System

Abstract

A system is provided.SOLUTION: A system comprising: means for analyzing a large amount of business documents and extracting risk factors using production artificial intelligence; means for matching with a historical risk database and performing comprehensive risk assessment; means for presenting risk information and recommended measures to a user; means for collecting feedback from the user and reflecting the feedback in the system; and means for setting a custom model to perform risk assessment specific to a specific industry.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] Risk management is a very important issue in a company's daily operations. In particular, reviewing large volumes of business documents and contracts takes time and effort, and increases the risk of human error. Furthermore, different industries and businesses each have their own unique risks, so general risk management systems may not be able to adequately address them. To solve these issues, a system is needed that can perform efficient and precise risk analysis and provide appropriate feedback to users. [Means for solving the problem]

[0005] This invention provides a system that uses generative artificial intelligence to analyze large volumes of business documents and extract risk factors. This system includes a means for comparing the results with a past risk database to perform a comprehensive risk assessment, and a means for presenting risk information and recommended countermeasures to the user. It also includes a means for collecting user feedback and incorporating it into the system, and a means for configuring custom models to perform risk assessments specific to specific industries. This allows users to reduce the effort of double-checking and address unreported risks. Furthermore, performing risk assessments tailored to specific industries can enhance business safety and improve efficiency.

[0006] "Generative artificial intelligence" is an algorithm that can automatically learn from large amounts of data and detect risk factors and errors.

[0007] "Business documents" are documents that describe business procedures in a company or organization, such as work procedures, contracts, and reports.

[0008] A "risk factor" is an element that poses a potential problem or danger to a particular document or procedure.

[0009] A "risk database" is a data storage that accumulates previously recorded risk information and error history.

[0010] "Risk assessment" is the process of assessing the severity and probability of occurrence of a problem based on detected risk factors.

[0011] "Feedback" refers to opinions and additional information on the analysis results provided by the user.

[0012] A "custom model" is an analytical algorithm configuration optimized specifically for a specific industry or business.

[0013] A "user" is a person or organization that uses this system to analyze business documents and perform risk assessments.

[0014] "Error detection" is the process of automatically identifying unclear statements and inconsistencies in a document.

[0015] "Risk information" refers to information about the details of risk factors and their effects obtained as a result of analysis.

[0016] An "alert" is a warning message that notifies the user when a risk is detected.

[0017] A "knowledge base" is a database that accumulates past risk information and feedback and uses it for future risk assessments. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0023] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

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

[0028] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0035] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0038] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0039] This invention relates to a risk management system for business documents that utilizes generative artificial intelligence. This system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. Specific embodiments of the system and details of the program processing are described below.

[0040] Overall system configuration

[0041] The system mainly consists of the following components:

[0042] Database server: stores business documents and risk data

[0043] Generative AI module: Performs document analysis and risk assessment

[0044] User Interface: A user interface for uploading documents and viewing analysis results.

[0045] Feedback module: To collect user feedback and reflect it in the system.

[0046] Program processing description

[0047] 1. Data input

[0048] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0049] The server receives the uploaded documents and stores them in a secure database, along with the document's metadata (creation date, author, tag information, etc.).

[0050] 2. Initial Risk Analysis

[0051] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[0052] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[0053] 3. Error Detection and Analysis

[0054] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[0055] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[0056] 4. Centralized risk management

[0057] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[0058] The server generates a comprehensive risk assessment result and presents it to the user in a user interface.

[0059] 5. Providing Feedback

[0060] Users can check the analysis results and risk assessments through a dashboard, and if necessary, enter feedback and request further analysis or corrections.

[0061] The server stores the received feedback in a database and reflects it in the next analysis.

[0062] 6. Applying a Custom Model

[0063] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0064] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[0065] Specific examples

[0066] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0067] 1. User uploads a contract

[0068] Users upload new supply contracts into the system, tag the contracts and select categories.

[0069] 2. The server receives and saves the contract

[0070] The server stores the contract in a secure database and generates metadata.

[0071] 3. Generative AI performs initial risk analysis

[0072] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0073] 4. Generative AI performs detailed analysis

[0074] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0075] 5. The server provides a comprehensive risk assessment

[0076] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0077] 6. User reviews the results and provides feedback

[0078] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0079] In this way, users can streamline risk assessment of contracts, recognize potential issues in advance, and take action.

[0080] The processing flow will be explained below.

[0081] Step 1:

[0082] Users upload business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0083] Step 2:

[0084] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0085] Step 3:

[0086] The server invokes a generative AI module to analyze the content of the stored documents.

[0087] Step 4:

[0088] Generative AI extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[0089] Step 5:

[0090] Generative AI analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadlines.

[0091] Step 6:

[0092] The server creates an initial risk report based on the risk factors and detailed issues provided by the generation AI and stores it in a database.

[0093] Step 7:

[0094] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[0095] Step 8:

[0096] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0097] Step 9:

[0098] Users can review the risk assessment results on the dashboard and enter any necessary corrections or additional feedback.

[0099] Step 10:

[0100] The server receives feedback from the user and reflects it in the database, which will be used in the next analysis.

[0101] Step 11:

[0102] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[0103] Step 12:

[0104] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs.

[0105] Step 13:

[0106] The server notifies the user of the results of applying the custom settings and provides improvements and recommendations.

[0107] Through these specific processing steps, the system quickly and accurately analyzes business documents, improving the efficiency and accuracy of risk management.

[0108] Example 1

[0109] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0110] This invention relates to a system that automatically extracts and analyzes risk factors in business documents that have been difficult to detect efficiently using conventional methods, and provides comprehensive risk assessments. Specifically, it improves business efficiency by providing a method for analyzing large volumes of business documents, identifying and assessing risk factors, and presenting appropriate risk information and recommended countermeasures to users. Another objective of this invention is to provide a means for supporting risk assessments specific to specific industries.

[0111] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0112] In this invention, the server includes: a means for users to upload business documents through a dedicated interface and store them in a secure database; a means for performing an initial risk analysis of the uploaded documents using a generative AI model and extracting keywords and phrases to identify risk factors; a means for performing contextual analysis of the identified risk factors using the generative AI model to detect and analyze specific errors; a means for performing a comprehensive risk assessment by comparing the results with a historical risk database and presenting a risk report to the user; a means for receiving feedback from the user, storing the feedback in a database and reflecting it in the system; and a means for receiving custom model settings and reflecting them in the generative AI model to perform a risk assessment specific to a specific industry. This makes it possible to efficiently extract and analyze risk factors from a large number of business documents and provide users with a comprehensive risk assessment and specific recommended measures.

[0113] "Business documents" refer to documents created and used by companies and organizations to carry out their business. Examples include contracts, work procedures, and reports.

[0114] A "generative AI model" refers to a type of artificial intelligence that uses machine learning and natural language processing to analyze text data and generate information.

[0115] "Risk factors" refer to keywords or phrases contained in business documents that could be potential risk factors, which could affect business operations.

[0116] "Initial risk analysis" refers to the initial analysis process using a generative AI model to extract keywords and phrases contained in business documents and identify risk factors.

[0117] "Contextual analysis" refers to the process of analyzing the text and context of documents containing identified risk factors to identify specific errors or issues.

[0118] "Database" refers to a system for systematically storing and managing business documents and risk-related information.

[0119] "Feedback" refers to comments and additional information provided by users based on the analysis results, which are reflected in improving the accuracy of the system and in the next analysis.

[0120] A "custom model" refers to a model that is configured and adjusted specifically for a specific industry or application, allowing the generative AI model to perform analysis tailored to those specific needs.

[0121] "Risk report" refers to a report that comprehensively summarizes the results of the initial risk analysis and context analysis, and is provided to the user.

[0122] "Dashboard" refers to the web application interface that allows users to view system analysis results, risk assessments, and feedback.

[0123] The business document risk management system of this invention is mainly composed of three entities: a server, a terminal, and a user. The entire system is realized by the main components such as a database server, a generation AI module, a user interface, and a feedback module.

[0124] System hardware and software examples

[0125] The system's database server is used to store business documents and risk data. This server should provide a secure and reliable data storage solution. Specifically, we recommend using an SQL database (e.g., MySQL, PostgreSQL) or a cloud storage service (e.g., Amazon S3).

[0126] The generative AI module includes models for document analysis and risk assessment. This module incorporates natural language processing (NLP) and machine learning (ML) techniques, such as generative AI models like GPT-3 and BERT.

[0127] The user interface is a tool that allows users to upload business documents and check the analysis results. This interface is implemented as a web application that can be accessed through a web browser and uses technologies such as HTML, CSS, and JavaScript.

[0128] The feedback module collects feedback from users and reflects it in the system. This module receives feedback data through form input or API submission and stores it in a database.

[0129] Specific operation of the system

[0130] Data Input

[0131] Users upload their company's business documents to the system through a dedicated interface. This involves clicking the document upload button and selecting the file. The server receives the uploaded document and stores it in a secure database. Document metadata (creation date, author, tag information) is also saved.

[0132] Initial Risk Analysis

[0133] The server applies a generative AI module to the stored documents. The generative AI model (e.g., GPT-3) extracts keywords and phrases from the documents and identifies potential risk factors. For example, in the case of a contract, important items such as "delivery date" and "quality standards" are analyzed.

[0134] Error Detection and Analysis

[0135] The generative AI module performs contextual analysis on the identified risk factors to identify specific errors or issues, such as vague quality requirements or unclear delivery deadlines, and reports the details of the risk to the user.

