risk of the following company and generate an action plan. Company name: Sample Co., Ltd., Past payment history: Monthly late payments, Industry information: Food and beverage industry."

[0280] Prompt message 2:

[0281] "Please generate a follow-up message for when a user is experiencing stress. Current message: 'Payment deadline is approaching. Please take action as soon as possible.'"

[0282] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0283] Step 1:

[0284] The user logs in to the system and uploads a list of target companies for collection. As a result, the user's input becomes a list of companies. The server receives this list and starts processing. Specifically, the user enters authentication information on the system's login screen and clicks the "Login" button. Then, the user clicks the "File Upload" button to select a company list file (e.g., CSV, Excel) and clicks the "Upload" button.

[0285] Step 2:

[0286] The server receives the uploaded company list and accesses external APIs or databases to collect the necessary information. The input is the company list, and the output is data such as the basic information of the company, payment history, and industry trends. Specifically, the server uses the company ID to send requests to the "Dun & Bradstreet API" or "Clearbit API" to obtain detailed company information. The information includes the company name, address, establishment date, payment history, etc.

[0287] Step 3:

[0288] The server normalizes and unifies the collected data. The input is multiple collected data sets, and the output is a normalized and consistent data set. Specifically, the server uses the "Pandas" library in Python to clean and format the data. For example, it unifies the date format and complements missing data.

[0289] Step 4:

[0290] The server saves the normalized data to the database. The input is the normalized data set, and the output is the data saved in the database. Specifically, the server executes SQL queries to insert data into the "PostgreSQL" or "MongoDB" database.

[0291] Step 5:

[0292] The server uses normalized data to run machine learning algorithms and assess companies' payment risk. The input is data read from a database, and the output is a risk assessment score for each company. Specifically, the server uses the Random Forest algorithm from the "Scikit-learn" library to predict payment risk and calculate the score.

[0293] Step 6:

[0294] The server generates customized action plans for each company based on the risk assessment results. The input is the risk assessment score, and the output is the action plan. Specifically, the server sets up frequent contact and payment reminders for high-risk companies, and proposes flexible payment plans for low-risk companies. Specific contact methods (email, phone) and contact frequency are set for each company.

[0295] Step 7:

[0296] The server uses an emotion engine to recognize the user's emotions and adaptively modify the action plan. The input is the user's text messages or voice data, and the output is the modified action plan. Specifically, the server uses "Text Analytics" and "Tone Analyzer" to analyze the user's emotional state, and if it determines that the user is experiencing high levels of stress or anxiety, it softens the content of the action plan and emphasizes support.

[0297] Step 8:

[0298] The terminal schedules and executes notifications based on the generated action plan. The input is the action plan, and the output is the notification to the company. Specifically, the terminal schedules and automatically executes weekly follow-up calls and emails to high-risk companies.

[0299] Step 9:

[0300] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The input is the data of the actions performed, and the output is a progress report. Specifically, the server periodically updates the progress data and saves it to the database.

[0301] Step 10:

[0302] The terminal displays the progress to the user and allows for additional follow-up actions as needed. The input is a progress report, and the output is a notification to the user and additional follow-up actions. Specifically, the terminal graphically displays the progress through a dashboard screen and provides an interface where the user can configure the necessary actions.

[0303] (Application Example 2)

[0304] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as a "server" and the smart device 14 as a "terminal".

[0305] Current debt collection systems, while capable of assessing a company's payment risk and generating action plans based on that assessment, struggle to provide adaptive responses that take into account the customer's emotional state. This results in one-sided communication with customers, hindering improvements in collection rates. Furthermore, it's difficult for employees to assess the situation in real time while interacting with customers. Solving these problems is essential to providing a more effective debt collection system.

[0306] The specific processing by the specific processing unit 290 of the data processing apparatus 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and enterprise information, means for evaluating the payment risk of enterprises using the collected information, means for generating an action plan suitable for each enterprise based on the evaluation result, means for automating notifications based on the generated action plan, means for monitoring the progress of the action plan and setting follow-up actions, means for recognizing the emotional state of the user, and means for adaptively changing the action plan based on the recognized emotional state. Thereby, a flexible and adaptive response considering the emotional state of the customer becomes possible, and an improvement in the recovery rate and customer satisfaction can be achieved.

[0307] "Industry-specific information" refers to information such as economic trends, competition status, and trading practices related to a specific industry.

[0308] "Payment status" refers to data such as past and current payment histories, outstanding amounts, and frequency of payment delays.

[0309] "Enterprise information" refers to information regarding the basic information, business status, financial condition, and creditworthiness of a corporation.

[0310] "Payment risk" refers to a risk assessment calculated based on factors that affect the likelihood of an enterprise fulfilling future payments.

[0311] "Action plan" refers to a plan that includes specific actions to be taken for an enterprise based on the evaluation result and their schedule.

[0312] "Automation of notifications" refers to a process of automatically sending notifications at specific timings based on a predetermined action plan.

[0313] "Follow-up action" refers to actions and measures to be additionally taken according to the progress after notification.

[0314] "Emotional state" refers to the user's current psychological and emotional condition, and is recognized through voice, text, facial expressions, etc.

[0315] "Adaptively modifying the action plan based on the perceived emotional state" means analyzing the user's emotions and flexibly changing the predetermined plan accordingly.

[0316] This invention relates to a debt collection system that assesses a company's payment risk and generates and executes an appropriate action plan. The system recognizes the user's emotional state and adaptively executes actions based on that state. Specifically, it collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[0317] System Overview

[0318] The server collects company information when users log in and upload a list of companies to be collected from. This information includes basic company information, past payment history, and industry economic trends. The collected data is normalized and centralized. Next, the server uses machine learning algorithms to assess the companies' payment risk. Based on this, high-risk companies will require more frequent follow-up, while low-risk companies will be offered flexible payment plans.

[0319] Emotional engine integration

[0320] The server uses an emotion engine to recognize the user's emotional state. For example, it can determine if a user is stressed through text message or voice analysis. Based on the recognized emotional state, it adaptively changes the action plan. For instance, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[0321] Automated notifications and progress monitoring

[0322] Based on the generated action plan, notifications are sent automatically. For example, high-risk companies may receive weekly follow-up calls or emails reminding them of late fees. The server monitors these notification processes and stores the execution history in a database. It also monitors progress in real time and sets up additional follow-up actions as needed.

[0323] Specific example

[0324] For example, consider a case where company B in the retail industry is included in the list. The server checks company B's payment history and discovers that it has a history of frequent payment delays. Based on this information, the server classifies company B as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. Additionally, during interactions with customers via smart glasses, the emotion engine analyzes the emotions of company B's representative, and if it determines that the representative is highly stressed, it changes the follow-up to be softer and more supportive. In this way, it is possible to increase the debt collection rate while simultaneously improving customer satisfaction.

[0325] Example prompt message

[0326] Let's say a customer enters a store, and an employee puts on smart glasses and begins serving them. The system detects the customer's stress level from their voice and determines they are high-risk based on their past payment history. Based on this information, the smart glasses display shows a message saying, "Use gentle language and emphasize support."

[0327] Prompt text for input example

[0328] Parameters:

[0329] customer_id: 12345

[0330] audio_input: "path / to / customer / voice / input.wav"

[0331] Query:

[0332] Please tell me about customer service methods that take into account the customer's payment risk level and emotional state.

[0333] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0334] Step 1: The user logs into the system and uploads a list of companies to be included in the collection process.

[0335] Input: User login information, list of companies to be included in the collection.

[0336] Output: A list of companies subject to recall is saved on the server.

[0337] Specific operation: The user logs in through the system interface and uploads a list of companies to be targeted for collection. The server receives this list and stores it in its internal database.

[0338] Step 2: The server collects corporate information.

[0339] Input: List of companies to be recalled

[0340] Output: Basic company information, payment history, industry economic trends

[0341] Specific operation: The server accesses external APIs and databases to collect basic information, past payment history, and industry economic trends of the target companies. The collected data is normalized and centralized.

[0342] Step 3: The server uses machine learning algorithms to assess the company's payment risk.

[0343] Input: Basic company information, payment history, industry economic trends

[0344] Output: Payment risk assessment results for each company

[0345] Specific operation: The server inputs the collected data into a machine learning algorithm to evaluate the payment risk of each company. The evaluation results are stored in a database.

[0346] Step 4: The server generates an action plan tailored to each company.

[0347] Input: Payment risk assessment results

[0348] Output: Action plans tailored to each company

[0349] Specific operation: Based on the payment risk assessment results, the server generates action plans that include frequent follow-up for high-risk companies and flexible payment plans for low-risk companies.

[0350] Step 5: The device automates notifications based on the generated action plan.

[0351] Input: Action Plan

[0352] Output: Automatically generated notifications (email, phone call, etc.)

[0353] Specific operation: The device automatically sends notifications via email or phone at the appropriate time based on the action plan.

[0354] Step 6: The server uses the emotion engine to recognize the user's emotional state.

[0355] Input: Text messages and audio data from the user.

[0356] Output: User's emotional state

[0357] Specific operation: The server uses an emotion engine to analyze text messages and voice data to detect the user's emotional state.

[0358] Step 7: The server adaptively modifies the action plan based on the recognized emotional state.

[0359] Input: User's emotional state

[0360] Output: Adaptively modified action plan

[0361] Specific operation: The server analyzes the user's emotional state and flexibly modifies the action plan accordingly. For example, if the user is feeling stressed, the follow-up content will be softened and emphasized to be more supportive.

[0362] Step 8: The server monitors the progress of the action plan and sets up follow-up actions.

[0363] Input: Action plan execution history

[0364] Output: Next follow-up action

[0365] Specific operation: The server saves the execution status of the action plan to a database and monitors it in real time. Based on the progress, it calculates the timing of the next follow-up and notifies the user.

[0366] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0367] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0368] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0369] [Second Embodiment]

[0370] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0371] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0372] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0373] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0374] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0375] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0376] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0377] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0378] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0379] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0380] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0381] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0382] System Overview

[0383] The debt collection system of the present invention aims to improve collection rates by evaluating a company's payment risk and generating and executing appropriate action plans. This system has the functionality to collect and analyze industry-specific information, payment status, company information, etc., and to automate follow-up actions.

[0384] Data collection

[0385] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0386] Data Analysis

[0387] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0388] Generating an action plan

[0389] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0390] Automated notifications

[0391] The terminal schedules notifications based on the generated action plan. For example, it can be configured to make weekly follow-up calls to high-risk companies and send overdue fee warning emails. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0392] Monitoring and follow-up on progress

[0393] The server monitors the progress of the action plan in real time and saves it to the database. The terminal displays the progress to the user and sets additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0394] Specific example

[0395] For example, if company A in the food and beverage industry is included in the list, the server will check company A's payment history and discover that it has a history of frequent payment delays. Based on this, the server will classify company A as "high risk" and generate an action plan that includes weekly follow-up calls and late fee warning emails. The terminal will automatically execute notifications based on this plan, and the server will track progress and provide reports to the user. If the payment situation does not improve, the server will readjust to increase the frequency of follow-ups.

[0396] In this way, the system of the present invention can efficiently manage a company's payment risk and increase the debt collection rate by taking appropriate action.

[0397] The following describes the processing flow.

[0398] Step 1:

[0399] The user logs into the system and uploads a list of companies to be included in the collection process.

[0400] Step 2:

[0401] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0402] Step 3:

[0403] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0404] Step 4:

[0405] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0406] Step 5:

[0407] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0408] Step 6:

[0409] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0410] Step 7:

[0411] The device notifies the user of the generated action plan and displays its contents.

[0412] Step 8:

[0413] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0414] Step 9:

[0415] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0416] Step 10:

[0417] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0418] Step 11:

[0419] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0420] Step 12:

[0421] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0422] In this way, by going through a series of steps, the efficiency and recovery rate of debt collection can be increased.

[0423] (Example 1)

[0424] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0425] Traditional debt collection systems have suffered from low collection rates due to a lack of means to accurately assess the payment risk of businesses and automatically execute appropriate collection actions. Furthermore, the process of prioritizing follow-up actions for each business is often done manually, resulting in inefficiencies.

[0426] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0427] In this invention, the server includes means for collecting industry-specific information, transaction status, and business information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of businesses using the normalized information; means for generating an action plan suitable for each business based on the evaluation results; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan and setting follow-up actions. This enables accurate evaluation of the payment risk of businesses, automation of the execution of optimal collection actions and follow-up of progress, and improves the efficiency of debt collection and the recovery rate.

[0428] "Industry-specific information" refers to information such as economic trends, market conditions, and performance indicators related to a specific industry.

[0429] "Transaction status" refers to the history of transactions a business has conducted in the past and the current status of those transactions, including payment history and delay information.

[0430] "Business information" refers to various data about a specific business, such as basic company information, financial status, and credit information.

[0431] "Normalization" is the process of maintaining data integrity and consistency by converting collected data into a unified format or standard.

[0432] "Centralization" is a process that improves data accessibility and management efficiency by integrating data collected from multiple sources and managing it centrally.

[0433] "Payment risk" is an assessment of the risk that a business will be able to pay its debts in time.

[0434] An "action plan" is a specific action plan customized for each business, including detailed measures such as payment reminders and follow-ups.

[0435] A "notification" is a message or warning that is automatically sent based on the generated action plan.

[0436] "Follow-up" is the process of monitoring the progress of an action plan that has already been implemented and taking additional actions as needed.

[0437] A "machine learning algorithm" is a mathematical model that automatically learns from data and is used to assess the payment risk of businesses.

[0438] The debt collection system of the present invention is a system that aims to improve the collection rate by collecting and analyzing industry-specific information, transaction status, and business information, assessing payment risk, and generating and executing appropriate action plans. This system includes the main elements of a server, terminals, and users.