[0136] Centralized risk management

[0137] The server compares the latest risk report with the existing risk database to see if similar risks have occurred in the past, and then compiles the relevant information to make a comprehensive assessment and present it to the user via a user interface.

[0138] Providing feedback

[0139] Users can check the analysis results and risk assessment on the dashboard, and if necessary, enter feedback to request further analysis or corrections. The server stores the received feedback in a database and reflects it in the next analysis.

[0140] Applying a Custom Model

[0141] On the device (such as a user's PC), custom settings specific to the industry are configured. For example, settings that emphasize quality control items for the manufacturing industry are configured. The server receives the custom settings and reflects them in the generated AI model. This enables risk assessment tailored to the needs of specific industries during analysis.

[0142] Examples of concrete examples and prompts

[0143] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0144] 1. User uploads a contract

[0145] Users upload new supply contracts into the system, tag the contracts and select categories.

[0146] 2. The server receives and saves the contract

[0147] The server stores the contract in a secure database and generates metadata.

[0148] 3. Generative AI performs initial risk analysis

[0149] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0150] 4. Generative AI performs detailed analysis

[0151] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0152] 5. The server provides a comprehensive risk assessment

[0153] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0154] 6. User reviews the results and provides feedback

[0155] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0156] In this way, users can streamline risk assessment of contracts and proactively identify and address potential issues. Specific examples of prompts include:

[0157] "I have uploaded a new supply contract. Please conduct an initial risk analysis and detailed analysis of this contract to identify risk factors related to delivery dates and quality standards."

[0158] Following this prompt, the system analyzes the contract, extracts risk factors, and provides feedback to the user, enabling quick and accurate risk management of business documents.

[0159] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0160] Step 1:

[0161] Uploading business documents

[0162] Users upload their company's business documents to the system through a dedicated interface. Specifically, users open a web browser, click the document upload button, and then specify the file in the file selection dialog or upload it by dragging and dropping.

[0163] Input: Business document file (e.g. PDF, Word)

[0164] Output: The uploaded document file is sent to the server.

[0165] Step 2:

[0166] Document storage and metadata registration

[0167] The server receives the uploaded document and stores it in a secure database, extracting the document's metadata (creation date, author, tag information) and registering it in the database.

[0168] Input: Uploaded document file

[0169] Data processing: file format check, metadata extraction

[0170] Output: Document files and metadata stored in a database

[0171] Step 3:

[0172] Initial Risk Analysis

[0173] The server applies a generative AI module (e.g., GPT-3) to the stored documents to extract keywords and phrases within the documents and identify potential risk factors. For example, in the case of a contract, it analyzes important items such as "delivery date" and "quality standards."

[0174] Input: Document files stored in the database

[0175] Data computation: Keyword extraction and risk factor identification using generative AI models

[0176] Output: Risk factor list

[0177] Step 4:

[0178] Detailed Risk Analysis

[0179] The generative AI module digs deeper into the identified risk factors and performs contextual analysis to identify specific issues or errors, leading to risk details such as "clarification of quality standards is required."

[0180] Input: Risk factor list, document text data

[0181] Data processing: Reanalysis of the context in which risk factors exist and identification of specific problems

[0182] Output: Detailed risk report

[0183] Step 5:

[0184] Centralized Risk Assessment

[0185] Based on the results of the generated AI, the server compares the latest risk report with the existing risk database to check whether similar risks have occurred in the past, and performs a comprehensive risk assessment and compiles the results into a risk report.

[0186] Inputs: Detailed risk reports, existing risk database

[0187] Data calculation: Database matching, calculation of comprehensive risk assessment

[0188] Output: Comprehensive risk report

[0189] Step 6:

[0190] Present risk assessment to users

[0191] The server presents a comprehensive risk report on a user interface, and the user can check the risk assessment results on a dashboard. Specifically, the risk report is displayed on the dashboard of a web application.

[0192] Input: Comprehensive Risk Report

[0193] Output: Risk report displayed in the user interface

[0194] Step 7:

[0195] User feedback

[0196] Users can check the analysis results and risk assessment on the dashboard, enter feedback as needed, and request further analysis or corrections.

[0197] Input: Risk report analysis results, feedback

[0198] Output: The received feedback is sent to the server

[0199] Step 8:

[0200] Feedback is saved and implemented

[0201] The server stores the feedback received from users in a database and reflects it in the next analysis. The feedback data is used to adjust the generative AI model.

[0202] Input: Feedback received

[0203] Data processing: saving feedback data and retuning the AI ​​model

[0204] Output: Updated AI model and feedback data

[0205] Step 9:

[0206] Configuring a Custom Model

[0207] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0208] Input: Custom setting item

[0209] Output: Custom settings are sent to the server

[0210] Step 10:

[0211] Reflecting custom settings

[0212] The server receives the custom configuration and feeds it into the generative AI model, which, upon analysis, tailors the risk assessment to specific industry needs.

[0213] Input: Custom setting item

[0214] Data processing: Adjusting the AI ​​model

[0215] Output: AI model with custom industry-specific settings

[0216] Through the above processing steps, this system can efficiently and accurately manage risks in business documents.

[0217] (Application example 1)

[0218] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0219] Traditionally, it has been extremely difficult to manually manage large volumes of business documents (e.g., delivery contracts, transportation regulations, customer requests) in logistics operations and identify risk factors. In particular, photographing documents, entering metadata, identifying and evaluating risk factors, and collecting and incorporating feedback were inefficient, often resulting in human error and time loss. This made it impossible to recognize potential risks in advance and take measures, leading to reduced operational efficiency and ultimately to a decline in service quality and financial losses.

[0220] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0221] In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database to perform a comprehensive risk assessment, means for presenting risk information and recommended countermeasures to users, means for collecting user feedback and reflecting it in the system, means for performing risk assessment specific to a specific industry by setting a custom model, terminal means for photographing business documents and inputting metadata, means for providing document analysis results on a dashboard, and means for users to input feedback based on the analysis results. This enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[0222] "Generative AI" refers to AI that has the ability to analyze and learn large amounts of text data and automatically generate new data and information.

[0223] "Business documents" are documents related to business within an organization or with external parties, including contracts, regulations, procedures, and customer requests.

[0224] "Risk factors" refer to elements that may cause potential risks, such as unclear clauses or uncertain elements contained in business documents.

[0225] A "risk database" is a database that accumulates and stores past business documents and risk assessment data, and is used to help with future risk assessments.

[0226] "Custom Model" refers to a risk assessment model that is customized to fit the needs of a specific industry or company.

[0227] "Feedback" refers to opinions and suggestions provided by users regarding analysis results and system usage, which are used to improve and update the system.

[0228] "Terminal" refers to a device (e.g., smartphone, tablet) used by a user to photograph, input, and upload business documents.

[0229] "Dashboard" refers to an interface that allows users to visually check analysis results and risk assessment information.

[0230] "Photographing" refers to taking an image of a business document using a digital camera or the camera function of a smartphone.

[0231] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings.

[0232] System Overview

[0233] The system of the present invention mainly comprises the following components:

[0234] 1. Server:

[0235] It includes a database server that stores business documents and risk data, a generative AI module, and a feedback module.

[0236] 2. Terminal:

[0237] A device (e.g., smartphone, tablet) that a user uses to capture, type, or upload business documents.

[0238] 3. User Interface:

[0239] A dashboard for users to view analysis results and risk assessments.

[0240] Program processing explanation

[0241] 1. Data input:

[0242] A user takes a photo of a business document using a terminal and inputs metadata (e.g., creation date and time, creator, tag information). The data is then sent from the terminal to the server.

[0243] The server stores the received business documents and metadata in a secure database.

[0244] 2. Initial risk analysis:

[0245] The server applies a generative AI module to the stored documents to extract risk factors. The generative AI analyzes keywords and phrases in the documents to identify potential risk factors (e.g., unclear clauses, delivery dates, quality standards).

[0246] 3. Risk Assessment and Dashboard View:

[0247] The server compares the extracted risk factors with a historical risk database to perform a comprehensive risk assessment, and the results are presented to the user through a user interface.

[0248] 4. Gather and incorporate feedback:

[0249] Users can check the analysis results displayed on the dashboard and enter feedback as needed, which is then sent to the system and stored on the server.

[0250] The server incorporates the collected feedback into the next risk analysis.

[0251] Hardware and software usage

[0252] Hardware:

[0253] Smartphone or tablet (used for taking photos and entering business documents)

[0254] Cloud server (used for data storage and analysis)

[0255] software:

[0256] Flask (Python web framework)

[0257] Generative AI module (used for document risk analysis)

[0258] Database module (used to store data)

[0259] Specific examples

[0260] For example, here's the scenario when a major logistics company wants to upload a new transportation contract into their system:

[0261] 1. Upload the contract:

[0262] Users take a photo of the shipping contract with their smartphone and upload it to the app, while also entering metadata such as the shipping date and the name of the carrier.

[0263] 2. Risk factor analysis:

[0264] The server uses a generative AI module to extract risk factors such as "transportation period" and "liability" from the contract.

[0265] 3. Providing risk assessments:

[0266] The extracted risk factors are displayed on the user's dashboard and compared with past risk data for evaluation.

[0267] 4. Enter your feedback:

[0268] The user checks the analysis results and inputs feedback such as "The terms of liability are unclear." The feedback is sent to the system and reflected in the next analysis.

[0269] Prompt Sentence Examples

[0270] "Analyze the risk factors in the transportation contract, especially those related to the duration of transportation and liability."

[0271] This invention enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[0272] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0273] Step 1:

[0274] Users take photos and upload business documents

[0275] The user takes a photo of a business document using a device (smartphone or tablet), enters the document's metadata (e.g., creation date and time, author, tag information) along with the image, and uploads it to the server via the app. This sends the business document image and its metadata to the server.

[0276] Step 2:

[0277] The server stores business documents and metadata

[0278] The server stores the received business document images and metadata in a secure database. The stored information includes the document content, creation date and time, author, tags, etc. This data storage ensures that the information necessary for analysis is secured.

[0279] Step 3:

[0280] The server applies the generated AI module to extract risk factors.

[0281] The server applies a generative AI module to the stored business document images. The generative AI module analyzes keywords and phrases (e.g., delivery dates, quality standards) within the document to extract potential risk factors. Based on the input document content, data processing and extraction of risk factors are performed. A list of risk factors is generated as the output.

[0282] Step 4:

[0283] The server compares the extracted risk factors with the risk database and performs a comprehensive evaluation.

[0284] The server compares the generated risk factor list with past risk data. The risk database stores past risk assessment data, and the server compares this data to evaluate the weighting and priority of risk factors. A comprehensive risk assessment is performed based on the input risk factor list, and the assessment results are output.

[0285] Step 5:

[0286] The server displays the risk assessment results on a dashboard.

[0287] The server displays the assessed risk factors and their details on a user interface (dashboard), where risk information and recommended measures are visually presented. The content displayed on the dashboard is generated and provided based on the input risk assessment results.

[0288] Step 6:

[0289] User enters feedback

[0290] The user checks the risk assessment results displayed on the dashboard and enters feedback as necessary. Specific examples (e.g., the liability conditions are unclear) are entered as feedback, and this is sent to the server.

[0291] Step 7:

[0292] The server collects feedback and applies it to the next analysis.

[0293] The server stores the received feedback in a database. The feedback is reflected in the next risk analysis and helps improve the analysis accuracy of the generative AI module. The database is updated based on the input feedback and reflected in the next analysis process.

[0294] In this way, the system enables efficient analysis and risk assessment of business documents through an automated risk management process.

[0295] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0296] This invention relates to a business document risk management system that utilizes generative artificial intelligence and an emotion engine. The system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. It also grasps the user's emotional state and reflects it in the risk assessment. A specific embodiment of the system and details of the program processing are described below.

[0297] Overall system configuration

[0298] The system mainly consists of the following components:

[0299] Database server: stores business documents and risk data

[0300] Generative AI module: Performs document analysis and risk assessment

[0301] Emotion Engine: Recognize and analyze the user's emotional state

[0302] User Interface: A user interface for uploading documents and viewing analysis results.

[0303] Feedback module: To collect user feedback and reflect it in the system.

[0304] Program processing description

[0305] 1. Data input

[0306] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0307] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0308] 2. Initial Risk Analysis

[0309] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[0310] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[0311] 3. Error Detection and Analysis

[0312] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[0313] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[0314] 4. Centralized risk management

[0315] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[0316] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0317] 5. Emotion recognition

[0318] The emotion engine analyzes emotions from the user's voice, text input, and facial expression data, and sends the results to the server.

[0319] For example, it recognizes the emotions (anxiety, relief, etc.) of users in real time when they review risk reports.

[0320] 6. Reflecting Emotional Data

[0321] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[0322] 7. Providing Feedback

[0323] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[0324] The server stores the received feedback in a database and reflects it in the next analysis.

[0325] 8. Applying Custom Models

[0326] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0327] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[0328] 9. Notification of risk assessment results and provision of countermeasures

[0329] The server notifies the user of the risk assessment results that reflect the custom settings and emotional data, and provides remediation measures and recommendations.

[0330] Specific examples

[0331] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0332] 1. User uploads a contract

[0333] Users upload new supply contracts into the system, tag the contracts and select categories.

[0334] 2. The server receives and saves the contract

[0335] The server stores the contract in a secure database and generates metadata.

[0336] 3. Generative AI performs initial risk analysis

[0337] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0338] 4. Generative AI performs detailed analysis

[0339] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0340] 5. The server provides a comprehensive risk assessment

[0341] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0342] 6. Emotion engine analyzes emotions

[0343] The emotion engine analyzes the user's facial expressions and voice when checking the risk report and determines that the user is feeling anxious.

[0344] 7. The server reflects the emotional data

[0345] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[0346] 8. User reviews the results and provides feedback

[0347] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0348] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[0349] The processing flow will be explained below.

[0350] Step 1:

[0351] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0352] Step 2:

[0353] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0354] Step 3:

[0355] The server invokes a generative AI module to analyze the content of the stored documents.

[0356] Step 4:

[0357] The generative AI module extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[0358] Step 5:

[0359] The generative AI module analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadline statements.

[0360] Step 6:

[0361] The server creates an initial risk report based on the risk information and issues obtained by the generative AI module and stores it in a database.

[0362] Step 7:

[0363] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[0364] Step 8:

[0365] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0366] Step 9:

[0367] The emotion engine collects the user's voice, text input, and facial expression data in real time through the user interface.

[0368] Step 10:

[0369] The emotion engine analyzes the collected data and recognizes the user's emotions (e.g., anxiety, relief).

[0370] Step 11:

[0371] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, if the user feels anxious, it may provide additional information or specific solutions to alleviate that anxiety.

[0372] Step 12:

[0373] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[0374] Step 13:

[0375] The server receives user feedback, records it in a database, and uses it in the next risk assessment.

[0376] Step 14:

[0377] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[0378] Step 15:

[0379] The server receives the custom configuration and feeds it into the generative AI, ensuring that the risk assessment meets industry-specific requirements.

[0380] Step 16:

[0381] The server generates a further optimized risk assessment result based on the custom settings and emotion data and notifies the user.

[0382] This allows users to efficiently and accurately assess the risks of business documents, enabling them to detect potential problems early and take measures.In addition, the emotion engine provides appropriate support according to the user's emotional state, improving the reliability of risk management and user satisfaction.

[0383] Example 2

[0384] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0385] Conventional risk management systems do not ensure sufficient accuracy and efficiency in risk assessment of business documents. Furthermore, they lack the functionality to reflect the user's emotional state in the risk assessment, which can prevent them from providing appropriate support. Furthermore, they lack flexibility when conducting risk assessments specific to specific industries.

[0386] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database and performing a comprehensive risk assessment, means for presenting risk information and recommended measures to the user, means for collecting feedback from the user and reflecting it in the system, means for performing a risk assessment specific to a specific industry by setting a custom model, and means for analyzing the user's emotional state and reflecting it in the risk assessment. This makes it possible to improve the accuracy and efficiency of risk assessment of business documents and provide appropriate support according to the user's emotional state.

[0387] "Generative AI" is AI that uses natural language processing and machine learning techniques to analyze text data and generate meaningful information.

[0388] "Business documents" refers to all documents created to carry out the daily work of a company or organization, including, for example, contracts and work procedures.

[0389] "Risk factors" refer to elements or items contained in documents or data that may cause problems or trouble.

[0390] A "risk database" is a database that collects information on risks that have occurred in the past and related data.

[0391] "Comprehensive risk assessment" is a comprehensive assessment based on extracted risk factors and compared with a past risk database.

[0392] "User" refers to the person who uses the system to upload business documents and receive the results of risk assessments.

[0393] "Feedback" refers to information such as evaluation information, improvements, and suggested corrections provided by users.

[0394] A "Custom Model" is an analytical model tailored to the unique needs and characteristics of a particular industry.

[0395] "Emotional state" refers to the psychological and emotional state a user is feeling at a particular moment.

[0396] An "emotion engine" is a technology or module that analyzes a user's voice, text, and facial expression data to recognize and evaluate their emotional state.

[0397] A "dashboard" refers to a graphical user interface that allows users to check system analysis results and risk assessment information at a glance.

[0398] "Error detection" refers to the process of identifying ambiguous statements and inconsistent items in business documents and pointing them out as specific problems.

[0399] The present invention relates to a risk management system for business documents that utilizes generative AI and an emotion engine. This system analyzes a large number of business documents, extracts risk factors, evaluates them, provides feedback, and grasps the user's emotional state, which is reflected in the risk assessment. Specific embodiments of the present invention are described below.

[0400] System Configuration

[0401] The system consists of the following main components:

[0402] Database server: Stores business documents and risk data.

[0403] Generative AI module: Performs document analysis and risk assessment.

[0404] Emotion Engine: Recognizes and analyzes the user's emotional state.

[0405] User Interface: The interface through which users can upload documents and view the analysis results.

[0406] Feedback module: Collects feedback from users and reflects it in the system.

[0407] Hardware and Software

[0408] Server: Hardware that stores and processes data. A high-performance server is desirable.

[0409] Generative AI: Software that uses natural language processing and machine learning techniques, such as models like BERT and GPT.

[0410] Emotion Engine: Software that analyzes a user's voice, text, and facial expression data to recognize their emotional state. For example, using the Emotion API or face recognition software.

[0411] Database: A system for storing risk data and business document metadata, such as a relational database like MySQL or PostgreSQL.

[0412] User interface: An interface for checking analysis results and feedback. These are primarily web-based applications, using JavaScript frameworks (React, Vue.js, etc.).