[0439] Data collection

[0440] First, the user logs into the system and uploads a list of businesses to be included in the collection process. This list may include Excel or CSV files. The terminal temporarily saves this file and then transfers it to the server. Based on the uploaded list of companies, the server accesses external APIs and databases to collect necessary information. For example, it uses APIs from credit rating agencies to obtain credit information on businesses. It also collects industry economic trends and market conditions from other data sources. Next, the server normalizes and centralizes the collected data to maintain data integrity and consistency.

[0441] Data Analysis

[0442] The server runs a machine learning algorithm using normalized data. This algorithm assesses the payment risk of each business based on past payment history, industry information, and business information. The assessment results are stored as a payment risk score, and high-risk businesses are identified.

[0443] Generating an action plan

[0444] Based on the risk assessment results, the server automatically generates customized action plans for each business. For example, high-risk businesses will be assigned frequent contact and payment reminders, while low-risk businesses will be assigned flexible payment plans and notifications. The generated action plans include specific contact methods (email, phone) and contact frequency.

[0445] Automated notifications

[0446] The terminal schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to company A every Monday" or "send a late payment warning email." The server monitors and executes these notifications to ensure they are delivered appropriately.

[0447] Monitoring and follow-up on progress

[0448] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. The user can set additional follow-up actions as needed while viewing the progress. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Based on the progress, the server calculates the timing of the next action and sets the follow-up action.

[0449] Specific example

[0450] For example, if a restaurant business A is included in the list, the user uploads the list containing business A to the system. The server retrieves business A's payment history from a credit bureau's API and confirms that there have been frequent payment delays in the past. Based on this, the server classifies business A as "high risk." The server generates an action plan for business A, including weekly follow-up calls and emails warning about late fees. The terminal automatically schedules notifications based on this action plan, and the server tracks the progress and provides reports to the user.

[0451] Examples of prompts for generative AI models

[0452] The following prompt statements can be used as input to the generated AI model:

[0453] "Please generate a debt collection action plan based on Business A's past payment history. Business A belongs to the food and beverage industry and has experienced multiple payment delays in the past year."

[0454] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0455] Step 1:

[0456] The user logs into the system and enters their username and password. The server authenticates the entered information and allows the user to log in. In this process, the user's authentication information is provided as input, and the authentication result and user session are generated as output.

[0457] Step 2:

[0458] Users upload a list of businesses to be included in the collection process to the system. Common file formats include Excel and CSV. The terminal temporarily stores the uploaded file and then transfers it to the server. The uploaded file is provided as input, and that file is transferred to the server as output.

[0459] Step 3:

[0460] The server accesses external APIs and databases based on the uploaded list of businesses to collect necessary information. Specifically, when making API calls, it sends requests to the APIs of credit rating agencies to retrieve credit information, past transaction history, and other data. The input is the list of businesses, and the output is detailed information about each collected business.

[0461] Step 4:

[0462] The server normalizes and centralizes the collected data. Specifically, it converts data collected in different formats into a unified format. This process ensures data integrity and consistency. Raw data is taken as input, and normalized, centralized data is generated as output.

[0463] Step 5:

[0464] The server runs a machine learning algorithm using normalized data. This algorithm assesses payment risk based on past payment history, industry information, and business information. Specifically, data is input into the algorithm, and a risk score is output.

[0465] Step 6:

[0466] The server saves the results of machine learning algorithm execution to a database. The input is the risk assessment results, and the output is the risk score and a list of high-risk businesses.

[0467] Step 7:

[0468] The server automatically generates customized action plans for each business based on the risk assessment results. Specifically, it sets up frequent contact and payment reminders for high-risk businesses. The input is the risk assessment results, and the output is the action plan.

[0469] Step 8:

[0470] The device schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to business A every Monday" or "send a late payment warning email." The action plan is set as the input, and the notification schedule is set as the output.

[0471] Step 9:

[0472] The server executes scheduled notifications and monitors their execution status. Specifically, it automatically sends emails and phone reminders at scheduled times. The notification schedule is the input, and a notification execution log is generated as the output.

[0473] Step 10:

[0474] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. Progress data is generated as input, and progress reports for the user are generated as output.

[0475] Step 11:

[0476] Users monitor the progress and set additional follow-up actions as needed. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Progress is set as the input, and additional follow-up actions are set as the output.

[0477] Step 12:

[0478] The server calculates the timing of the next action based on the progress and sets up follow-up actions. Progress data is taken as input, and the next action plan is generated as output.

[0479] (Application Example 1)

[0480] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0481] Debt collection systems, designed to assess a company's payment risk and generate and implement effective action plans, currently suffer from problems such as insufficient real-time monitoring and follow-up, or limited automation, making efficient debt collection difficult. Furthermore, automatically generating and reliably executing notifications remains a challenge.

[0482] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0483] In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating a company's payment risk using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for acquiring company data using an external API; means for quantifying a company's payment risk using the collected data; and means for automatically generating the content and timing of notifications based on the risk assessment. This makes it possible to evaluate a company's payment risk in real time, generate an optimal action plan according to that risk, and reliably follow up.

[0484] "Industry-specific information" refers to data and statistics related to a specific industry.

[0485] "Payment status" refers to data that shows a company's past or present payment history and payment status.

[0486] "Company information" refers to detailed information about a company, including its basic data, financial status, and management status.

[0487] An "external API" refers to an interface for obtaining information in real time from external databases or services.

[0488] "Company data" refers to various types of data related to a company, such as basic company information, payment history, and credit information.

[0489] "Risk assessment" refers to the process of analyzing and evaluating a company's ability to pay its debts and its associated risks based on collected information.

[0490] An "action plan" refers to a plan that specifies the optimal response methods for each company based on the results of a risk assessment.

[0491] "Automated notifications" refers to a feature that automatically sends notifications based on an action plan generated by the system, without requiring manual action from the user.

[0492] "Follow-up" refers to the activity of monitoring the progress of an action plan after its implementation and taking additional actions as needed.

[0493] "Progress monitoring" refers to the process of tracking in real time whether the action plan is being implemented, and the subsequent responses and progress of the company.

[0494] "Quantification" refers to the process of converting evaluation results and data into specific numbers or scores.

[0495] "Notification content" refers to the main body of the warning or reminder message that is sent.

[0496] "Notification timing" refers to the timing and schedule of when notifications will be sent.

[0497] This invention relates to a system for evaluating a company's payment risk and generating and executing an appropriate action plan. To specifically implement this invention, various means of servers, terminals, and users work together.

[0498] System Overview

[0499] Data collection

[0500] The server receives user-uploaded lists and collects industry-specific information, payment status, and company information through external APIs and internal databases. External APIs retrieve information in real time, and the retrieved data is normalized and centralized within the server.

[0501] Data Analysis

[0502] The server runs machine learning algorithms using normalized data to assess each company's payment risk. The risk assessment is based on collected payment history, industry information, and company information. Specifically, the assessment process quantifies each company's payment risk and stores the risk assessment results in a database.

[0503] Generating an action plan

[0504] The server automatically generates customized action plans for each company based on the assessment results. These action plans include specific contact methods (e.g., email, phone) and contact frequency. Companies at high risk are configured to receive frequent contact and warning emails.

[0505] Automated notifications

[0506] The device schedules and automatically executes notifications based on the generated action plan. The server uses the Twilio API to send SMS and phone notifications, automatically generating the content and timing of the notifications. This ensures that the appropriate actions are taken.

[0507] Monitoring and follow-up on progress

[0508] The server monitors the progress of the action plan in real time and stores the data in a database. The terminal displays the progress to the user and allows them to set additional follow-up actions as needed. This ensures that appropriate tracking and response are always based on the latest information.

[0509] Specific example

[0510] For example, consider the case where the company list includes "CompanyA" and "CompanyB" as follows.

[0511] Python

[0512] company_list = ["CompanyA", "CompanyB"]

[0513] api_key = "your_api_key"

[0514] account_sid = "your_account_sid"

[0515] auth_token = "your_auth_token"

[0516] secu_collect = SecuCollect(api_key, account_sid, auth_token)

[0517] secu_collect.run(company_list)

[0518] Here, the server retrieves payment history and basic information for "CompanyA" and "CompanyB" and performs a risk assessment. Subsequently, it sends warning emails and SMS reminders to high-risk companies and monitors the progress of the action plan. This process is performed automatically by the system, and users can check the results in real time.

[0519] Example of a prompt

[0520] The following are examples of prompts used when utilizing generative AI models:

[0521] Develop an application that assesses a company's payment risks and automates appropriate response actions based on those assessments. This application should handle everything from data collection and risk assessment to action plan generation, automated notifications, and follow-up.

[0522] Based on this text, the system provides a set of functions to efficiently manage a company's payment risk and improve its debt collection rate.

[0523] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0524] Step 1:

[0525] The server receives input from users who log in to the system and upload a list of companies. Based on this list, the server accesses external APIs and internal databases to collect industry-specific information, payment status, and company information. This collected data is obtained in its raw, unprocessed state.

[0526] Step 2:

[0527] The server normalizes and centralizes the collected raw data. Specifically, it standardizes each data field according to the database schema and transforms data from multiple sources into an integrated dataset. This process formats the input data into a consistent format.

[0528] Step 3:

[0529] The server inputs normalized data into a machine learning algorithm to quantify payment risk. This risk score is calculated based on payment history, the company's financial status, industry trends, and other factors. The risk score is stored in the server's database and used to generate subsequent action plans.

[0530] Step 4:

[0531] The server generates customized action plans for each company based on the risk assessment results. For example, it creates plans that include frequent contact and warning emails for high-risk companies, and regular reminders for medium-risk companies. These action plans are automatically generated within the server.

[0532] Step 5:

[0533] The device receives an action plan from the server and schedules notifications. Scheduled notifications are delivered via the Twilio API as SMS or phone calls. The content and timing of each notification are automatically determined based on the action plan.

[0534] Step 6:

[0535] The server monitors the progress of the action plan in real time. It tracks data such as whether each notification was sent successfully and whether there was a response from the company, and stores this data in a database. This information is then used to inform the next follow-up actions.

[0536] Step 7:

[0537] The terminal displays the progress to the user and allows for additional follow-up actions as needed. Users can stay informed in real time through progress reports and take necessary corrections or additional actions. This information is fed back to the server and considered when generating the next action plan.

[0538] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0539] System Overview

[0540] The debt collection system of the present invention aims to improve the collection rate by evaluating a company's payment risk and generating and executing an appropriate action plan. It also incorporates an emotion engine to recognize the user's emotions and adaptively execute actions. This system collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[0541] Data collection

[0542] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0543] Data Analysis

[0544] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0545] Generating an action plan

[0546] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0547] Emotional engine integration

[0548] The server uses an emotion engine to recognize the user's emotions. For example, it detects the user's emotional state through text message and voice analysis. Based on the recognized emotions, the server adaptively changes its action plan. For example, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[0549] Automated notifications

[0550] The terminal schedules notifications based on the generated action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0551] Monitoring and follow-up on progress

[0552] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The terminal displays the progress to the user and provides an interface for setting additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0553] Specific example

[0554] For example, consider a case where company A in the food and beverage industry is included in the list. The server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of the person in charge at company A, and if it determines that they are under high stress, it modifies the follow-up content to be softer and more supportive.

[0555] In this way, the system of the present invention efficiently manages a company's payment risk, increases the debt collection rate by taking appropriate actions, and enables flexible responses that respond to the user's emotions.

[0556] The following describes the processing flow.

[0557] Step 1:

[0558] The user logs into the system and uploads a list of companies to be included in the collection process.

[0559] Step 2:

[0560] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0561] Step 3:

[0562] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0563] Step 4:

[0564] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0565] Step 5:

[0566] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0567] Step 6:

[0568] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0569] Step 7:

[0570] The server uses an emotion engine to recognize the emotions of company representatives and users from text messages and voice data.

[0571] Step 8:

[0572] The server adaptively modifies the action plan based on the recognized emotion data. For example, if the person in charge is experiencing stress, the communication method and message content will be changed to softer language.

[0573] Step 9:

[0574] The device notifies the user of the generated action plan and displays its contents.

[0575] Step 10:

[0576] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0577] Step 11:

[0578] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0579] Step 12:

[0580] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0581] Step 13:

[0582] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0583] Step 14:

[0584] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0585] By following this series of steps, it is possible not only to improve the efficiency and recovery rate of debt collection, but also to provide flexible responses that respond to the user's emotions.

[0586] (Example 2)

[0587] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0588] In modern businesses, efficient and effective debt collection is crucial. However, assessing payment risk and setting appropriate follow-up actions requires considerable effort and time. Furthermore, traditional systems struggle to respond flexibly to user emotions, often leading to stress. This results in challenges such as declining debt collection rates and increased burden on users.

[0589] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of companies using a machine learning algorithm with the collected information; means for generating customized action plans suitable for each company based on the evaluation results; means for recognizing the user's emotions using an emotion engine and adaptively changing the action plan; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan, saving the history of executed actions, and setting follow-up actions. This enables efficient and effective debt collection and allows for flexible responses that respond to the user's emotions.

[0590] "Industry-specific information" refers to information such as economic conditions, market trends, and industry-specific management indicators related to a particular industry.

[0591] "Payment status" refers to information about a company's past and present payment history, payment patterns, and whether or not there have been any payment delays.

[0592] "Company information" refers to basic company data such as name, address, date of establishment, financial status, and number of employees.

[0593] "Normalization" refers to the process of converting collected data into a consistent format and supplementing any missing fields.