[0413] Emotion recognition and feedback reflection

[0414] The emotion engine analyzes facial, voice, and text data in real time as users review risk reports to recognize their emotional state (e.g., anxiety, relief).

[0415] The server then incorporates the emotion data into the analysis results and provides additional support information or solutions according to the user's emotional state. For example, if the user is feeling anxious, the server may provide more detailed explanations or additional materials.

[0416] Gathering and implementing feedback

[0417] Users can check the analysis results and risk assessments from the dashboard and enter any necessary corrections or additional feedback.

[0418] The server stores the feedback received from users in a database and reflects it in the next analysis, allowing the system to continuously improve and provide more accurate risk assessments.

[0419] Specific examples

[0420] For example, consider a scenario in which a large manufacturing company uploads a new supply contract to the system. The user uploads the new supply contract to the system, tags the contract, and selects a category. The server stores the contract in a secure database and generates metadata. The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors. The generation AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors. The server performs a comprehensive evaluation and notifies the user through the user interface. At this time, the emotion engine analyzes the user's emotions and provides appropriate additional information.

[0421] Prompt Sentence Examples

[0422] The generative AI model performs a risk assessment of a specific business document by providing prompts such as:

[0423] "Please identify the risk factors contained in this contract and conduct a specific risk assessment. Please analyze with particular emphasis on important items such as delivery dates, quality standards, and payment terms."

[0424] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[0425] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0426] Step 1: Data Input

[0427] Users upload their company's business documents to the system through a dedicated interface. Specifically, they click the "Upload" button on the user interface and select the business document file.

[0428] Input: Business document file, metadata (creation date and time, creator, tag information)

[0429] The server receives the uploaded documents and stores them in a secure database, along with metadata about the documents.

[0430] Output: Document files and metadata stored in a database

[0431] Step 2: Initial risk analysis

[0432] The server applies the generative AI module to documents stored in the database, using AI analysis algorithms to extract keywords and phrases within the documents and identify risk factors.

[0433] Input: Document files in the database

[0434] The server lists potential risk factors within the document based on the keywords and phrases extracted by the generative AI module.

[0435] Output: List of identified risk factors

[0436] Step 3: Error detection and analysis

[0437] The generative AI module digs deep into specific risk areas within a document, analyzes the context, and identifies specific issues, generating error messages such as "Delivery date unclear" or "Quality standards unclear."

[0438] Input: List of identified risk factors, document files in the database

[0439] The generative AI module analyzes the context of the document and identifies specific error locations.

[0440] Output: Detailed risk report with specific error messages

[0441] Step 4: Centralized risk management

[0442] The server compares the risk report created by the generative AI module with an existing risk database, checking whether similar risks have occurred in the past and adding any relevant information.

[0443] Input: Detailed risk report, existing risk database

[0444] The server generates a comprehensive risk assessment and stores the information in a database.

[0445] Output: Comprehensive risk assessment results

[0446] Step 5: Emotion Recognition

[0447] The emotion engine analyzes the user's voice, text input, and facial expression data to recognize their emotional state. For example, it captures the user's facial expressions with a camera and uses a facial expression recognition algorithm to determine their emotions.

[0448] Input: User's voice data, text data, facial expression data

[0449] The emotion engine analyzes this data and sends the results to the server.

[0450] Output: Parsed emotion data

[0451] Step 6: Reflecting emotional data

[0452] The server then converts the emotional data from the emotion engine into a detailed risk report, providing more specific explanations and additional resources if the user is feeling anxious, for example.

[0453] Inputs: Analyzed sentiment data, detailed risk reports

[0454] The server adds additional support information to the risk report according to the emotion data.

[0455] Output: Final risk report reflecting sentiment data

[0456] Step 7: Provide feedback

[0457] Users can review the final risk report from the dashboard and enter feedback such as comments and correction requests.

[0458] Input: User feedback data

[0459] The server stores the feedback received from the user in a database and reflects it in the next risk assessment.

[0460] Output: Feedback data stored in a database

[0461] Step 8: Applying the custom model

[0462] Terminals (such as users' PCs) are custom configured for specific industries, such as manufacturing, to strictly evaluate quality standards.

[0463] Input: User's custom setting data

[0464] The server receives the custom settings and reflects them in the generated AI.

[0465] Output: A generative AI model with your custom settings

[0466] Step 9: Communicate risk assessment results and provide countermeasures

[0467] The server notifies the user of the final risk assessment result, which reflects the custom settings and emotion data, by displaying the result on a dashboard through the user interface.

[0468] Inputs: Custom Settings, Emotion Data, Final Risk Report

[0469] The server notifies the user of the risk assessment results to provide appropriate remediation measures and recommendations.

[0470] Output: Notified risk assessment results and countermeasures

[0471] (Application example 2)

[0472] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0473] Modern content distribution services require automatic analysis of the content being viewed by viewers to verify its safety and appropriateness. However, conventional systems lack the means to efficiently analyze large amounts of content, making it difficult to provide effective feedback that reflects viewers' emotional state. Furthermore, content improvement and risk assessment based on viewer feedback are also inadequate. To solve these problems, an efficient content analysis system that combines an emotion engine and generative AI is needed.

[0474] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0475] In this invention, the server includes means for analyzing risk factors of viewing content using generative artificial intelligence, means for analyzing viewer feedback using an emotion engine to identify an emotional state, and means for providing recommendations to the viewer based on the emotion data and the risk assessment results, thereby making it possible to automatically evaluate the safety and appropriateness of viewing content and provide feedback that reflects the viewer's emotional state.

[0476] "Generative artificial intelligence" refers to artificial intelligence techniques used to automatically analyze data and extract and assess risk factors.

[0477] "Risk factors" are elements that may have an impact on users, such as potential problems, inappropriate information, or violent expressions in the viewing content.

[0478] "Content" is a general term for various digital media provided to viewers, such as video, audio, images, and text.

[0479] A "viewer" is a user who uses a content distribution service to view content.

[0480] An "emotion engine" is a technology that analyzes the emotional state of a user from their feedback and reactions and reflects that in the system.

[0481] "Feedback" refers to information such as impressions, opinions, and evaluations provided by viewers regarding the content they watch.

[0482] A "dashboard" is an interface that allows users and viewers to check analysis results and risk assessments.

[0483] "Comprehensive risk assessment" is a procedure for assessing the safety and appropriateness of content based on analyzed risk factors and emotional data.

[0484] "Recommendations" are specific suggestions or advice provided by the system to improve viewing content or ensure viewer safety.

[0485] This invention is a system that improves the safety and quality of content distribution services by analyzing risk factors of viewing content and providing feedback based on the viewer's emotional state. Specific embodiments and program processing of this system are described below.

[0486] Overall system configuration

[0487] The system mainly consists of the following components:

[0488] Database server: Stores content data and risk data.

[0489] Generative AI module: Analyzes and evaluates risk factors of viewed content.

[0490] Emotion Engine: Analyzes viewer feedback and identifies their emotional state.

[0491] User Interface: The interface for viewers to provide feedback and see the analysis results.

[0492] Feedback module: Generates and provides feedback to viewers based on sentiment data and risk assessment.

[0493] Program processing description

[0494] 1. Data input

[0495] Users upload content to the system from their own devices via a dedicated interface. The server receives the uploaded content and stores it in a secure database, along with the content's metadata (creation date, creator, etc.).

[0496] 2. Initial Risk Analysis

[0497] The server then applies a generative AI module to the content stored in the database. The generative AI extracts keywords and phrases from the content and identifies potential risk factors, such as violent scenes or inappropriate language in a video.

[0498] 3. Error Detection and Analysis

[0499] The generative AI module drills down into specific risk areas and analyzes the context to identify specific issues, which then generates a specific risk report and provides risk details to the user.

[0500] 4. Emotion recognition

[0501] The emotion engine analyzes voice, text input, and facial expression data from viewers and sends the results to the server. For example, it can recognize the viewer's emotions (anxiety, relief, etc.) in real time when watching content.

[0502] 5. Reflecting Emotional Data

[0503] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[0504] 6. Providing Feedback

[0505] The user can check the analysis results and risk assessment through the user interface and input any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[0506] Specific examples

[0507] For example, consider a scenario where a video streaming company uploads a new documentary to their system:

[0508] 1. User uploads a video

[0509] Users upload new documentary videos to the system, tag the videos and select categories.

[0510] 2. The server receives and saves the video

[0511] The server stores the videos in a secure database and generates metadata.

[0512] 3. Generative AI performs initial risk analysis

[0513] The generative AI analyzes important items in the video, such as "violent scenes" and "inappropriate expressions," and extracts risk factors.

[0514] 4. Generative AI performs detailed analysis

[0515] Generative AI detects ambiguities and inconsistencies and identifies specific risks and errors, such as "This scene contains violent content."

[0516] 5. The server provides a comprehensive risk assessment

[0517] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0518] 6. Emotion engine analyzes emotions

[0519] The emotion engine analyzes the viewer's facial expressions and voice when watching a video and determines whether the viewer is feeling anxious.

[0520] 7. The server reflects the emotional data

[0521] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[0522] 8. User reviews the results and provides feedback

[0523] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0524] In this way, viewers can efficiently assess the risk of distributed content, recognize potential problems in advance, and take measures. In addition, by using the emotion engine, appropriate support can be provided according to the viewer's emotional state, further improving the accuracy and efficiency of risk management.

[0525] Prompt Sentence Examples

[0526] "This video contains violent content, but what specifically is the problem?"

[0527] "Tell us your thoughts after watching: How did the video make you feel?"

[0528] In this way, it will be possible to provide risk assessment and support based on viewer feedback, which is expected to improve the quality of content and the viewer experience.

[0529] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0530] Step 1:

[0531] A user uploads viewing content.

[0532] Input: File path of the content you want to watch

[0533] Output: Content data stored in the database

[0534] Specific behavior:

[0535] Users upload content to the server via a dedicated interface from their device, and the server stores the content data and metadata (e.g., creation date and creator) in a secure database.

[0536] Step 2:

[0537] The server uses a generative AI module to perform an initial risk analysis of the content.

[0538] Input: Content data stored in the database

[0539] Output: Initial risk report with risk factors extracted

[0540] Specific behavior:

[0541] The server retrieves the content from the database and uses a generative AI module to analyze keywords and phrases in the content, identifying risk factors such as violent scenes or inappropriate language, and generating an initial risk report.

[0542] Step 3:

[0543] The server uses a generated AI module to perform detailed error detection and analysis.

[0544] Input: Initial Risk Report

[0545] Output: Detailed risk report identifying specific risks and errors

[0546] Specific behavior:

[0547] Based on the initial risk report, the server uses a generative AI module to perform further analysis, digging deeper into ambiguities and inconsistencies to identify specific issues and risks and generate a detailed risk report.

[0548] Step 4:

[0549] An emotion engine analyzes viewer feedback to identify emotional states.

[0550] Input: Audio and text feedback from viewers

[0551] Output: Emotional state data

[0552] Specific behavior:

[0553] Viewers provide their thoughts and opinions through a feedback interface, and the server uses an emotion engine to analyze the audio and text data of the feedback to identify the viewer's emotional state (e.g., anxiety, relief).

[0554] Step 5:

[0555] The server incorporates the emotional data into risk assessment.

[0556] Input: Emotional state data, detailed risk report

[0557] Output: Comprehensive risk assessment results reflecting emotional data

[0558] Specific behavior:

[0559] The server then incorporates the emotional state data obtained from the emotion engine into a detailed risk report, adding risk solutions and supplemental information if the viewer is feeling anxious, and generating a comprehensive risk assessment result.

[0560] Step 6:

[0561] Provide comprehensive risk assessment results and recommendations, and gather user feedback.

[0562] Input: Overall risk assessment results displayed on the user interface

[0563] Output: More feedback from the user

[0564] Specific behavior:

[0565] The server generates a comprehensive risk assessment result and notifies the viewer through a user interface. The viewer can check the results on the dashboard and enter any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[0566] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0567] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0568] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0569] [Second embodiment]

[0570] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0571] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0572] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0573] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0574] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0575] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0576] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0577] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0578] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0579] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0580] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0581] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0582] This invention relates to a risk management system for business documents that utilizes generative artificial intelligence. This system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. Specific embodiments of the system and details of the program processing are described below.

[0583] Overall system configuration

[0584] The system mainly consists of the following components:

[0585] Database server: stores business documents and risk data

[0586] Generative AI module: Performs document analysis and risk assessment

[0587] User Interface: A user interface for uploading documents and viewing analysis results.

[0588] Feedback module: To collect user feedback and reflect it in the system.

[0589] Program processing description

[0590] 1. Data input

[0591] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0592] The server receives the uploaded documents and stores them in a secure database, along with the document's metadata (creation date, author, tag information, etc.).

[0593] 2. Initial Risk Analysis

[0594] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[0595] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[0596] 3. Error Detection and Analysis

[0597] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[0598] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[0599] 4. Centralized risk management

[0600] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[0601] The server generates a comprehensive risk assessment result and presents it to the user in a user interface.

[0602] 5. Providing Feedback

[0603] Users can check the analysis results and risk assessments through a dashboard, and if necessary, enter feedback and request further analysis or corrections.

[0604] The server stores the received feedback in a database and reflects it in the next analysis.

[0605] 6. Applying a Custom Model

[0606] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0607] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[0608] Specific examples

[0609] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0610] 1. User uploads a contract

[0611] Users upload new supply contracts into the system, tag the contracts and select categories.

[0612] 2. The server receives and saves the contract

[0613] The server stores the contract in a secure database and generates metadata.

[0614] 3. Generative AI performs initial risk analysis

[0615] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0616] 4. Generative AI performs detailed analysis

[0617] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0618] 5. The server provides a comprehensive risk assessment

[0619] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0620] 6. User reviews the results and provides feedback

[0621] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0622] In this way, users can streamline risk assessment of contracts, recognize potential issues in advance, and take action.

[0623] The processing flow will be explained below.

[0624] Step 1:

[0625] Users upload business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0626] Step 2:

[0627] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0628] Step 3:

[0629] The server invokes a generative AI module to analyze the content of the stored documents.

[0630] Step 4:

[0631] Generative AI extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[0632] Step 5:

[0633] Generative AI analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadlines.

[0634] Step 6:

[0635] The server creates an initial risk report based on the risk factors and detailed issues provided by the generation AI and stores it in a database.

[0636] Step 7:

[0637] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[0638] Step 8:

[0639] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0640] Step 9:

[0641] Users can review the risk assessment results on the dashboard and enter any necessary corrections or additional feedback.

[0642] Step 10:

[0643] The server receives feedback from the user and reflects it in the database, which will be used in the next analysis.

[0644] Step 11:

[0645] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[0646] Step 12:

[0647] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs.

[0648] Step 13:

[0649] The server notifies the user of the results of applying the custom settings and provides improvements and recommendations.

[0650] Through these specific processing steps, the system quickly and accurately analyzes business documents, improving the efficiency and accuracy of risk management.

[0651] Example 1

[0652] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0653] This invention relates to a system that automatically extracts and analyzes risk factors in business documents that have been difficult to detect efficiently using conventional methods, and provides comprehensive risk assessments. Specifically, it improves business efficiency by providing a method for analyzing large volumes of business documents, identifying and assessing risk factors, and presenting appropriate risk information and recommended countermeasures to users. Another objective of this invention is to provide a means for supporting risk assessments specific to specific industries.

[0654] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0655] In this invention, the server includes: a means for users to upload business documents through a dedicated interface and store them in a secure database; a means for performing an initial risk analysis of the uploaded documents using a generative AI model and extracting keywords and phrases to identify risk factors; a means for performing contextual analysis of the identified risk factors using the generative AI model to detect and analyze specific errors; a means for performing a comprehensive risk assessment by comparing the results with a historical risk database and presenting a risk report to the user; a means for receiving feedback from the user, storing the feedback in a database and reflecting it in the system; and a means for receiving custom model settings and reflecting them in the generative AI model to perform a risk assessment specific to a specific industry. This makes it possible to efficiently extract and analyze risk factors from a large number of business documents and provide users with a comprehensive risk assessment and specific recommended measures.

[0656] "Business documents" refer to documents created and used by companies and organizations to carry out their business. Examples include contracts, work procedures, and reports.

[0657] A "generative AI model" refers to a type of artificial intelligence that uses machine learning and natural language processing to analyze text data and generate information.

[0658] "Risk factors" refer to keywords or phrases contained in business documents that could be potential risk factors, which could affect business operations.

[0659] "Initial risk analysis" refers to the initial analysis process using a generative AI model to extract keywords and phrases contained in business documents and identify risk factors.

[0660] "Contextual analysis" refers to the process of analyzing the text and context of documents containing identified risk factors to identify specific errors or issues.

[0661] "Database" refers to a system for systematically storing and managing business documents and risk-related information.

[0662] "Feedback" refers to comments and additional information provided by users based on the analysis results, which are reflected in improving the accuracy of the system and in the next analysis.

[0663] A "custom model" refers to a model that is configured and adjusted specifically for a specific industry or application, allowing the generative AI model to perform analysis tailored to those specific needs.

[0664] "Risk report" refers to a report that comprehensively summarizes the results of the initial risk analysis and context analysis, and is provided to the user.

[0665] "Dashboard" refers to the web application interface that allows users to view system analysis results, risk assessments, and feedback.

[0666] The business document risk management system of this invention is mainly composed of three entities: a server, a terminal, and a user. The entire system is realized by the main components such as a database server, a generation AI module, a user interface, and a feedback module.

[0667] System hardware and software examples

[0668] The system's database server is used to store business documents and risk data. This server should provide a secure and reliable data storage solution. Specifically, we recommend using an SQL database (e.g., MySQL, PostgreSQL) or a cloud storage service (e.g., Amazon S3).

[0669] The generative AI module includes models for document analysis and risk assessment. This module incorporates natural language processing (NLP) and machine learning (ML) techniques, such as generative AI models like GPT-3 and BERT.

[0670] The user interface is a tool that allows users to upload business documents and check the analysis results. This interface is implemented as a web application that can be accessed through a web browser and uses technologies such as HTML, CSS, and JavaScript.

[0671] The feedback module collects feedback from users and reflects it in the system. This module receives feedback data through form input or API submission and stores it in a database.

[0672] Specific operation of the system

[0673] Data Input

[0674] Users upload their company's business documents to the system through a dedicated interface. This involves clicking the document upload button and selecting the file. The server receives the uploaded document and stores it in a secure database. Document metadata (creation date, author, tag information) is also saved.

[0675] Initial Risk Analysis

[0676] The server applies a generative AI module to the stored documents. The generative AI model (e.g., GPT-3) extracts keywords and phrases from the documents and identifies potential risk factors. For example, in the case of a contract, important items such as "delivery date" and "quality standards" are analyzed.