[0594] "Unification" refers to the process of integrating information collected from different data sources and consolidating it into a single database.

[0595] A "machine learning algorithm" refers to an algorithm used to analyze data, identify patterns, and perform predictions and classifications.

[0596] "Payment risk" refers to the possibility that a particular company may delay or default on payments in the future.

[0597] A "customized action plan" refers to a plan that includes specific countermeasures, developed based on each company's payment risk assessment.

[0598] An "emotion engine" refers to a technology that analyzes a user's text and voice to identify their emotional state.

[0599] "Automated notifications" refers to the process of automatically sending notifications to users or companies based on a pre-configured action plan.

[0600] "Progress monitoring" refers to the process of tracking and evaluating in real time how well the set action plan is being implemented.

[0601] "Follow-up action" refers to taking additional steps or making contact based on the established action plan.

[0602] The debt collection system of the present invention is implemented with the following configuration. This system evaluates a company's payment risk, generates and executes an appropriate action plan, aims to improve the collection rate, and also has the function of recognizing the user's emotions by combining it with an emotion engine and performing actions adaptively.

[0603] Software and hardware to be used

[0604] This system is implemented using the following software and hardware:

[0605] 1. Server: Data collection, data normalization, execution of machine learning algorithms, generation of action plans, sentiment recognition by sentiment engine, monitoring and execution of notifications.

[0606] 2. Terminal: Provides the user interface, schedules and executes notifications, and displays progress.

[0607] 3. Software used:

[0608] External APIs for data collection: Examples include "Dun & Bradstreet API" and "Clearbit API".

[0609] Data analysis and machine learning: Python's "Pandas," "Scikit-learn," and "TensorFlow"

[0610] Sentiment recognition: Microsoft Azure's "Text Analytics" and IBM Watson's "Tone Analyzer"

[0611] Databases: PostgreSQL and MongoDB for centralized data storage.

[0612] System operation

[0613] To assess a company's payment risk, the server first requires the user to log in and upload a list of companies to be collected from. Based on the uploaded list, the server accesses external APIs and databases to collect necessary data such as basic company information, payment history, and industry trends.

[0614] The collected data is normalized and centralized on the server using the "Pandas" library, and then stored in "PostgreSQL" or "MongoDB". The server then uses "Scikit-learn" or "TensorFlow" to run machine learning algorithms to assess a company's payment risk. The assessment criteria include historical payment history, industry economic trends, and company information.

[0615] Based on risk assessment results from machine learning algorithms, the server generates customized action plans for each company. Specifically, it sets up frequent contact and payment reminders for high-risk companies and proposes flexible payment plans for low-risk companies. The generated action plans include contact methods (email, phone) and contact frequency.

[0616] Furthermore, the server uses emotion engines such as "Text Analytics" and "Tone Analyzer" to recognize the user's emotions and adaptively change the action plan. For example, if it detects that the user is stressed, it softens the notification content and emphasizes support.

[0617] Based on the generated action plan, the device schedules and executes notifications. Specifically, it sets up weekly follow-up calls and reminders for high-risk companies and executes them automatically. The server monitors the execution status of notifications and stores progress data in a database. The device displays the progress to the user and sets additional follow-up actions as needed.

[0618] Specific example

[0619] For example, if company A in the food and beverage industry is included in the list, the server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of company A's representative, and if it determines that the representative is stressed, it changes the content of the follow-up to be softer and more supportive.

[0620] Prompts for Generative AI Models

[0621] The following are specific examples of prompt statements to input into a generative AI model:

[0622] Prompt message 1:

[0623] "Assess the payment risk of the following company and generate an action plan. Company name: Sample Co., Ltd., Past payment history: Monthly late payments, Industry information: Food and beverage industry."

[0624] Prompt message 2:

[0625] "Please generate a follow-up message for when a user is experiencing stress. Current message: 'Payment deadline is approaching. Please take action as soon as possible.'"

[0626] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0627] Step 1:

[0628] The user logs into the system and uploads a list of companies to be included in the collection process. This means the user's input becomes a list of companies. The server receives this list and begins processing. Specifically, the user enters their authentication information on the system's login screen and clicks the "Login" button. Then, they click the "File Upload" button, select a company list file (e.g., CSV, Excel), and click the "Upload" button.

[0629] Step 2:

[0630] The server receives the uploaded company list and accesses external APIs and databases to collect the necessary information. The input is the company list, and the output is data such as basic company information, payment history, and industry trends. Specifically, the server uses the company ID to send requests to the "Dun & Bradstreet API" and "Clearbit API" to retrieve detailed company information. This information includes company name, address, establishment date, and payment history.

[0631] Step 3:

[0632] The server normalizes and centralizes the collected data. The input is multiple collected datasets, and the output is a normalized, consistent dataset. Specifically, the server uses the Python library "Pandas" to clean and standardize the data format. For example, it standardizes date formats and fills in missing data.

[0633] Step 4:

[0634] The server stores normalized data in a database. The input is a normalized dataset, and the output is the data stored in the database. Specifically, the server executes SQL queries to insert data into databases such as PostgreSQL or MongoDB.

[0635] Step 5:

[0636] The server uses normalized data to run machine learning algorithms and assess companies' payment risk. The input is data read from a database, and the output is a risk assessment score for each company. Specifically, the server uses the Random Forest algorithm from the "Scikit-learn" library to predict payment risk and calculate the score.

[0637] Step 6:

[0638] The server generates customized action plans for each company based on the risk assessment results. The input is the risk assessment score, and the output is the action plan. Specifically, the server sets up frequent contact and payment reminders for high-risk companies, and proposes flexible payment plans for low-risk companies. Specific contact methods (email, phone) and contact frequency are set for each company.

[0639] Step 7:

[0640] The server uses an emotion engine to recognize the user's emotions and adaptively modify the action plan. The input is the user's text messages or voice data, and the output is the modified action plan. Specifically, the server uses "Text Analytics" and "Tone Analyzer" to analyze the user's emotional state, and if it determines that the user is experiencing high levels of stress or anxiety, it softens the content of the action plan and emphasizes support.

[0641] Step 8:

[0642] The terminal schedules and executes notifications based on the generated action plan. The input is the action plan, and the output is the notification to the company. Specifically, the terminal schedules and automatically executes weekly follow-up calls and emails to high-risk companies.

[0643] Step 9:

[0644] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The input is the data of the actions performed, and the output is a progress report. Specifically, the server periodically updates the progress data and saves it to the database.

[0645] Step 10:

[0646] The terminal displays the progress to the user and allows for additional follow-up actions as needed. The input is a progress report, and the output is a notification to the user and additional follow-up actions. Specifically, the terminal graphically displays the progress through a dashboard screen and provides an interface where the user can configure the necessary actions.

[0647] (Application Example 2)

[0648] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0649] Current debt collection systems, while capable of assessing a company's payment risk and generating action plans based on that assessment, struggle to provide adaptive responses that take into account the customer's emotional state. This results in one-sided communication with customers, hindering improvements in collection rates. Furthermore, it's difficult for employees to assess the situation in real time while interacting with customers. Solving these problems is essential to providing a more effective debt collection system.

[0650] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating the payment risk of companies using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for recognizing the emotional state of the user; and means for adaptively changing the action plan based on the recognized emotional state. This enables flexible and adaptive responses that take into account the emotional state of the customer, thereby improving the collection rate and customer satisfaction.

[0651] "Industry-specific information" refers to information such as economic trends, competitive conditions, and business practices related to a specific industry.

[0652] "Payment status" refers to data such as past and present payment history, outstanding amounts, and frequency of payment delays.

[0653] "Company information" refers to information about a corporation's basic information, management status, financial condition, and creditworthiness.

[0654] "Payment risk" refers to a risk assessment calculated based on factors that affect a company's likelihood of fulfilling future payments.

[0655] An "action plan" is a plan that includes specific actions to be taken by the company based on the evaluation results, as well as the schedule for those actions.

[0656] "Automated notifications" is the process of automatically sending notifications at specific times based on a predetermined action plan.

[0657] "Follow-up actions" refer to additional actions or measures taken depending on the progress made after notification.

[0658] "Emotional state" refers to the user's current psychological and emotional condition, and is recognized through voice, text, facial expressions, etc.

[0659] "Adaptively modifying the action plan based on the perceived emotional state" means analyzing the user's emotions and flexibly changing the predetermined plan accordingly.

[0660] This invention relates to a debt collection system that assesses a company's payment risk and generates and executes an appropriate action plan. The system recognizes the user's emotional state and adaptively executes actions based on that state. Specifically, it collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[0661] System Overview

[0662] The server collects company information when users log in and upload a list of companies to be collected from. This information includes basic company information, past payment history, and industry economic trends. The collected data is normalized and centralized. Next, the server uses machine learning algorithms to assess the companies' payment risk. Based on this, high-risk companies will require more frequent follow-up, while low-risk companies will be offered flexible payment plans.

[0663] Emotional engine integration

[0664] The server uses an emotion engine to recognize the user's emotional state. For example, it can determine if a user is stressed through text message or voice analysis. Based on the recognized emotional state, it adaptively changes the action plan. For instance, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[0665] Automated notifications and progress monitoring

[0666] Based on the generated action plan, notifications are sent automatically. For example, high-risk companies may receive weekly follow-up calls or emails reminding them of late fees. The server monitors these notification processes and stores the execution history in a database. It also monitors progress in real time and sets up additional follow-up actions as needed.

[0667] Specific example

[0668] For example, consider a case where company B in the retail industry is included in the list. The server checks company B's payment history and discovers that it has a history of frequent payment delays. Based on this information, the server classifies company B as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. Additionally, during interactions with customers via smart glasses, the emotion engine analyzes the emotions of company B's representative, and if it determines that the representative is highly stressed, it changes the follow-up to be softer and more supportive. In this way, it is possible to increase the debt collection rate while simultaneously improving customer satisfaction.

[0669] Example prompt message

[0670] Let's say a customer enters a store, and an employee puts on smart glasses and begins serving them. The system detects the customer's stress level from their voice and determines they are high-risk based on their past payment history. Based on this information, the smart glasses display shows a message saying, "Use gentle language and emphasize support."

[0671] Prompt text for input example

[0672] Parameters:

[0673] customer_id: 12345

[0674] audio_input: "path / to / customer / voice / input.wav"

[0675] Query:

[0676] Please tell me about customer service methods that take into account the customer's payment risk level and emotional state.

[0677] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0678] Step 1: The user logs into the system and uploads a list of companies to be included in the collection process.

[0679] Input: User login information, list of companies to be included in the collection.

[0680] Output: A list of companies subject to recall is saved on the server.

[0681] Specific operation: The user logs in through the system interface and uploads a list of companies to be targeted for collection. The server receives this list and stores it in its internal database.

[0682] Step 2: The server collects corporate information.

[0683] Input: List of companies to be recalled

[0684] Output: Basic company information, payment history, industry economic trends

[0685] Specific operation: The server accesses external APIs and databases to collect basic information, past payment history, and industry economic trends of the target companies. The collected data is normalized and centralized.

[0686] Step 3: The server uses machine learning algorithms to assess the company's payment risk.

[0687] Input: Basic company information, payment history, industry economic trends

[0688] Output: Payment risk assessment results for each company

[0689] Specific operation: The server inputs the collected data into a machine learning algorithm to evaluate the payment risk of each company. The evaluation results are stored in a database.

[0690] Step 4: The server generates an action plan tailored to each company.

[0691] Input: Payment risk assessment results

[0692] Output: Action plans tailored to each company

[0693] Specific operation: Based on the payment risk assessment results, the server generates action plans that include frequent follow-up for high-risk companies and flexible payment plans for low-risk companies.

[0694] Step 5: The device automates notifications based on the generated action plan.

[0695] Input: Action Plan

[0696] Output: Automatically generated notifications (email, phone call, etc.)

[0697] Specific operation: The device automatically sends notifications via email or phone at the appropriate time based on the action plan.

[0698] Step 6: The server uses the emotion engine to recognize the user's emotional state.

[0699] Input: Text messages and audio data from the user.

[0700] Output: User's emotional state

[0701] Specific operation: The server uses an emotion engine to analyze text messages and voice data to detect the user's emotional state.

[0702] Step 7: The server adaptively modifies the action plan based on the recognized emotional state.

[0703] Input: User's emotional state

[0704] Output: Adaptively modified action plan

[0705] Specific operation: The server analyzes the user's emotional state and flexibly modifies the action plan accordingly. For example, if the user is feeling stressed, the follow-up content will be softened and emphasized to be more supportive.

[0706] Step 8: The server monitors the progress of the action plan and sets up follow-up actions.

[0707] Input: Action plan execution history

[0708] Output: Next follow-up action

[0709] Specific operation: The server saves the execution status of the action plan to a database and monitors it in real time. Based on the progress, it calculates the timing of the next follow-up and notifies the user.

[0710] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0711] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0712] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0713] [Third Embodiment]

[0714] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0715] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0716] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0717] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0718] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0719] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0720] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0721] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0722] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0723] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0724] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0725] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0726] System Overview

[0727] The debt collection system of the present invention aims to improve collection rates by evaluating a company's payment risk and generating and executing appropriate action plans. This system has the functionality to collect and analyze industry-specific information, payment status, company information, etc., and to automate follow-up actions.