[0677] Error Detection and Analysis

[0678] The generative AI module performs contextual analysis on the identified risk factors to identify specific errors or issues, such as vague quality requirements or unclear delivery deadlines, and reports the details of the risk to the user.

[0679] Centralized risk management

[0680] The server compares the latest risk report with the existing risk database to see if similar risks have occurred in the past, and then compiles the relevant information to make a comprehensive assessment and present it to the user via a user interface.

[0681] Providing feedback

[0682] Users can check the analysis results and risk assessment on the dashboard, and if necessary, enter feedback to request further analysis or corrections. The server stores the received feedback in a database and reflects it in the next analysis.

[0683] Applying a Custom Model

[0684] On the device (such as a user's PC), custom settings specific to the industry are configured. For example, settings that emphasize quality control items for the manufacturing industry are configured. The server receives the custom settings and reflects them in the generated AI model. This enables risk assessment tailored to the needs of specific industries during analysis.

[0685] Examples of concrete examples and prompts

[0686] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0687] 1. User uploads a contract

[0688] Users upload new supply contracts into the system, tag the contracts and select categories.

[0689] 2. The server receives and saves the contract

[0690] The server stores the contract in a secure database and generates metadata.

[0691] 3. Generative AI performs initial risk analysis

[0692] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0693] 4. Generative AI performs detailed analysis

[0694] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0695] 5. The server provides a comprehensive risk assessment

[0696] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0697] 6. User reviews the results and provides feedback

[0698] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0699] In this way, users can streamline risk assessment of contracts and proactively identify and address potential issues. Specific examples of prompts include:

[0700] "I have uploaded a new supply contract. Please conduct an initial risk analysis and detailed analysis of this contract to identify risk factors related to delivery dates and quality standards."

[0701] Following this prompt, the system analyzes the contract, extracts risk factors, and provides feedback to the user, enabling quick and accurate risk management of business documents.

[0702] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0703] Step 1:

[0704] Uploading business documents

[0705] Users upload their company's business documents to the system through a dedicated interface. Specifically, users open a web browser, click the document upload button, and then specify the file in the file selection dialog or upload it by dragging and dropping.

[0706] Input: Business document file (e.g. PDF, Word)

[0707] Output: The uploaded document file is sent to the server.

[0708] Step 2:

[0709] Document storage and metadata registration

[0710] The server receives the uploaded document and stores it in a secure database, extracting the document's metadata (creation date, author, tag information) and registering it in the database.

[0711] Input: Uploaded document file

[0712] Data processing: file format check, metadata extraction

[0713] Output: Document files and metadata stored in a database

[0714] Step 3:

[0715] Initial Risk Analysis

[0716] The server applies a generative AI module (e.g., GPT-3) to the stored documents to extract keywords and phrases within the documents and identify potential risk factors. For example, in the case of a contract, it analyzes important items such as "delivery date" and "quality standards."

[0717] Input: Document files stored in the database

[0718] Data computation: Keyword extraction and risk factor identification using generative AI models

[0719] Output: Risk factor list

[0720] Step 4:

[0721] Detailed Risk Analysis

[0722] The generative AI module digs deeper into the identified risk factors and performs contextual analysis to identify specific issues or errors, leading to risk details such as "clarification of quality standards is required."

[0723] Input: Risk factor list, document text data

[0724] Data processing: Reanalysis of the context in which risk factors exist and identification of specific problems

[0725] Output: Detailed risk report

[0726] Step 5:

[0727] Centralized Risk Assessment

[0728] Based on the results of the generated AI, the server compares the latest risk report with the existing risk database to check whether similar risks have occurred in the past, and performs a comprehensive risk assessment and compiles the results into a risk report.

[0729] Inputs: Detailed risk reports, existing risk database

[0730] Data calculation: Database matching, calculation of comprehensive risk assessment

[0731] Output: Comprehensive risk report

[0732] Step 6:

[0733] Present risk assessment to users

[0734] The server presents a comprehensive risk report on a user interface, and the user can check the risk assessment results on a dashboard. Specifically, the risk report is displayed on the dashboard of a web application.

[0735] Input: Comprehensive Risk Report

[0736] Output: Risk report displayed in the user interface

[0737] Step 7:

[0738] User feedback

[0739] Users can check the analysis results and risk assessment on the dashboard, enter feedback as needed, and request further analysis or corrections.

[0740] Input: Risk report analysis results, feedback

[0741] Output: The received feedback is sent to the server

[0742] Step 8:

[0743] Feedback is saved and implemented

[0744] The server stores the feedback received from users in a database and reflects it in the next analysis. The feedback data is used to adjust the generative AI model.

[0745] Input: Feedback received

[0746] Data processing: saving feedback data and retuning the AI ​​model

[0747] Output: Updated AI model and feedback data

[0748] Step 9:

[0749] Configuring a Custom Model

[0750] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0751] Input: Custom setting item

[0752] Output: Custom settings are sent to the server

[0753] Step 10:

[0754] Reflecting custom settings

[0755] The server receives the custom configuration and feeds it into the generative AI model, which, upon analysis, tailors the risk assessment to specific industry needs.

[0756] Input: Custom setting item

[0757] Data processing: Adjusting the AI ​​model

[0758] Output: AI model with custom industry-specific settings

[0759] Through the above processing steps, this system can efficiently and accurately manage risks in business documents.

[0760] (Application example 1)

[0761] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0762] Traditionally, it has been extremely difficult to manually manage large volumes of business documents (e.g., delivery contracts, transportation regulations, customer requests) in logistics operations and identify risk factors. In particular, photographing documents, entering metadata, identifying and evaluating risk factors, and collecting and incorporating feedback were inefficient, often resulting in human error and time loss. This made it impossible to recognize potential risks in advance and take measures, leading to reduced operational efficiency and ultimately to a decline in service quality and financial losses.

[0763] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0764] In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database to perform a comprehensive risk assessment, means for presenting risk information and recommended countermeasures to users, means for collecting user feedback and reflecting it in the system, means for performing risk assessment specific to a specific industry by setting a custom model, terminal means for photographing business documents and inputting metadata, means for providing document analysis results on a dashboard, and means for users to input feedback based on the analysis results. This enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[0765] "Generative AI" refers to AI that has the ability to analyze and learn large amounts of text data and automatically generate new data and information.

[0766] "Business documents" are documents related to business within an organization or with external parties, including contracts, regulations, procedures, and customer requests.

[0767] "Risk factors" refer to elements that may cause potential risks, such as unclear clauses or uncertain elements contained in business documents.

[0768] A "risk database" is a database that accumulates and stores past business documents and risk assessment data, and is used to help with future risk assessments.

[0769] "Custom Model" refers to a risk assessment model that is customized to fit the needs of a specific industry or company.

[0770] "Feedback" refers to opinions and suggestions provided by users regarding analysis results and system usage, which are used to improve and update the system.

[0771] "Terminal" refers to a device (e.g., smartphone, tablet) used by a user to photograph, input, and upload business documents.

[0772] "Dashboard" refers to an interface that allows users to visually check analysis results and risk assessment information.

[0773] "Photographing" refers to taking an image of a business document using a digital camera or the camera function of a smartphone.

[0774] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings.

[0775] System Overview

[0776] The system of the present invention mainly comprises the following components:

[0777] 1. Server:

[0778] It includes a database server that stores business documents and risk data, a generative AI module, and a feedback module.

[0779] 2. Terminal:

[0780] A device (e.g., smartphone, tablet) that a user uses to capture, type, or upload business documents.

[0781] 3. User Interface:

[0782] A dashboard for users to view analysis results and risk assessments.

[0783] Program processing explanation

[0784] 1. Data input:

[0785] A user takes a photo of a business document using a terminal and inputs metadata (e.g., creation date and time, creator, tag information). The data is then sent from the terminal to the server.

[0786] The server stores the received business documents and metadata in a secure database.

[0787] 2. Initial risk analysis:

[0788] The server applies a generative AI module to the stored documents to extract risk factors. The generative AI analyzes keywords and phrases in the documents to identify potential risk factors (e.g., unclear clauses, delivery dates, quality standards).

[0789] 3. Risk Assessment and Dashboard View:

[0790] The server compares the extracted risk factors with a historical risk database to perform a comprehensive risk assessment, and the results are presented to the user through a user interface.

[0791] 4. Gather and incorporate feedback:

[0792] Users can check the analysis results displayed on the dashboard and enter feedback as needed, which is then sent to the system and stored on the server.

[0793] The server incorporates the collected feedback into the next risk analysis.

[0794] Hardware and software usage

[0795] Hardware:

[0796] Smartphone or tablet (used for taking photos and entering business documents)

[0797] Cloud server (used for data storage and analysis)

[0798] software:

[0799] Flask (Python web framework)

[0800] Generative AI module (used for document risk analysis)

[0801] Database module (used to store data)

[0802] Specific examples

[0803] For example, here's the scenario when a major logistics company wants to upload a new transportation contract into their system:

[0804] 1. Upload the contract:

[0805] Users take a photo of the shipping contract with their smartphone and upload it to the app, while also entering metadata such as the shipping date and the name of the carrier.

[0806] 2. Risk factor analysis:

[0807] The server uses a generative AI module to extract risk factors such as "transportation period" and "liability" from the contract.

[0808] 3. Providing risk assessments:

[0809] The extracted risk factors are displayed on the user's dashboard and compared with past risk data for evaluation.