[0728] Data collection

[0729] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0730] Data Analysis

[0731] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0732] Generating an action plan

[0733] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0734] Automated notifications

[0735] The terminal schedules notifications based on the generated action plan. For example, it can be configured to make weekly follow-up calls to high-risk companies and send overdue fee warning emails. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0736] Monitoring and follow-up on progress

[0737] The server monitors the progress of the action plan in real time and saves it to the database. The terminal displays the progress to the user and sets additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0738] Specific example

[0739] For example, if company A in the food and beverage industry is included in the list, the server will check company A's payment history and discover that it has a history of frequent payment delays. Based on this, the server will classify company A as "high risk" and generate an action plan that includes weekly follow-up calls and late fee warning emails. The terminal will automatically execute notifications based on this plan, and the server will track progress and provide reports to the user. If the payment situation does not improve, the server will readjust to increase the frequency of follow-ups.

[0740] In this way, the system of the present invention can efficiently manage a company's payment risk and increase the debt collection rate by taking appropriate action.

[0741] The following describes the processing flow.

[0742] Step 1:

[0743] The user logs into the system and uploads a list of companies to be included in the collection process.

[0744] Step 2:

[0745] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0746] Step 3:

[0747] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0748] Step 4:

[0749] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0750] Step 5:

[0751] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0752] Step 6:

[0753] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0754] Step 7:

[0755] The device notifies the user of the generated action plan and displays its contents.

[0756] Step 8:

[0757] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0758] Step 9:

[0759] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0760] Step 10:

[0761] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0762] Step 11:

[0763] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0764] Step 12:

[0765] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0766] In this way, by going through a series of steps, the efficiency and recovery rate of debt collection can be increased.

[0767] (Example 1)

[0768] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0769] Traditional debt collection systems have suffered from low collection rates due to a lack of means to accurately assess the payment risk of businesses and automatically execute appropriate collection actions. Furthermore, the process of prioritizing follow-up actions for each business is often done manually, resulting in inefficiencies.

[0770] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0771] In this invention, the server includes means for collecting industry-specific information, transaction status, and business information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of businesses using the normalized information; means for generating an action plan suitable for each business based on the evaluation results; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan and setting follow-up actions. This enables accurate evaluation of the payment risk of businesses, automation of the execution of optimal collection actions and follow-up of progress, and improves the efficiency of debt collection and the recovery rate.

[0772] "Industry-specific information" refers to information such as economic trends, market conditions, and performance indicators related to a specific industry.

[0773] "Transaction status" refers to the history of transactions a business has conducted in the past and the current status of those transactions, including payment history and delay information.

[0774] "Business information" refers to various data about a specific business, such as basic company information, financial status, and credit information.

[0775] "Normalization" is the process of maintaining data integrity and consistency by converting collected data into a unified format or standard.

[0776] "Centralization" is a process that improves data accessibility and management efficiency by integrating data collected from multiple sources and managing it centrally.

[0777] "Payment risk" is an assessment of the risk that a business will be able to pay its debts in time.

[0778] An "action plan" is a specific action plan customized for each business, including detailed measures such as payment reminders and follow-ups.

[0779] A "notification" is a message or warning that is automatically sent based on the generated action plan.

[0780] "Follow-up" is the process of monitoring the progress of an action plan that has already been implemented and taking additional actions as needed.

[0781] A "machine learning algorithm" is a mathematical model that automatically learns from data and is used to assess the payment risk of businesses.

[0782] The debt collection system of the present invention is a system that aims to improve the collection rate by collecting and analyzing industry-specific information, transaction status, and business information, assessing payment risk, and generating and executing appropriate action plans. This system includes the main elements of a server, terminals, and users.

[0783] Data collection

[0784] First, the user logs into the system and uploads a list of businesses to be included in the collection process. This list may include Excel or CSV files. The terminal temporarily saves this file and then transfers it to the server. Based on the uploaded list of companies, the server accesses external APIs and databases to collect necessary information. For example, it uses APIs from credit rating agencies to obtain credit information on businesses. It also collects industry economic trends and market conditions from other data sources. Next, the server normalizes and centralizes the collected data to maintain data integrity and consistency.

[0785] Data Analysis

[0786] The server runs a machine learning algorithm using normalized data. This algorithm assesses the payment risk of each business based on past payment history, industry information, and business information. The assessment results are stored as a payment risk score, and high-risk businesses are identified.

[0787] Generating an action plan

[0788] Based on the risk assessment results, the server automatically generates customized action plans for each business. For example, high-risk businesses will be assigned frequent contact and payment reminders, while low-risk businesses will be assigned flexible payment plans and notifications. The generated action plans include specific contact methods (email, phone) and contact frequency.

[0789] Automated notifications

[0790] The terminal schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to company A every Monday" or "send a late payment warning email." The server monitors and executes these notifications to ensure they are delivered appropriately.

[0791] Monitoring and follow-up on progress

[0792] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. The user can set additional follow-up actions as needed while viewing the progress. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Based on the progress, the server calculates the timing of the next action and sets the follow-up action.

[0793] Specific example

[0794] For example, if a restaurant business A is included in the list, the user uploads the list containing business A to the system. The server retrieves business A's payment history from a credit bureau's API and confirms that there have been frequent payment delays in the past. Based on this, the server classifies business A as "high risk." The server generates an action plan for business A, including weekly follow-up calls and emails warning about late fees. The terminal automatically schedules notifications based on this action plan, and the server tracks the progress and provides reports to the user.

[0795] Examples of prompts for generative AI models

[0796] The following prompt statements can be used as input to the generated AI model:

[0797] "Please generate a debt collection action plan based on Business A's past payment history. Business A belongs to the food and beverage industry and has experienced multiple payment delays in the past year."

[0798] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0799] Step 1:

[0800] The user logs into the system and enters their username and password. The server authenticates the entered information and allows the user to log in. In this process, the user's authentication information is provided as input, and the authentication result and user session are generated as output.

[0801] Step 2:

[0802] Users upload a list of businesses to be included in the collection process to the system. Common file formats include Excel and CSV. The terminal temporarily stores the uploaded file and then transfers it to the server. The uploaded file is provided as input, and that file is transferred to the server as output.

[0803] Step 3:

[0804] The server accesses external APIs and databases based on the uploaded list of businesses to collect necessary information. Specifically, when making API calls, it sends requests to the APIs of credit rating agencies to retrieve credit information, past transaction history, and other data. The input is the list of businesses, and the output is detailed information about each collected business.

[0805] Step 4:

[0806] The server normalizes and centralizes the collected data. Specifically, it converts data collected in different formats into a unified format. This process ensures data integrity and consistency. Raw data is taken as input, and normalized, centralized data is generated as output.

[0807] Step 5:

[0808] The server runs a machine learning algorithm using normalized data. This algorithm assesses payment risk based on past payment history, industry information, and business information. Specifically, data is input into the algorithm, and a risk score is output.

[0809] Step 6:

[0810] The server saves the results of machine learning algorithm execution to a database. The input is the risk assessment results, and the output is the risk score and a list of high-risk businesses.

[0811] Step 7:

[0812] The server automatically generates customized action plans for each business based on the risk assessment results. Specifically, it sets up frequent contact and payment reminders for high-risk businesses. The input is the risk assessment results, and the output is the action plan.

[0813] Step 8:

[0814] The device schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to business A every Monday" or "send a late payment warning email." The action plan is set as the input, and the notification schedule is set as the output.

[0815] Step 9:

[0816] The server executes scheduled notifications and monitors their execution status. Specifically, it automatically sends emails and phone reminders at scheduled times. The notification schedule is the input, and a notification execution log is generated as the output.

[0817] Step 10:

[0818] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. Progress data is generated as input, and progress reports for the user are generated as output.

[0819] Step 11:

[0820] Users monitor the progress and set additional follow-up actions as needed. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Progress is set as the input, and additional follow-up actions are set as the output.

[0821] Step 12:

[0822] The server calculates the timing of the next action based on the progress and sets up follow-up actions. Progress data is taken as input, and the next action plan is generated as output.

[0823] (Application Example 1)

[0824] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0825] Debt collection systems, designed to assess a company's payment risk and generate and implement effective action plans, currently suffer from problems such as insufficient real-time monitoring and follow-up, or limited automation, making efficient debt collection difficult. Furthermore, automatically generating and reliably executing notifications remains a challenge.

[0826] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0827] In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating a company's payment risk using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for acquiring company data using an external API; means for quantifying a company's payment risk using the collected data; and means for automatically generating the content and timing of notifications based on the risk assessment. This makes it possible to evaluate a company's payment risk in real time, generate an optimal action plan according to that risk, and reliably follow up.

[0828] "Industry-specific information" refers to data and statistics related to a specific industry.

[0829] "Payment status" refers to data that shows a company's past or present payment history and payment status.

[0830] "Company information" refers to detailed information about a company, including its basic data, financial status, and management status.

[0831] An "external API" refers to an interface for obtaining information in real time from external databases or services.

[0832] "Company data" refers to various types of data related to a company, such as basic company information, payment history, and credit information.

[0833] "Risk assessment" refers to the process of analyzing and evaluating a company's ability to pay its debts and its associated risks based on collected information.

[0834] An "action plan" refers to a plan that specifies the optimal response methods for each company based on the results of a risk assessment.

[0835] "Automated notifications" refers to a feature that automatically sends notifications based on an action plan generated by the system, without requiring manual action from the user.

[0836] "Follow-up" refers to the activity of monitoring the progress of an action plan after its implementation and taking additional actions as needed.

[0837] "Progress monitoring" refers to the process of tracking in real time whether the action plan is being implemented, and the subsequent responses and progress of the company.

[0838] "Quantification" refers to the process of converting evaluation results and data into specific numbers or scores.

[0839] "Notification content" refers to the main body of the warning or reminder message that is sent.

[0840] "Notification timing" refers to the timing and schedule of when notifications will be sent.

[0841] This invention relates to a system for evaluating a company's payment risk and generating and executing an appropriate action plan. To specifically implement this invention, various means of servers, terminals, and users work together.

[0842] System Overview

[0843] Data collection

[0844] The server receives user-uploaded lists and collects industry-specific information, payment status, and company information through external APIs and internal databases. External APIs retrieve information in real time, and the retrieved data is normalized and centralized within the server.

[0845] Data Analysis

[0846] The server runs machine learning algorithms using normalized data to assess each company's payment risk. The risk assessment is based on collected payment history, industry information, and company information. Specifically, the assessment process quantifies each company's payment risk and stores the risk assessment results in a database.

[0847] Generating an action plan

[0848] The server automatically generates customized action plans for each company based on the assessment results. These action plans include specific contact methods (e.g., email, phone) and contact frequency. Companies at high risk are configured to receive frequent contact and warning emails.

[0849] Automated notifications

[0850] The device schedules and automatically executes notifications based on the generated action plan. The server uses the Twilio API to send SMS and phone notifications, automatically generating the content and timing of the notifications. This ensures that the appropriate actions are taken.

[0851] Monitoring and follow-up on progress

[0852] The server monitors the progress of the action plan in real time and stores the data in a database. The terminal displays the progress to the user and allows them to set additional follow-up actions as needed. This ensures that appropriate tracking and response are always based on the latest information.

[0853] Specific example

[0854] For example, consider the case where the company list includes "CompanyA" and "CompanyB" as follows.

[0855] Python

[0856] company_list = ["CompanyA", "CompanyB"]

[0857] api_key = "your_api_key"

[0858] account_sid = "your_account_sid"

[0859] auth_token = "your_auth_token"

[0860] secu_collect = SecuCollect(api_key, account_sid, auth_token)

[0861] secu_collect.run(company_list)

[0862] Here, the server retrieves payment history and basic information for "CompanyA" and "CompanyB" and performs a risk assessment. Subsequently, it sends warning emails and SMS reminders to high-risk companies and monitors the progress of the action plan. This process is performed automatically by the system, and users can check the results in real time.

[0863] Example of a prompt

[0864] The following are examples of prompts used when utilizing generative AI models:

[0865] Develop an application that assesses a company's payment risks and automates appropriate response actions based on those assessments. This application should handle everything from data collection and risk assessment to action plan generation, automated notifications, and follow-up.

[0866] Based on this text, the system provides a set of functions to efficiently manage a company's payment risk and improve its debt collection rate.

[0867] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0868] Step 1:

[0869] The server receives input from users who log in to the system and upload a list of companies. Based on this list, the server accesses external APIs and internal databases to collect industry-specific information, payment status, and company information. This collected data is obtained in its raw, unprocessed state.

[0870] Step 2:

[0871] The server normalizes and centralizes the collected raw data. Specifically, it standardizes each data field according to the database schema and transforms data from multiple sources into an integrated dataset. This process formats the input data into a consistent format.

[0872] Step 3:

[0873] The server inputs normalized data into a machine learning algorithm to quantify payment risk. This risk score is calculated based on payment history, the company's financial status, industry trends, and other factors. The risk score is stored in the server's database and used to generate subsequent action plans.

[0874] Step 4:

[0875] The server generates customized action plans for each company based on the risk assessment results. For example, it creates plans that include frequent contact and warning emails for high-risk companies, and regular reminders for medium-risk companies. These action plans are automatically generated within the server.

[0876] Step 5:

[0877] The device receives an action plan from the server and schedules notifications. Scheduled notifications are delivered via the Twilio API as SMS or phone calls. The content and timing of each notification are automatically determined based on the action plan.

[0878] Step 6:

[0879] The server monitors the progress of the action plan in real time. It tracks data such as whether each notification was sent successfully and whether there was a response from the company, and stores this data in a database. This information is then used to inform the next follow-up actions.

[0880] Step 7:

[0881] The terminal displays the progress to the user and allows for additional follow-up actions as needed. Users can stay informed in real time through progress reports and take necessary corrections or additional actions. This information is fed back to the server and considered when generating the next action plan.