[0810] 4. Enter your feedback:

[0811] The user checks the analysis results and inputs feedback such as "The terms of liability are unclear." The feedback is sent to the system and reflected in the next analysis.

[0812] Prompt Sentence Examples

[0813] "Analyze the risk factors in the transportation contract, especially those related to the duration of transportation and liability."

[0814] This invention enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[0815] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0816] Step 1:

[0817] Users take photos and upload business documents

[0818] The user takes a photo of a business document using a device (smartphone or tablet), enters the document's metadata (e.g., creation date and time, author, tag information) along with the image, and uploads it to the server via the app. This sends the business document image and its metadata to the server.

[0819] Step 2:

[0820] The server stores business documents and metadata

[0821] The server stores the received business document images and metadata in a secure database. The stored information includes the document content, creation date and time, author, tags, etc. This data storage ensures that the information necessary for analysis is secured.

[0822] Step 3:

[0823] The server applies the generated AI module to extract risk factors.

[0824] The server applies a generative AI module to the stored business document images. The generative AI module analyzes keywords and phrases (e.g., delivery dates, quality standards) within the document to extract potential risk factors. Based on the input document content, data processing and extraction of risk factors are performed. A list of risk factors is generated as the output.

[0825] Step 4:

[0826] The server compares the extracted risk factors with the risk database and performs a comprehensive evaluation.

[0827] The server compares the generated risk factor list with past risk data. The risk database stores past risk assessment data, and the server compares this data to evaluate the weighting and priority of risk factors. A comprehensive risk assessment is performed based on the input risk factor list, and the assessment results are output.

[0828] Step 5:

[0829] The server displays the risk assessment results on a dashboard.

[0830] The server displays the assessed risk factors and their details on a user interface (dashboard), where risk information and recommended measures are visually presented. The content displayed on the dashboard is generated and provided based on the input risk assessment results.

[0831] Step 6:

[0832] User enters feedback

[0833] The user checks the risk assessment results displayed on the dashboard and enters feedback as necessary. Specific examples (e.g., the liability conditions are unclear) are entered as feedback, and this is sent to the server.

[0834] Step 7:

[0835] The server collects feedback and applies it to the next analysis.

[0836] The server stores the received feedback in a database. The feedback is reflected in the next risk analysis and helps improve the analysis accuracy of the generative AI module. The database is updated based on the input feedback and reflected in the next analysis process.

[0837] In this way, the system enables efficient analysis and risk assessment of business documents through an automated risk management process.

[0838] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0839] This invention relates to a business document risk management system that utilizes generative artificial intelligence and an emotion engine. The system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. It also grasps the user's emotional state and reflects it in the risk assessment. A specific embodiment of the system and details of the program processing are described below.

[0840] Overall system configuration

[0841] The system mainly consists of the following components:

[0842] Database server: stores business documents and risk data

[0843] Generative AI module: Performs document analysis and risk assessment

[0844] Emotion Engine: Recognize and analyze the user's emotional state

[0845] User Interface: A user interface for uploading documents and viewing analysis results.

[0846] Feedback module: To collect user feedback and reflect it in the system.

[0847] Program processing description

[0848] 1. Data input

[0849] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0850] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0851] 2. Initial Risk Analysis

[0852] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[0853] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[0854] 3. Error Detection and Analysis

[0855] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[0856] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[0857] 4. Centralized risk management

[0858] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[0859] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0860] 5. Emotion recognition

[0861] The emotion engine analyzes emotions from the user's voice, text input, and facial expression data, and sends the results to the server.

[0862] For example, it recognizes the emotions (anxiety, relief, etc.) of users in real time when they review risk reports.

[0863] 6. Reflecting Emotional Data

[0864] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[0865] 7. Providing Feedback

[0866] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[0867] The server stores the received feedback in a database and reflects it in the next analysis.

[0868] 8. Applying Custom Models

[0869] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[0870] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[0871] 9. Notification of risk assessment results and provision of countermeasures

[0872] The server notifies the user of the risk assessment results that reflect the custom settings and emotional data, and provides remediation measures and recommendations.

[0873] Specific examples

[0874] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[0875] 1. User uploads a contract

[0876] Users upload new supply contracts into the system, tag the contracts and select categories.

[0877] 2. The server receives and saves the contract

[0878] The server stores the contract in a secure database and generates metadata.

[0879] 3. Generative AI performs initial risk analysis

[0880] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[0881] 4. Generative AI performs detailed analysis

[0882] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[0883] 5. The server provides a comprehensive risk assessment

[0884] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[0885] 6. Emotion engine analyzes emotions

[0886] The emotion engine analyzes the user's facial expressions and voice when checking the risk report and determines that the user is feeling anxious.

[0887] 7. The server reflects the emotional data

[0888] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[0889] 8. User reviews the results and provides feedback

[0890] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[0891] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[0892] The processing flow will be explained below.

[0893] Step 1:

[0894] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[0895] Step 2:

[0896] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[0897] Step 3:

[0898] The server invokes a generative AI module to analyze the content of the stored documents.

[0899] Step 4:

[0900] The generative AI module extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[0901] Step 5:

[0902] The generative AI module analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadline statements.

[0903] Step 6:

[0904] The server creates an initial risk report based on the risk information and issues obtained by the generative AI module and stores it in a database.

[0905] Step 7:

[0906] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[0907] Step 8:

[0908] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[0909] Step 9:

[0910] The emotion engine collects the user's voice, text input, and facial expression data in real time through the user interface.

[0911] Step 10:

[0912] The emotion engine analyzes the collected data and recognizes the user's emotions (e.g., anxiety, relief).

[0913] Step 11:

[0914] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, if the user feels anxious, it may provide additional information or specific solutions to alleviate that anxiety.

[0915] Step 12:

[0916] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[0917] Step 13:

[0918] The server receives user feedback, records it in a database, and uses it in the next risk assessment.

[0919] Step 14:

[0920] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[0921] Step 15:

[0922] The server receives the custom configuration and feeds it into the generative AI, ensuring that the risk assessment meets industry-specific requirements.

[0923] Step 16:

[0924] The server generates a further optimized risk assessment result based on the custom settings and emotion data and notifies the user.

[0925] This allows users to efficiently and accurately assess the risks of business documents, enabling them to detect potential problems early and take measures.In addition, the emotion engine provides appropriate support according to the user's emotional state, improving the reliability of risk management and user satisfaction.

[0926] Example 2

[0927] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0928] Conventional risk management systems do not ensure sufficient accuracy and efficiency in risk assessment of business documents. Furthermore, they lack the functionality to reflect the user's emotional state in the risk assessment, which can prevent them from providing appropriate support. Furthermore, they lack flexibility when conducting risk assessments specific to specific industries.

[0929] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database and performing a comprehensive risk assessment, means for presenting risk information and recommended measures to the user, means for collecting feedback from the user and reflecting it in the system, means for performing a risk assessment specific to a specific industry by setting a custom model, and means for analyzing the user's emotional state and reflecting it in the risk assessment. This makes it possible to improve the accuracy and efficiency of risk assessment of business documents and provide appropriate support according to the user's emotional state.

[0930] "Generative AI" is AI that uses natural language processing and machine learning techniques to analyze text data and generate meaningful information.

[0931] "Business documents" refers to all documents created to carry out the daily work of a company or organization, including, for example, contracts and work procedures.

[0932] "Risk factors" refer to elements or items contained in documents or data that may cause problems or trouble.

[0933] A "risk database" is a database that collects information on risks that have occurred in the past and related data.

[0934] "Comprehensive risk assessment" is a comprehensive assessment based on extracted risk factors and compared with a past risk database.

[0935] "User" refers to the person who uses the system to upload business documents and receive the results of risk assessments.

[0936] "Feedback" refers to information such as evaluation information, improvements, and suggested corrections provided by users.

[0937] A "Custom Model" is an analytical model tailored to the unique needs and characteristics of a particular industry.

[0938] "Emotional state" refers to the psychological and emotional state a user is feeling at a particular moment.

[0939] An "emotion engine" is a technology or module that analyzes a user's voice, text, and facial expression data to recognize and evaluate their emotional state.

[0940] A "dashboard" refers to a graphical user interface that allows users to check system analysis results and risk assessment information at a glance.

[0941] "Error detection" refers to the process of identifying ambiguous statements and inconsistent items in business documents and pointing them out as specific problems.

[0942] The present invention relates to a risk management system for business documents that utilizes generative AI and an emotion engine. This system analyzes a large number of business documents, extracts risk factors, evaluates them, provides feedback, and grasps the user's emotional state, which is reflected in the risk assessment. Specific embodiments of the present invention are described below.

[0943] System Configuration

[0944] The system consists of the following main components:

[0945] Database server: Stores business documents and risk data.

[0946] Generative AI module: Performs document analysis and risk assessment.

[0947] Emotion Engine: Recognizes and analyzes the user's emotional state.

[0948] User Interface: The interface through which users can upload documents and view the analysis results.

[0949] Feedback module: Collects feedback from users and reflects it in the system.

[0950] Hardware and Software

[0951] Server: Hardware that stores and processes data. A high-performance server is desirable.

[0952] Generative AI: Software that uses natural language processing and machine learning techniques, such as models like BERT and GPT.

[0953] Emotion Engine: Software that analyzes a user's voice, text, and facial expression data to recognize their emotional state. For example, using the Emotion API or face recognition software.

[0954] Database: A system for storing risk data and business document metadata, such as a relational database like MySQL or PostgreSQL.