[0882] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0883] System Overview

[0884] The debt collection system of the present invention aims to improve the collection rate by evaluating a company's payment risk and generating and executing an appropriate action plan. It also incorporates an emotion engine to recognize the user's emotions and adaptively execute actions. This system collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[0885] Data collection

[0886] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[0887] Data Analysis

[0888] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[0889] Generating an action plan

[0890] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[0891] Emotional engine integration

[0892] The server uses an emotion engine to recognize the user's emotions. For example, it detects the user's emotional state through text message and voice analysis. Based on the recognized emotions, the server adaptively changes its action plan. For example, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[0893] Automated notifications

[0894] The terminal schedules notifications based on the generated action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees. The server monitors and executes these notifications to ensure they are delivered appropriately.

[0895] Monitoring and follow-up on progress

[0896] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The terminal displays the progress to the user and provides an interface for setting additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[0897] Specific example

[0898] For example, consider a case where company A in the food and beverage industry is included in the list. The server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of the person in charge at company A, and if it determines that they are under high stress, it modifies the follow-up content to be softer and more supportive.

[0899] In this way, the system of the present invention efficiently manages a company's payment risk, increases the debt collection rate by taking appropriate actions, and enables flexible responses that respond to the user's emotions.

[0900] The following describes the processing flow.

[0901] Step 1:

[0902] The user logs into the system and uploads a list of companies to be included in the collection process.

[0903] Step 2:

[0904] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[0905] Step 3:

[0906] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[0907] Step 4:

[0908] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[0909] Step 5:

[0910] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[0911] Step 6:

[0912] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[0913] Step 7:

[0914] The server uses an emotion engine to recognize the emotions of company representatives and users from text messages and voice data.

[0915] Step 8:

[0916] The server adaptively modifies the action plan based on the recognized emotion data. For example, if the person in charge is experiencing stress, the communication method and message content will be changed to softer language.

[0917] Step 9:

[0918] The device notifies the user of the generated action plan and displays its contents.

[0919] Step 10:

[0920] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[0921] Step 11:

[0922] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[0923] Step 12:

[0924] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[0925] Step 13:

[0926] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[0927] Step 14:

[0928] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[0929] By following this series of steps, it is possible not only to improve the efficiency and recovery rate of debt collection, but also to provide flexible responses that respond to the user's emotions.

[0930] (Example 2)

[0931] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0932] In modern businesses, efficient and effective debt collection is crucial. However, assessing payment risk and setting appropriate follow-up actions requires considerable effort and time. Furthermore, traditional systems struggle to respond flexibly to user emotions, often leading to stress. This results in challenges such as declining debt collection rates and increased burden on users.

[0933] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of companies using a machine learning algorithm with the collected information; means for generating customized action plans suitable for each company based on the evaluation results; means for recognizing the user's emotions using an emotion engine and adaptively changing the action plan; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan, saving the history of executed actions, and setting follow-up actions. This enables efficient and effective debt collection and allows for flexible responses that respond to the user's emotions.

[0934] "Industry-specific information" refers to information such as economic conditions, market trends, and industry-specific management indicators related to a particular industry.

[0935] "Payment status" refers to information about a company's past and present payment history, payment patterns, and whether or not there have been any payment delays.

[0936] "Company information" refers to basic company data such as name, address, date of establishment, financial status, and number of employees.

[0937] "Normalization" refers to the process of converting collected data into a consistent format and supplementing any missing fields.

[0938] "Unification" refers to the process of integrating information collected from different data sources and consolidating it into a single database.

[0939] A "machine learning algorithm" refers to an algorithm used to analyze data, identify patterns, and perform predictions and classifications.

[0940] "Payment risk" refers to the possibility that a particular company may delay or default on payments in the future.

[0941] A "customized action plan" refers to a plan that includes specific countermeasures, developed based on each company's payment risk assessment.

[0942] An "emotion engine" refers to a technology that analyzes a user's text and voice to identify their emotional state.

[0943] "Automated notifications" refers to the process of automatically sending notifications to users or companies based on a pre-configured action plan.

[0944] "Progress monitoring" refers to the process of tracking and evaluating in real time how well the set action plan is being implemented.

[0945] "Follow-up action" refers to taking additional steps or making contact based on the established action plan.

[0946] The debt collection system of the present invention is implemented with the following configuration. This system evaluates a company's payment risk, generates and executes an appropriate action plan, aims to improve the collection rate, and also has the function of recognizing the user's emotions by combining it with an emotion engine and performing actions adaptively.

[0947] Software and hardware to be used

[0948] This system is implemented using the following software and hardware:

[0949] 1. Server: Data collection, data normalization, execution of machine learning algorithms, generation of action plans, sentiment recognition by sentiment engine, monitoring and execution of notifications.

[0950] 2. Terminal: Provides the user interface, schedules and executes notifications, and displays progress.

[0951] 3. Software used:

[0952] External APIs for data collection: Examples include "Dun & Bradstreet API" and "Clearbit API".

[0953] Data analysis and machine learning: Python's "Pandas," "Scikit-learn," and "TensorFlow"

[0954] Sentiment recognition: Microsoft Azure's "Text Analytics" and IBM Watson's "Tone Analyzer"

[0955] Databases: PostgreSQL and MongoDB for centralized data storage.

[0956] System operation

[0957] To assess a company's payment risk, the server first requires the user to log in and upload a list of companies to be collected from. Based on the uploaded list, the server accesses external APIs and databases to collect necessary data such as basic company information, payment history, and industry trends.

[0958] The collected data is normalized and centralized on the server using the "Pandas" library, and then stored in "PostgreSQL" or "MongoDB". The server then uses "Scikit-learn" or "TensorFlow" to run machine learning algorithms to assess a company's payment risk. The assessment criteria include historical payment history, industry economic trends, and company information.

[0959] Based on risk assessment results from machine learning algorithms, the server generates customized action plans for each company. Specifically, it sets up frequent contact and payment reminders for high-risk companies and proposes flexible payment plans for low-risk companies. The generated action plans include contact methods (email, phone) and contact frequency.

[0960] Furthermore, the server uses emotion engines such as "Text Analytics" and "Tone Analyzer" to recognize the user's emotions and adaptively change the action plan. For example, if it detects that the user is stressed, it softens the notification content and emphasizes support.

[0961] Based on the generated action plan, the device schedules and executes notifications. Specifically, it sets up weekly follow-up calls and reminders for high-risk companies and executes them automatically. The server monitors the execution status of notifications and stores progress data in a database. The device displays the progress to the user and sets additional follow-up actions as needed.

[0962] Specific example

[0963] For example, if company A in the food and beverage industry is included in the list, the server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of company A's representative, and if it determines that the representative is stressed, it changes the content of the follow-up to be softer and more supportive.

[0964] Prompts for Generative AI Models

[0965] The following are specific examples of prompt statements to input into a generative AI model:

[0966] Prompt message 1:

[0967] "Assess the payment risk of the following company and generate an action plan. Company name: Sample Co., Ltd., Past payment history: Monthly late payments, Industry information: Food and beverage industry."

[0968] Prompt message 2:

[0969] "Please generate a follow-up message for when a user is experiencing stress. Current message: 'Payment deadline is approaching. Please take action as soon as possible.'"

[0970] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0971] Step 1:

[0972] The user logs into the system and uploads a list of companies to be included in the collection process. This means the user's input becomes a list of companies. The server receives this list and begins processing. Specifically, the user enters their authentication information on the system's login screen and clicks the "Login" button. Then, they click the "File Upload" button, select a company list file (e.g., CSV, Excel), and click the "Upload" button.

[0973] Step 2:

[0974] The server receives the uploaded company list and accesses external APIs and databases to collect the necessary information. The input is the company list, and the output is data such as basic company information, payment history, and industry trends. Specifically, the server uses the company ID to send requests to the "Dun & Bradstreet API" and "Clearbit API" to retrieve detailed company information. This information includes company name, address, establishment date, and payment history.

[0975] Step 3:

[0976] The server normalizes and centralizes the collected data. The input is multiple collected datasets, and the output is a normalized, consistent dataset. Specifically, the server uses the Python library "Pandas" to clean and standardize the data format. For example, it standardizes date formats and fills in missing data.

[0977] Step 4:

[0978] The server stores normalized data in a database. The input is a normalized dataset, and the output is the data stored in the database. Specifically, the server executes SQL queries to insert data into databases such as PostgreSQL or MongoDB.

[0979] Step 5:

[0980] The server uses normalized data to run machine learning algorithms and assess companies' payment risk. The input is data read from a database, and the output is a risk assessment score for each company. Specifically, the server uses the Random Forest algorithm from the "Scikit-learn" library to predict payment risk and calculate the score.

[0981] Step 6:

[0982] The server generates customized action plans for each company based on the risk assessment results. The input is the risk assessment score, and the output is the action plan. Specifically, the server sets up frequent contact and payment reminders for high-risk companies, and proposes flexible payment plans for low-risk companies. Specific contact methods (email, phone) and contact frequency are set for each company.

[0983] Step 7:

[0984] The server uses an emotion engine to recognize the user's emotions and adaptively modify the action plan. The input is the user's text messages or voice data, and the output is the modified action plan. Specifically, the server uses "Text Analytics" and "Tone Analyzer" to analyze the user's emotional state, and if it determines that the user is experiencing high levels of stress or anxiety, it softens the content of the action plan and emphasizes support.

[0985] Step 8:

[0986] The terminal schedules and executes notifications based on the generated action plan. The input is the action plan, and the output is the notification to the company. Specifically, the terminal schedules and automatically executes weekly follow-up calls and emails to high-risk companies.

[0987] Step 9:

[0988] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The input is the data of the actions performed, and the output is a progress report. Specifically, the server periodically updates the progress data and saves it to the database.

[0989] Step 10:

[0990] The terminal displays the progress to the user and allows for additional follow-up actions as needed. The input is a progress report, and the output is a notification to the user and additional follow-up actions. Specifically, the terminal graphically displays the progress through a dashboard screen and provides an interface where the user can configure the necessary actions.

[0991] (Application Example 2)

[0992] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0993] Current debt collection systems, while capable of assessing a company's payment risk and generating action plans based on that assessment, struggle to provide adaptive responses that take into account the customer's emotional state. This results in one-sided communication with customers, hindering improvements in collection rates. Furthermore, it's difficult for employees to assess the situation in real time while interacting with customers. Solving these problems is essential to providing a more effective debt collection system.

[0994] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating the payment risk of companies using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for recognizing the emotional state of the user; and means for adaptively changing the action plan based on the recognized emotional state. This enables flexible and adaptive responses that take into account the emotional state of the customer, thereby improving the collection rate and customer satisfaction.

[0995] "Industry-specific information" refers to information such as economic trends, competitive conditions, and business practices related to a specific industry.

[0996] "Payment status" refers to data such as past and present payment history, outstanding amounts, and frequency of payment delays.

[0997] "Company information" refers to information about a corporation's basic information, management status, financial condition, and creditworthiness.

[0998] "Payment risk" refers to a risk assessment calculated based on factors that affect a company's likelihood of fulfilling future payments.

[0999] An "action plan" is a plan that includes specific actions to be taken by the company based on the evaluation results, as well as the schedule for those actions.

[1000] "Automated notifications" is the process of automatically sending notifications at specific times based on a predetermined action plan.

[1001] "Follow-up actions" refer to additional actions or measures taken depending on the progress made after notification.

[1002] "Emotional state" refers to the user's current psychological and emotional condition, and is recognized through voice, text, facial expressions, etc.

[1003] "Adaptively modifying the action plan based on the perceived emotional state" means analyzing the user's emotions and flexibly changing the predetermined plan accordingly.

[1004] This invention relates to a debt collection system that assesses a company's payment risk and generates and executes an appropriate action plan. The system recognizes the user's emotional state and adaptively executes actions based on that state. Specifically, it collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[1005] System Overview

[1006] The server collects company information when users log in and upload a list of companies to be collected from. This information includes basic company information, past payment history, and industry economic trends. The collected data is normalized and centralized. Next, the server uses machine learning algorithms to assess the companies' payment risk. Based on this, high-risk companies will require more frequent follow-up, while low-risk companies will be offered flexible payment plans.

[1007] Emotional engine integration

[1008] The server uses an emotion engine to recognize the user's emotional state. For example, it can determine if a user is stressed through text message or voice analysis. Based on the recognized emotional state, it adaptively changes the action plan. For instance, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[1009] Automated notifications and progress monitoring

[1010] Based on the generated action plan, notifications are sent automatically. For example, high-risk companies may receive weekly follow-up calls or emails reminding them of late fees. The server monitors these notification processes and stores the execution history in a database. It also monitors progress in real time and sets up additional follow-up actions as needed.

[1011] Specific example

[1012] For example, consider a case where company B in the retail industry is included in the list. The server checks company B's payment history and discovers that it has a history of frequent payment delays. Based on this information, the server classifies company B as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. Additionally, during interactions with customers via smart glasses, the emotion engine analyzes the emotions of company B's representative, and if it determines that the representative is highly stressed, it changes the follow-up to be softer and more supportive. In this way, it is possible to increase the debt collection rate while simultaneously improving customer satisfaction.

[1013] Example prompt message

[1014] Let's say a customer enters a store, and an employee puts on smart glasses and begins serving them. The system detects the customer's stress level from their voice and determines they are high-risk based on their past payment history. Based on this information, the smart glasses display shows a message saying, "Use gentle language and emphasize support."

[1015] Prompt text for input example

[1016] Parameters:

[1017] customer_id: 12345

[1018] audio_input: "path / to / customer / voice / input.wav"

[1019] Query:

[1020] Please tell me about customer service methods that take into account the customer's payment risk level and emotional state.

[1021] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1022] Step 1: The user logs into the system and uploads a list of companies to be included in the collection process.

[1023] Input: User login information, list of companies to be included in the collection.

[1024] Output: A list of companies subject to recall is saved on the server.

[1025] Specific operation: The user logs in through the system interface and uploads a list of companies to be targeted for collection. The server receives this list and stores it in its internal database.

[1026] Step 2: The server collects corporate information.

[1027] Input: List of companies to be recalled

[1028] Output: Basic company information, payment history, industry economic trends

[1029] Specific operation: The server accesses external APIs and databases to collect basic information, past payment history, and industry economic trends of the target companies. The collected data is normalized and centralized.

[1030] Step 3: The server uses machine learning algorithms to assess the company's payment risk.

[1031] Input: Basic company information, payment history, industry economic trends

[1032] Output: Payment risk assessment results for each company

[1033] Specific operation: The server inputs the collected data into a machine learning algorithm to evaluate the payment risk of each company. The evaluation results are stored in a database.

[1034] Step 4: The server generates an action plan tailored to each company.

[1035] Input: Payment risk assessment results

[1036] Output: Action plans tailored to each company

[1037] Specific operation: Based on the payment risk assessment results, the server generates action plans that include frequent follow-up for high-risk companies and flexible payment plans for low-risk companies.

[1038] Step 5: The device automates notifications based on the generated action plan.

[1039] Input: Action Plan

[1040] Output: Automatically generated notifications (email, phone call, etc.)

[1041] Specific operation: The device automatically sends notifications via email or phone at the appropriate time based on the action plan.

[1042] Step 6: The server uses the emotion engine to recognize the user's emotional state.

[1043] Input: Text messages and audio data from the user.

[1044] Output: User's emotional state

[1045] Specific operation: The server uses an emotion engine to analyze text messages and voice data to detect the user's emotional state.

[1046] Step 7: The server adaptively modifies the action plan based on the recognized emotional state.

[1047] Input: User's emotional state

[1048] Output: Adaptively modified action plan

[1049] Specific operation: The server analyzes the user's emotional state and flexibly modifies the action plan accordingly. For example, if the user is feeling stressed, the follow-up content will be softened and emphasized to be more supportive.

[1050] Step 8: The server monitors the progress of the action plan and sets up follow-up actions.

[1051] Input: Action plan execution history

[1052] Output: Next follow-up action

[1053] Specific operation: The server saves the execution status of the action plan to a database and monitors it in real time. Based on the progress, it calculates the timing of the next follow-up and notifies the user.

[1054] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1055] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1056] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[1057] [Fourth Embodiment]

[1058] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[1059] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1060] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1061] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[1062] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[1063] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[1064] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[1065] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[1066] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[1067] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1068] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1069] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[1070] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1071] System Overview

[1072] The debt collection system of the present invention aims to improve collection rates by evaluating a company's payment risk and generating and executing appropriate action plans. This system has the functionality to collect and analyze industry-specific information, payment status, company information, etc., and to automate follow-up actions.

[1073] Data collection

[1074] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[1075] Data Analysis

[1076] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[1077] Generating an action plan

[1078] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[1079] Automated notifications

[1080] The terminal schedules notifications based on the generated action plan. For example, it can be configured to make weekly follow-up calls to high-risk companies and send overdue fee warning emails. The server monitors and executes these notifications to ensure they are delivered appropriately.

[1081] Monitoring and follow-up on progress

[1082] The server monitors the progress of the action plan in real time and saves it to the database. The terminal displays the progress to the user and sets additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[1083] Specific example

[1084] For example, if company A in the food and beverage industry is included in the list, the server will check company A's payment history and discover that it has a history of frequent payment delays. Based on this, the server will classify company A as "high risk" and generate an action plan that includes weekly follow-up calls and late fee warning emails. The terminal will automatically execute notifications based on this plan, and the server will track progress and provide reports to the user. If the payment situation does not improve, the server will readjust to increase the frequency of follow-ups.

[1085] In this way, the system of the present invention can efficiently manage a company's payment risk and increase the debt collection rate by taking appropriate action.

[1086] The following describes the processing flow.

[1087] Step 1:

[1088] The user logs into the system and uploads a list of companies to be included in the collection process.

[1089] Step 2:

[1090] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[1091] Step 3:

[1092] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[1093] Step 4:

[1094] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[1095] Step 5:

[1096] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[1097] Step 6:

[1098] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[1099] Step 7:

[1100] The device notifies the user of the generated action plan and displays its contents.

[1101] Step 8:

[1102] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[1103] Step 9:

[1104] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[1105] Step 10:

[1106] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[1107] Step 11:

[1108] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[1109] Step 12:

[1110] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[1111] In this way, by going through a series of steps, the efficiency and recovery rate of debt collection can be increased.

[1112] (Example 1)

[1113] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1114] Traditional debt collection systems have suffered from low collection rates due to a lack of means to accurately assess the payment risk of businesses and automatically execute appropriate collection actions. Furthermore, the process of prioritizing follow-up actions for each business is often done manually, resulting in inefficiencies.

[1115] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[1116] In this invention, the server includes means for collecting industry-specific information, transaction status, and business information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of businesses using the normalized information; means for generating an action plan suitable for each business based on the evaluation results; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan and setting follow-up actions. This enables accurate evaluation of the payment risk of businesses, automation of the execution of optimal collection actions and follow-up of progress, and improves the efficiency of debt collection and the recovery rate.

[1117] "Industry-specific information" refers to information such as economic trends, market conditions, and performance indicators related to a specific industry.

[1118] "Transaction status" refers to the history of transactions a business has conducted in the past and the current status of those transactions, including payment history and delay information.

[1119] "Business information" refers to various data about a specific business, such as basic company information, financial status, and credit information.

[1120] "Normalization" is the process of maintaining data integrity and consistency by converting collected data into a unified format or standard.

[1121] "Centralization" is a process that improves data accessibility and management efficiency by integrating data collected from multiple sources and managing it centrally.

[1122] "Payment risk" is an assessment of the risk that a business will be able to pay its debts in time.

[1123] An "action plan" is a specific action plan customized for each business, including detailed measures such as payment reminders and follow-ups.

[1124] A "notification" is a message or warning that is automatically sent based on the generated action plan.

[1125] "Follow-up" is the process of monitoring the progress of an action plan that has already been implemented and taking additional actions as needed.

[1126] A "machine learning algorithm" is a mathematical model that automatically learns from data and is used to assess the payment risk of businesses.

[1127] The debt collection system of the present invention is a system that aims to improve the collection rate by collecting and analyzing industry-specific information, transaction status, and business information, assessing payment risk, and generating and executing appropriate action plans. This system includes the main elements of a server, terminals, and users.

[1128] Data collection

[1129] First, the user logs into the system and uploads a list of businesses to be included in the collection process. This list may include Excel or CSV files. The terminal temporarily saves this file and then transfers it to the server. Based on the uploaded list of companies, the server accesses external APIs and databases to collect necessary information. For example, it uses APIs from credit rating agencies to obtain credit information on businesses. It also collects industry economic trends and market conditions from other data sources. Next, the server normalizes and centralizes the collected data to maintain data integrity and consistency.

[1130] Data Analysis

[1131] The server runs a machine learning algorithm using normalized data. This algorithm assesses the payment risk of each business based on past payment history, industry information, and business information. The assessment results are stored as a payment risk score, and high-risk businesses are identified.

[1132] Generating an action plan

[1133] Based on the risk assessment results, the server automatically generates customized action plans for each business. For example, high-risk businesses will be assigned frequent contact and payment reminders, while low-risk businesses will be assigned flexible payment plans and notifications. The generated action plans include specific contact methods (email, phone) and contact frequency.

[1134] Automated notifications

[1135] The terminal schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to company A every Monday" or "send a late payment warning email." The server monitors and executes these notifications to ensure they are delivered appropriately.

[1136] Monitoring and follow-up on progress

[1137] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. The user can set additional follow-up actions as needed while viewing the progress. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Based on the progress, the server calculates the timing of the next action and sets the follow-up action.

[1138] Specific example

[1139] For example, if a restaurant business A is included in the list, the user uploads the list containing business A to the system. The server retrieves business A's payment history from a credit bureau's API and confirms that there have been frequent payment delays in the past. Based on this, the server classifies business A as "high risk." The server generates an action plan for business A, including weekly follow-up calls and emails warning about late fees. The terminal automatically schedules notifications based on this action plan, and the server tracks the progress and provides reports to the user.

[1140] Examples of prompts for generative AI models

[1141] The following prompt statements can be used as input to the generated AI model:

[1142] "Please generate a debt collection action plan based on Business A's past payment history. Business A belongs to the food and beverage industry and has experienced multiple payment delays in the past year."

[1143] The flow of the specific processing in Example 1 will be explained using Figure 11.

[1144] Step 1:

[1145] The user logs into the system and enters their username and password. The server authenticates the entered information and allows the user to log in. In this process, the user's authentication information is provided as input, and the authentication result and user session are generated as output.

[1146] Step 2:

[1147] Users upload a list of businesses to be included in the collection process to the system. Common file formats include Excel and CSV. The terminal temporarily stores the uploaded file and then transfers it to the server. The uploaded file is provided as input, and that file is transferred to the server as output.

[1148] Step 3:

[1149] The server accesses external APIs and databases based on the uploaded list of businesses to collect necessary information. Specifically, when making API calls, it sends requests to the APIs of credit rating agencies to retrieve credit information, past transaction history, and other data. The input is the list of businesses, and the output is detailed information about each collected business.

[1150] Step 4:

[1151] The server normalizes and centralizes the collected data. Specifically, it converts data collected in different formats into a unified format. This process ensures data integrity and consistency. Raw data is taken as input, and normalized, centralized data is generated as output.

[1152] Step 5:

[1153] The server runs a machine learning algorithm using normalized data. This algorithm assesses payment risk based on past payment history, industry information, and business information. Specifically, data is input into the algorithm, and a risk score is output.

[1154] Step 6:

[1155] The server saves the results of machine learning algorithm execution to a database. The input is the risk assessment results, and the output is the risk score and a list of high-risk businesses.

[1156] Step 7:

[1157] The server automatically generates customized action plans for each business based on the risk assessment results. Specifically, it sets up frequent contact and payment reminders for high-risk businesses. The input is the risk assessment results, and the output is the action plan.

[1158] Step 8:

[1159] The device schedules notifications based on the generated action plan. For example, it sets specific tasks such as "make a follow-up phone call to business A every Monday" or "send a late payment warning email." The action plan is set as the input, and the notification schedule is set as the output.

[1160] Step 9:

[1161] The server executes scheduled notifications and monitors their execution status. Specifically, it automatically sends emails and phone reminders at scheduled times. The notification schedule is the input, and a notification execution log is generated as the output.

[1162] Step 10:

[1163] The server monitors the progress of the action plan in real time and stores it in a database. The terminal displays the progress to the user, allowing them to see which actions have been taken for which businesses. Progress data is generated as input, and progress reports for the user are generated as output.

[1164] Step 11:

[1165] Users monitor the progress and set additional follow-up actions as needed. For example, if a particular business does not respond to a reminder, they can set more frequent contact. Progress is set as the input, and additional follow-up actions are set as the output.

[1166] Step 12:

[1167] The server calculates the timing of the next action based on the progress and sets up follow-up actions. Progress data is taken as input, and the next action plan is generated as output.

[1168] (Application Example 1)

[1169] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1170] Debt collection systems, designed to assess a company's payment risk and generate and implement effective action plans, currently suffer from problems such as insufficient real-time monitoring and follow-up, or limited automation, making efficient debt collection difficult. Furthermore, automatically generating and reliably executing notifications remains a challenge.

[1171] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[1172] In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating a company's payment risk using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for acquiring company data using an external API; means for quantifying a company's payment risk using the collected data; and means for automatically generating the content and timing of notifications based on the risk assessment. This makes it possible to evaluate a company's payment risk in real time, generate an optimal action plan according to that risk, and reliably follow up.

[1173] "Industry-specific information" refers to data and statistics related to a specific industry.

[1174] "Payment status" refers to data that shows a company's past or present payment history and payment status.

[1175] "Company information" refers to detailed information about a company, including its basic data, financial status, and management status.

[1176] An "external API" refers to an interface for obtaining information in real time from external databases or services.

[1177] "Company data" refers to various types of data related to a company, such as basic company information, payment history, and credit information.

[1178] "Risk assessment" refers to the process of analyzing and evaluating a company's ability to pay its debts and its associated risks based on collected information.

[1179] An "action plan" refers to a plan that specifies the optimal response methods for each company based on the results of a risk assessment.

[1180] "Automated notifications" refers to a feature that automatically sends notifications based on an action plan generated by the system, without requiring manual action from the user.

[1181] "Follow-up" refers to the activity of monitoring the progress of an action plan after its implementation and taking additional actions as needed.

[1182] "Progress monitoring" refers to the process of tracking in real time whether the action plan is being implemented, and the subsequent responses and progress of the company.

[1183] "Quantification" refers to the process of converting evaluation results and data into specific numbers or scores.

[1184] "Notification content" refers to the main body of the warning or reminder message that is sent.

[1185] "Notification timing" refers to the timing and schedule of when notifications will be sent.

[1186] This invention relates to a system for evaluating a company's payment risk and generating and executing an appropriate action plan. To specifically implement this invention, various means of servers, terminals, and users work together.

[1187] System Overview

[1188] Data collection

[1189] The server receives user-uploaded lists and collects industry-specific information, payment status, and company information through external APIs and internal databases. External APIs retrieve information in real time, and the retrieved data is normalized and centralized within the server.

[1190] Data Analysis

[1191] The server runs machine learning algorithms using normalized data to assess each company's payment risk. The risk assessment is based on collected payment history, industry information, and company information. Specifically, the assessment process quantifies each company's payment risk and stores the risk assessment results in a database.

[1192] Generating an action plan

[1193] The server automatically generates customized action plans for each company based on the assessment results. These action plans include specific contact methods (e.g., email, phone) and contact frequency. Companies at high risk are configured to receive frequent contact and warning emails.

[1194] Automated notifications

[1195] The device schedules and automatically executes notifications based on the generated action plan. The server uses the Twilio API to send SMS and phone notifications, automatically generating the content and timing of the notifications. This ensures that the appropriate actions are taken.

[1196] Monitoring and follow-up on progress

[1197] The server monitors the progress of the action plan in real time and stores the data in a database. The terminal displays the progress to the user and allows them to set additional follow-up actions as needed. This ensures that appropriate tracking and response are always based on the latest information.

[1198] Specific example

[1199] For example, consider the case where the company list includes "CompanyA" and "CompanyB" as follows.

[1200] Python

[1201] company_list = ["CompanyA", "CompanyB"]

[1202] api_key = "your_api_key"

[1203] account_sid = "your_account_sid"

[1204] auth_token = "your_auth_token"

[1205] secu_collect = SecuCollect(api_key, account_sid, auth_token)

[1206] secu_collect.run(company_list)

[1207] Here, the server retrieves payment history and basic information for "CompanyA" and "CompanyB" and performs a risk assessment. Subsequently, it sends warning emails and SMS reminders to high-risk companies and monitors the progress of the action plan. This process is performed automatically by the system, and users can check the results in real time.

[1208] Example of a prompt

[1209] The following are examples of prompts used when utilizing generative AI models:

[1210] Develop an application that assesses a company's payment risks and automates appropriate response actions based on those assessments. This application should handle everything from data collection and risk assessment to action plan generation, automated notifications, and follow-up.

[1211] Based on this text, the system provides a set of functions to efficiently manage a company's payment risk and improve its debt collection rate.

[1212] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[1213] Step 1:

[1214] The server receives input from users who log in to the system and upload a list of companies. Based on this list, the server accesses external APIs and internal databases to collect industry-specific information, payment status, and company information. This collected data is obtained in its raw, unprocessed state.

[1215] Step 2:

[1216] The server normalizes and centralizes the collected raw data. Specifically, it standardizes each data field according to the database schema and transforms data from multiple sources into an integrated dataset. This process formats the input data into a consistent format.

[1217] Step 3:

[1218] The server inputs normalized data into a machine learning algorithm to quantify payment risk. This risk score is calculated based on payment history, the company's financial status, industry trends, and other factors. The risk score is stored in the server's database and used to generate subsequent action plans.

[1219] Step 4:

[1220] The server generates customized action plans for each company based on the risk assessment results. For example, it creates plans that include frequent contact and warning emails for high-risk companies, and regular reminders for medium-risk companies. These action plans are automatically generated within the server.

[1221] Step 5:

[1222] The device receives an action plan from the server and schedules notifications. Scheduled notifications are delivered via the Twilio API as SMS or phone calls. The content and timing of each notification are automatically determined based on the action plan.

[1223] Step 6:

[1224] The server monitors the progress of the action plan in real time. It tracks data such as whether each notification was sent successfully and whether there was a response from the company, and stores this data in a database. This information is then used to inform the next follow-up actions.

[1225] Step 7:

[1226] The terminal displays the progress to the user and allows for additional follow-up actions as needed. Users can stay informed in real time through progress reports and take necessary corrections or additional actions. This information is fed back to the server and considered when generating the next action plan.

[1227] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[1228] System Overview

[1229] The debt collection system of the present invention aims to improve the collection rate by evaluating a company's payment risk and generating and executing an appropriate action plan. It also incorporates an emotion engine to recognize the user's emotions and adaptively execute actions. This system collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[1230] Data collection

[1231] Users log into the system and upload a list of companies to be included in the collection process. The server receives the uploaded list of companies and accesses external APIs and databases to collect necessary information. Specifically, this includes basic company information, past payment history, and industry economic trends. The server normalizes and centralizes the collected data.

[1232] Data Analysis

[1233] The server uses normalized data to run machine learning algorithms and assess each company's payment risk. The assessment is based on payment history, industry information, and company information. The server stores the assessment results and identifies high-risk companies, allowing for priority action on which companies should be addressed.

[1234] Generating an action plan

[1235] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent contact and payment reminders, while low-risk companies will receive flexible payment plans and notifications. The generated action plans will include specific contact methods (e.g., email, phone) and contact frequency.

[1236] Emotional engine integration

[1237] The server uses an emotion engine to recognize the user's emotions. For example, it detects the user's emotional state through text message and voice analysis. Based on the recognized emotions, the server adaptively changes its action plan. For example, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[1238] Automated notifications

[1239] The terminal schedules notifications based on the generated action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees. The server monitors and executes these notifications to ensure they are delivered appropriately.

[1240] Monitoring and follow-up on progress

[1241] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The terminal displays the progress to the user and provides an interface for setting additional follow-up actions as needed. The server calculates the timing of the next action based on the progress and sets the follow-up action.

[1242] Specific example

[1243] For example, consider a case where company A in the food and beverage industry is included in the list. The server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of the person in charge at company A, and if it determines that they are under high stress, it modifies the follow-up content to be softer and more supportive.

[1244] In this way, the system of the present invention efficiently manages a company's payment risk, increases the debt collection rate by taking appropriate actions, and enables flexible responses that respond to the user's emotions.

[1245] The following describes the processing flow.

[1246] Step 1:

[1247] The user logs into the system and uploads a list of companies to be included in the collection process.

[1248] Step 2:

[1249] The server receives the uploaded list of companies and collects industry-specific information, payment status, and company information for each company in the list from external APIs and databases.

[1250] Step 3:

[1251] The server normalizes the collected data, unifies different formats and units, and stores it as a unified dataset.

[1252] Step 4:

[1253] The server inputs normalized data into a machine learning algorithm to assess each company's payment risk. The assessment is performed based on past payment history and industry information.

[1254] Step 5:

[1255] The server stores the results of the risk assessment in a database and identifies companies with a high payment risk.

[1256] Step 6:

[1257] The server automatically generates customized action plans for each company based on the risk assessment results. High-risk companies will receive frequent communication and payment reminders, while low-risk companies will be offered flexible payment plans.

[1258] Step 7:

[1259] The server uses an emotion engine to recognize the emotions of company representatives and users from text messages and voice data.

[1260] Step 8:

[1261] The server adaptively modifies the action plan based on the recognized emotion data. For example, if the person in charge is experiencing stress, the communication method and message content will be changed to softer language.

[1262] Step 9:

[1263] The device notifies the user of the generated action plan and displays its contents.

[1264] Step 10:

[1265] The device schedules and automatically executes notifications based on the action plan. For example, it might make weekly follow-up calls to high-risk companies and send emails reminding them of late fees.

[1266] Step 11:

[1267] The server monitors the progress of the action plan in real time and saves a history of the actions taken to a database.

[1268] Step 12:

[1269] The device provides an interface that displays the progress to the user and allows them to set additional follow-up actions as needed.

[1270] Step 13:

[1271] The server automatically calculates the timing of the next action based on the progress and schedules follow-up actions.

[1272] Step 14:

[1273] The device will send a follow-up notification to the user and, if there is no improvement in progress, will suggest additional actions.

[1274] By following this series of steps, it is possible not only to improve the efficiency and recovery rate of debt collection, but also to provide flexible responses that respond to the user's emotions.

[1275] (Example 2)

[1276] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1277] In modern businesses, efficient and effective debt collection is crucial. However, assessing payment risk and setting appropriate follow-up actions requires considerable effort and time. Furthermore, traditional systems struggle to respond flexibly to user emotions, often leading to stress. This results in challenges such as declining debt collection rates and increased burden on users.

[1278] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for normalizing and centralizing the collected information; means for evaluating the payment risk of companies using a machine learning algorithm with the collected information; means for generating customized action plans suitable for each company based on the evaluation results; means for recognizing the user's emotions using an emotion engine and adaptively changing the action plan; means for automating notifications based on the generated action plan; and means for monitoring the progress of the action plan, saving the history of executed actions, and setting follow-up actions. This enables efficient and effective debt collection and allows for flexible responses that respond to the user's emotions.

[1279] "Industry-specific information" refers to information such as economic conditions, market trends, and industry-specific management indicators related to a particular industry.

[1280] "Payment status" refers to information about a company's past and present payment history, payment patterns, and whether or not there have been any payment delays.

[1281] "Company information" refers to basic company data such as name, address, date of establishment, financial status, and number of employees.

[1282] "Normalization" refers to the process of converting collected data into a consistent format and supplementing any missing fields.

[1283] "Unification" refers to the process of integrating information collected from different data sources and consolidating it into a single database.

[1284] A "machine learning algorithm" refers to an algorithm used to analyze data, identify patterns, and perform predictions and classifications.

[1285] "Payment risk" refers to the possibility that a particular company may delay or default on payments in the future.

[1286] A "customized action plan" refers to a plan that includes specific countermeasures, developed based on each company's payment risk assessment.

[1287] An "emotion engine" refers to a technology that analyzes a user's text and voice to identify their emotional state.

[1288] "Automated notifications" refers to the process of automatically sending notifications to users or companies based on a pre-configured action plan.

[1289] "Progress monitoring" refers to the process of tracking and evaluating in real time how well the set action plan is being implemented.

[1290] "Follow-up action" refers to taking additional steps or making contact based on the established action plan.

[1291] The debt collection system of the present invention is implemented with the following configuration. This system evaluates a company's payment risk, generates and executes an appropriate action plan, aims to improve the collection rate, and also has the function of recognizing the user's emotions by combining it with an emotion engine and performing actions adaptively.

[1292] Software and hardware to be used

[1293] This system is implemented using the following software and hardware:

[1294] 1. Server: Data collection, data normalization, execution of machine learning algorithms, generation of action plans, sentiment recognition by sentiment engine, monitoring and execution of notifications.

[1295] 2. Terminal: Provides the user interface, schedules and executes notifications, and displays progress.

[1296] 3. Software used:

[1297] External APIs for data collection: Examples include "Dun & Bradstreet API" and "Clearbit API".

[1298] Data analysis and machine learning: Python's "Pandas," "Scikit-learn," and "TensorFlow"

[1299] Sentiment recognition: Microsoft Azure's "Text Analytics" and IBM Watson's "Tone Analyzer"

[1300] Databases: PostgreSQL and MongoDB for centralized data storage.

[1301] System operation

[1302] To assess a company's payment risk, the server first requires the user to log in and upload a list of companies to be collected from. Based on the uploaded list, the server accesses external APIs and databases to collect necessary data such as basic company information, payment history, and industry trends.

[1303] The collected data is normalized and centralized on the server using the "Pandas" library, and then stored in "PostgreSQL" or "MongoDB". The server then uses "Scikit-learn" or "TensorFlow" to run machine learning algorithms to assess a company's payment risk. The assessment criteria include historical payment history, industry economic trends, and company information.

[1304] Based on risk assessment results from machine learning algorithms, the server generates customized action plans for each company. Specifically, it sets up frequent contact and payment reminders for high-risk companies and proposes flexible payment plans for low-risk companies. The generated action plans include contact methods (email, phone) and contact frequency.

[1305] Furthermore, the server uses emotion engines such as "Text Analytics" and "Tone Analyzer" to recognize the user's emotions and adaptively change the action plan. For example, if it detects that the user is stressed, it softens the notification content and emphasizes support.

[1306] Based on the generated action plan, the device schedules and executes notifications. Specifically, it sets up weekly follow-up calls and reminders for high-risk companies and executes them automatically. The server monitors the execution status of notifications and stores progress data in a database. The device displays the progress to the user and sets additional follow-up actions as needed.

[1307] Specific example

[1308] For example, if company A in the food and beverage industry is included in the list, the server checks company A's payment history and discovers that it has a history of frequent payment delays. Based on this, the server classifies company A as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. The terminal automatically executes notifications based on this plan, and the server tracks the progress and provides reports to the user. At the same time, the emotion engine analyzes the emotions of company A's representative, and if it determines that the representative is stressed, it changes the content of the follow-up to be softer and more supportive.

[1309] Prompts for Generative AI Models

[1310] The following are specific examples of prompt statements to input into a generative AI model:

[1311] Prompt message 1:

[1312] "Assess the payment risk of the following company and generate an action plan. Company name: Sample Co., Ltd., Past payment history: Monthly late payments, Industry information: Food and beverage industry."

[1313] Prompt message 2:

[1314] "Please generate a follow-up message for when a user is experiencing stress. Current message: 'Payment deadline is approaching. Please take action as soon as possible.'"

[1315] The flow of the specific processing in Example 2 will be explained using Figure 13.

[1316] Step 1:

[1317] The user logs into the system and uploads a list of companies to be included in the collection process. This means the user's input becomes a list of companies. The server receives this list and begins processing. Specifically, the user enters their authentication information on the system's login screen and clicks the "Login" button. Then, they click the "File Upload" button, select a company list file (e.g., CSV, Excel), and click the "Upload" button.

[1318] Step 2:

[1319] The server receives the uploaded company list and accesses external APIs and databases to collect the necessary information. The input is the company list, and the output is data such as basic company information, payment history, and industry trends. Specifically, the server uses the company ID to send requests to the "Dun & Bradstreet API" and "Clearbit API" to retrieve detailed company information. This information includes company name, address, establishment date, and payment history.

[1320] Step 3:

[1321] The server normalizes and centralizes the collected data. The input is multiple collected datasets, and the output is a normalized, consistent dataset. Specifically, the server uses the Python library "Pandas" to clean and standardize the data format. For example, it standardizes date formats and fills in missing data.

[1322] Step 4:

[1323] The server stores normalized data in a database. The input is a normalized dataset, and the output is the data stored in the database. Specifically, the server executes SQL queries to insert data into databases such as PostgreSQL or MongoDB.

[1324] Step 5:

[1325] The server uses normalized data to run machine learning algorithms and assess companies' payment risk. The input is data read from a database, and the output is a risk assessment score for each company. Specifically, the server uses the Random Forest algorithm from the "Scikit-learn" library to predict payment risk and calculate the score.

[1326] Step 6:

[1327] The server generates customized action plans for each company based on the risk assessment results. The input is the risk assessment score, and the output is the action plan. Specifically, the server sets up frequent contact and payment reminders for high-risk companies, and proposes flexible payment plans for low-risk companies. Specific contact methods (email, phone) and contact frequency are set for each company.

[1328] Step 7:

[1329] The server uses an emotion engine to recognize the user's emotions and adaptively modify the action plan. The input is the user's text messages or voice data, and the output is the modified action plan. Specifically, the server uses "Text Analytics" and "Tone Analyzer" to analyze the user's emotional state, and if it determines that the user is experiencing high levels of stress or anxiety, it softens the content of the action plan and emphasizes support.

[1330] Step 8:

[1331] The terminal schedules and executes notifications based on the generated action plan. The input is the action plan, and the output is the notification to the company. Specifically, the terminal schedules and automatically executes weekly follow-up calls and emails to high-risk companies.

[1332] Step 9:

[1333] The server monitors the progress of the action plan in real time and saves a history of the actions performed to a database. The input is the data of the actions performed, and the output is a progress report. Specifically, the server periodically updates the progress data and saves it to the database.

[1334] Step 10:

[1335] The terminal displays the progress to the user and allows for additional follow-up actions as needed. The input is a progress report, and the output is a notification to the user and additional follow-up actions. Specifically, the terminal graphically displays the progress through a dashboard screen and provides an interface where the user can configure the necessary actions.

[1336] (Application Example 2)

[1337] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[1338] Current debt collection systems, while capable of assessing a company's payment risk and generating action plans based on that assessment, struggle to provide adaptive responses that take into account the customer's emotional state. This results in one-sided communication with customers, hindering improvements in collection rates. Furthermore, it's difficult for employees to assess the situation in real time while interacting with customers. Solving these problems is essential to providing a more effective debt collection system.

[1339] In Application Example 2, the specific processing performed by the specific processing unit 290 of the data processing device 12 is realized by the following means. In this invention, the server includes means for collecting industry-specific information, payment status, and company information; means for evaluating the payment risk of companies using the collected information; means for generating an action plan suitable for each company based on the evaluation results; means for automating notifications based on the generated action plan; means for monitoring the progress of the action plan and setting follow-up actions; means for recognizing the emotional state of the user; and means for adaptively changing the action plan based on the recognized emotional state. This enables flexible and adaptive responses that take into account the emotional state of the customer, thereby improving the collection rate and customer satisfaction.

[1340] "Industry-specific information" refers to information such as economic trends, competitive conditions, and business practices related to a specific industry.

[1341] "Payment status" refers to data such as past and present payment history, outstanding amounts, and frequency of payment delays.

[1342] "Company information" refers to information about a corporation's basic information, management status, financial condition, and creditworthiness.

[1343] "Payment risk" refers to a risk assessment calculated based on factors that affect a company's likelihood of fulfilling future payments.

[1344] An "action plan" is a plan that includes specific actions to be taken by the company based on the evaluation results, as well as the schedule for those actions.

[1345] "Automated notifications" is the process of automatically sending notifications at specific times based on a predetermined action plan.

[1346] "Follow-up actions" refer to additional actions or measures taken depending on the progress made after notification.

[1347] "Emotional state" refers to the user's current psychological and emotional condition, and is recognized through voice, text, facial expressions, etc.

[1348] "Adaptively modifying the action plan based on the perceived emotional state" means analyzing the user's emotions and flexibly changing the predetermined plan accordingly.

[1349] This invention relates to a debt collection system that assesses a company's payment risk and generates and executes an appropriate action plan. The system recognizes the user's emotional state and adaptively executes actions based on that state. Specifically, it collects and analyzes industry-specific information, payment status, company information, etc., and automates follow-up actions.

[1350] System Overview

[1351] The server collects company information when users log in and upload a list of companies to be collected from. This information includes basic company information, past payment history, and industry economic trends. The collected data is normalized and centralized. Next, the server uses machine learning algorithms to assess the companies' payment risk. Based on this, high-risk companies will require more frequent follow-up, while low-risk companies will be offered flexible payment plans.

[1352] Emotional engine integration

[1353] The server uses an emotion engine to recognize the user's emotional state. For example, it can determine if a user is stressed through text message or voice analysis. Based on the recognized emotional state, it adaptively changes the action plan. For instance, if it determines that the user is stressed, it changes the message to use softer language and emphasize support.

[1354] Automated notifications and progress monitoring

[1355] Based on the generated action plan, notifications are sent automatically. For example, high-risk companies may receive weekly follow-up calls or emails reminding them of late fees. The server monitors these notification processes and stores the execution history in a database. It also monitors progress in real time and sets up additional follow-up actions as needed.

[1356] Specific example

[1357] For example, consider a case where company B in the retail industry is included in the list. The server checks company B's payment history and discovers that it has a history of frequent payment delays. Based on this information, the server classifies company B as "high risk" and generates an action plan that includes weekly follow-up calls and emails warning about late fees. Additionally, during interactions with customers via smart glasses, the emotion engine analyzes the emotions of company B's representative, and if it determines that the representative is highly stressed, it changes the follow-up to be softer and more supportive. In this way, it is possible to increase the debt collection rate while simultaneously improving customer satisfaction.

[1358] Example prompt message

[1359] Let's say a customer enters a store, and an employee puts on smart glasses and begins serving them. The system detects the customer's stress level from their voice and determines they are high-risk based on their past payment history. Based on this information, the smart glasses display shows a message saying, "Use gentle language and emphasize support."

[1360] Prompt text for input example

[1361] Parameters:

[1362] customer_id: 12345

[1363] audio_input: "path / to / customer / voice / input.wav"

[1364] Query:

[1365] Please tell me about customer service methods that take into account the customer's payment risk level and emotional state.

[1366] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[1367] Step 1: The user logs into the system and uploads a list of companies to be included in the collection process.

[1368] Input: User login information, list of companies to be included in the collection.

[1369] Output: A list of companies subject to recall is saved on the server.

[1370] Specific operation: The user logs in through the system interface and uploads a list of companies to be targeted for collection. The server receives this list and stores it in its internal database.

[1371] Step 2: The server collects corporate information.

[1372] Input: List of companies to be recalled

[1373] Output: Basic company information, payment history, industry economic trends

[1374] Specific operation: The server accesses external APIs and databases to collect basic information, past payment history, and industry economic trends of the target companies. The collected data is normalized and centralized.

[1375] Step 3: The server uses machine learning algorithms to assess the company's payment risk.

[1376] Input: Basic company information, payment history, industry economic trends

[1377] Output: Payment risk assessment results for each company

[1378] Specific operation: The server inputs the collected data into a machine learning algorithm to evaluate the payment risk of each company. The evaluation results are stored in a database.

[1379] Step 4: The server generates an action plan tailored to each company.

[1380] Input: Payment risk assessment results

[1381] Output: Action plans tailored to each company

[1382] Specific operation: Based on the payment risk assessment results, the server generates action plans that include frequent follow-up for high-risk companies and flexible payment plans for low-risk companies.

[1383] Step 5: The device automates notifications based on the generated action plan.

[1384] Input: Action Plan

[1385] Output: Automatically generated notifications (email, phone call, etc.)

[1386] Specific operation: The device automatically sends notifications via email or phone at the appropriate time based on the action plan.

[1387] Step 6: The server uses the emotion engine to recognize the user's emotional state.

[1388] Input: Text messages and audio data from the user.

[1389] Output: User's emotional state

[1390] Specific operation: The server uses an emotion engine to analyze text messages and voice data to detect the user's emotional state.

[1391] Step 7: The server adaptively modifies the action plan based on the recognized emotional state.

[1392] Input: User's emotional state

[1393] Output: Adaptively modified action plan

[1394] Specific operation: The server analyzes the user's emotional state and flexibly modifies the action plan accordingly. For example, if the user is feeling stressed, the follow-up content will be softened and emphasized to be more supportive.

[1395] Step 8: The server monitors the progress of the action plan and sets up follow-up actions.

[1396] Input: Action plan execution history

[1397] Output: Next follow-up action

[1398] Specific operation: The server saves the execution status of the action plan to a database and monitors it in real time. Based on the progress, it calculates the timing of the next follow-up and notifies the user.

[1399] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[1400] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1401] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[1402] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[1403] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[1404] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[1405] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[1406] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[1407] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[1408] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[1409] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[1410] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[1411] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[1412] 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.

[1413] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[1414] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[1415] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[1416] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[1417] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[1418] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[1419] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference.

[1420] The following is further disclosed regarding the embodiments described above.

[1421] (Claim 1)

[1422] Means of collecting industry-specific information, payment status, and company information,

[1423] A means of evaluating a company's payment risk using the collected information,

[1424] A means of generating an action plan suitable for each company based on the evaluation results,

[1425] A means to automate notifications based on the generated action plan,

[1426] A means of monitoring the progress of the action plan and setting follow-up actions,

[1427] A system that includes this.

[1428] (Claim 2)

[1429] The system according to claim 1, further comprising means for normalizing and centralizing the collected information.

[1430] (Claim 3)

[1431] The system according to claim 1, comprising means for evaluating a company's payment risk using a machine learning algorithm.

[1432] "Example 1"

[1433] (Claim 1)

[1434] Means for collecting industry-specific information, transaction status, and business information,

[1435] A means of normalizing and centralizing the collected information,

[1436] A means of evaluating a business's payment risk using normalized information,

[1437] A means of generating an action plan suitable for each business based on the evaluation results,

[1438] A means of automating notifications based on the generated action plan,

[1439] A means of monitoring the progress of the action plan and setting follow-up actions,

[1440] A system that includes this.

[1441] (Claim 2)

[1442] The system according to claim 1, comprising means for evaluating the payment risk of a business operator using a machine learning algorithm.

[1443] (Claim 3)

[1444] The system according to claim 1, further comprising means for monitoring the progress of notifications in real time and calculating and setting the timing of the next action.

[1445] "Application Example 1"

[1446] (Claim 1)

[1447] Means of collecting industry-specific information, payment status, and company information,

[1448] A means of evaluating a company's payment risk using the collected information,

[1449] A means of generating an action plan suitable for each company based on the evaluation results,

[1450] A means to automate notifications based on the generated action plan,

[1451] A means of monitoring the progress of the action plan and setting follow-up actions,

[1452] Methods for obtaining company data using external APIs,

[1453] A method for quantifying a company's payment risk using collected data,

[1454] A means of automatically generating the content and timing of notifications based on risk assessment,

[1455] A system that includes this.

[1456] (Claim 2)

[1457] The system according to claim 1, further comprising means for normalizing and centralizing the collected information.

[1458] (Claim 3)

[1459] The system according to claim 1, comprising means for evaluating a company's payment risk using a machine learning algorithm, and means for using prompt statements in automating notifications.

[1460] "Example 2 of combining an emotion engine"

[1461] (Claim 1)

[1462] Means of collecting industry-specific information, payment status, and company information,

[1463] A means of normalizing and centralizing the collected information,

[1464] A means of evaluating a company's payment risk using a machine learning algorithm with collected information,

[1465] A means of generating customized action plans suitable for each company based on the evaluation results,

[1466] A means of recognizing user emotions using an emotion engine and adaptively changing the action plan,

[1467] A means to automate notifications based on the generated action plan,

[1468] A means to monitor the progress of the action plan, save a history of the actions taken, and set up follow-up actions,

[1469] A system that includes this.

[1470] (Claim 2)

[1471] The system according to claim 1, further comprising means for normalizing and centralizing the collected information.

[1472] (Claim 3)

[1473] The system according to claim 1, further comprising means for evaluating a company's payment risk using a machine learning algorithm.

[1474] "Application example 2 when combining with an emotional engine"

[1475] (Claim 1)

[1476] Means of collecting industry-specific information, payment status, and company information,

[1477] A means of evaluating a company's payment risk using the collected information,

[1478] A means of generating an action plan suitable for each company based on the evaluation results,

[1479] A means to automate notifications based on the generated action plan,

[1480] A means of monitoring the progress of the action plan and setting follow-up actions,

[1481] A means of recognizing the user's emotional state,

[1482] Means for adaptively modifying action plans based on recognized emotional states,

[1483] A system that includes this.

[1484] (Claim 2)

[1485] The system according to claim 1, further comprising means for normalizing and centralizing the collected information.

[1486] (Claim 3)

[1487] A method for evaluating a company's payment risk using machine learning algorithms,

[1488] The system according to claim 1, comprising means of using an algorithm for analyzing the emotional state of a customer. [Explanation of symbols]

[1489] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. Means of collecting industry-specific information, payment status, and company information, A means of evaluating a company's payment risk using the collected information, A means of generating an action plan suitable for each company based on the evaluation results, A means to automate notifications based on the generated action plan, A means of monitoring the progress of the action plan and setting follow-up actions, A system that includes this.

2. The system according to claim 1, further comprising means for normalizing and centralizing the collected information.

3. The system according to claim 1, comprising means for evaluating a company's payment risk using a machine learning algorithm.

Images