[0955] User interface: An interface for checking analysis results and feedback. These are primarily web-based applications, using JavaScript frameworks (React, Vue.js, etc.).

[0956] Emotion recognition and feedback reflection

[0957] The emotion engine analyzes facial, voice, and text data in real time as users review risk reports to recognize their emotional state (e.g., anxiety, relief).

[0958] The server then incorporates the emotion data into the analysis results and provides additional support information or solutions according to the user's emotional state. For example, if the user is feeling anxious, the server may provide more detailed explanations or additional materials.

[0959] Gathering and implementing feedback

[0960] Users can check the analysis results and risk assessments from the dashboard and enter any necessary corrections or additional feedback.

[0961] The server stores the feedback received from users in a database and reflects it in the next analysis, allowing the system to continuously improve and provide more accurate risk assessments.

[0962] Specific examples

[0963] For example, consider a scenario in which a large manufacturing company uploads a new supply contract to the system. The user uploads the new supply contract to the system, tags the contract, and selects a category. The server stores the contract in a secure database and generates metadata. The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors. The generation AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors. The server performs a comprehensive evaluation and notifies the user through the user interface. At this time, the emotion engine analyzes the user's emotions and provides appropriate additional information.

[0964] Prompt Sentence Examples

[0965] The generative AI model performs a risk assessment of a specific business document by providing prompts such as:

[0966] "Please identify the risk factors contained in this contract and conduct a specific risk assessment. Please analyze with particular emphasis on important items such as delivery dates, quality standards, and payment terms."

[0967] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[0968] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0969] Step 1: Data Input

[0970] Users upload their company's business documents to the system through a dedicated interface. Specifically, they click the "Upload" button on the user interface and select the business document file.

[0971] Input: Business document file, metadata (creation date and time, creator, tag information)

[0972] The server receives the uploaded documents and stores them in a secure database, along with metadata about the documents.

[0973] Output: Document files and metadata stored in a database

[0974] Step 2: Initial risk analysis

[0975] The server applies the generative AI module to documents stored in the database, using AI analysis algorithms to extract keywords and phrases within the documents and identify risk factors.

[0976] Input: Document files in the database

[0977] The server lists potential risk factors within the document based on the keywords and phrases extracted by the generative AI module.

[0978] Output: List of identified risk factors

[0979] Step 3: Error detection and analysis

[0980] The generative AI module digs deep into specific risk areas within a document, analyzes the context, and identifies specific issues, generating error messages such as "Delivery date unclear" or "Quality standards unclear."

[0981] Input: List of identified risk factors, document files in the database

[0982] The generative AI module analyzes the context of the document and identifies specific error locations.

[0983] Output: Detailed risk report with specific error messages

[0984] Step 4: Centralized risk management

[0985] The server compares the risk report created by the generative AI module with an existing risk database, checking whether similar risks have occurred in the past and adding any relevant information.

[0986] Input: Detailed risk report, existing risk database

[0987] The server generates a comprehensive risk assessment and stores the information in a database.

[0988] Output: Comprehensive risk assessment results

[0989] Step 5: Emotion Recognition

[0990] The emotion engine analyzes the user's voice, text input, and facial expression data to recognize their emotional state. For example, it captures the user's facial expressions with a camera and uses a facial expression recognition algorithm to determine their emotions.

[0991] Input: User's voice data, text data, facial expression data

[0992] The emotion engine analyzes this data and sends the results to the server.

[0993] Output: Parsed emotion data

[0994] Step 6: Reflecting emotional data

[0995] The server then converts the emotional data from the emotion engine into a detailed risk report, providing more specific explanations and additional resources if the user is feeling anxious, for example.

[0996] Inputs: Analyzed sentiment data, detailed risk reports

[0997] The server adds additional support information to the risk report according to the emotion data.

[0998] Output: Final risk report reflecting sentiment data

[0999] Step 7: Provide feedback

[1000] Users can review the final risk report from the dashboard and enter feedback such as comments and correction requests.

[1001] Input: User feedback data

[1002] The server stores the feedback received from the user in a database and reflects it in the next risk assessment.

[1003] Output: Feedback data stored in a database

[1004] Step 8: Applying the custom model

[1005] Terminals (such as users' PCs) are custom configured for specific industries, such as manufacturing, to strictly evaluate quality standards.

[1006] Input: User's custom setting data

[1007] The server receives the custom settings and reflects them in the generated AI.

[1008] Output: A generative AI model with your custom settings

[1009] Step 9: Communicate risk assessment results and provide countermeasures

[1010] The server notifies the user of the final risk assessment result, which reflects the custom settings and emotion data, by displaying the result on a dashboard through the user interface.

[1011] Inputs: Custom Settings, Emotion Data, Final Risk Report

[1012] The server notifies the user of the risk assessment results to provide appropriate remediation measures and recommendations.

[1013] Output: Notified risk assessment results and countermeasures

[1014] (Application example 2)

[1015] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1016] Modern content distribution services require automatic analysis of the content being viewed by viewers to verify its safety and appropriateness. However, conventional systems lack the means to efficiently analyze large amounts of content, making it difficult to provide effective feedback that reflects viewers' emotional state. Furthermore, content improvement and risk assessment based on viewer feedback are also inadequate. To solve these problems, an efficient content analysis system that combines an emotion engine and generative AI is needed.

[1017] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1018] In this invention, the server includes means for analyzing risk factors of viewing content using generative artificial intelligence, means for analyzing viewer feedback using an emotion engine to identify an emotional state, and means for providing recommendations to the viewer based on the emotion data and the risk assessment results, thereby making it possible to automatically evaluate the safety and appropriateness of viewing content and provide feedback that reflects the viewer's emotional state.

[1019] "Generative artificial intelligence" refers to artificial intelligence techniques used to automatically analyze data and extract and assess risk factors.

[1020] "Risk factors" are elements that may have an impact on users, such as potential problems, inappropriate information, or violent expressions in the viewing content.

[1021] "Content" is a general term for various digital media provided to viewers, such as video, audio, images, and text.

[1022] A "viewer" is a user who uses a content distribution service to view content.

[1023] An "emotion engine" is a technology that analyzes the emotional state of a user from their feedback and reactions and reflects that in the system.

[1024] "Feedback" refers to information such as impressions, opinions, and evaluations provided by viewers regarding the content they watch.

[1025] A "dashboard" is an interface that allows users and viewers to check analysis results and risk assessments.

[1026] "Comprehensive risk assessment" is a procedure for assessing the safety and appropriateness of content based on analyzed risk factors and emotional data.

[1027] "Recommendations" are specific suggestions or advice provided by the system to improve viewing content or ensure viewer safety.

[1028] This invention is a system that improves the safety and quality of content distribution services by analyzing risk factors of viewing content and providing feedback based on the viewer's emotional state. Specific embodiments and program processing of this system are described below.

[1029] Overall system configuration

[1030] The system mainly consists of the following components:

[1031] Database server: Stores content data and risk data.

[1032] Generative AI module: Analyzes and evaluates risk factors of viewed content.

[1033] Emotion Engine: Analyzes viewer feedback and identifies their emotional state.

[1034] User Interface: The interface for viewers to provide feedback and see the analysis results.

[1035] Feedback module: Generates and provides feedback to viewers based on sentiment data and risk assessment.

[1036] Program processing description

[1037] 1. Data input

[1038] Users upload content to the system from their own devices via a dedicated interface. The server receives the uploaded content and stores it in a secure database, along with the content's metadata (creation date, creator, etc.).

[1039] 2. Initial Risk Analysis

[1040] The server then applies a generative AI module to the content stored in the database. The generative AI extracts keywords and phrases from the content and identifies potential risk factors, such as violent scenes or inappropriate language in a video.

[1041] 3. Error Detection and Analysis

[1042] The generative AI module drills down into specific risk areas and analyzes the context to identify specific issues, which then generates a specific risk report and provides risk details to the user.

[1043] 4. Emotion recognition

[1044] The emotion engine analyzes voice, text input, and facial expression data from viewers and sends the results to the server. For example, it can recognize the viewer's emotions (anxiety, relief, etc.) in real time when watching content.

[1045] 5. Reflecting Emotional Data

[1046] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[1047] 6. Providing Feedback

[1048] The user can check the analysis results and risk assessment through the user interface and input any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[1049] Specific examples

[1050] For example, consider a scenario where a video streaming company uploads a new documentary to their system:

[1051] 1. User uploads a video

[1052] Users upload new documentary videos to the system, tag the videos and select categories.

[1053] 2. The server receives and saves the video

[1054] The server stores the videos in a secure database and generates metadata.

[1055] 3. Generative AI performs initial risk analysis

[1056] The generative AI analyzes important items in the video, such as "violent scenes" and "inappropriate expressions," and extracts risk factors.

[1057] 4. Generative AI performs detailed analysis

[1058] Generative AI detects ambiguities and inconsistencies and identifies specific risks and errors, such as "This scene contains violent content."

[1059] 5. The server provides a comprehensive risk assessment

[1060] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1061] 6. Emotion engine analyzes emotions

[1062] The emotion engine analyzes the viewer's facial expressions and voice when watching a video and determines whether the viewer is feeling anxious.

[1063] 7. The server reflects the emotional data

[1064] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[1065] 8. User reviews the results and provides feedback

[1066] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1067] In this way, viewers can efficiently assess the risk of distributed content, recognize potential problems in advance, and take measures. In addition, by using the emotion engine, appropriate support can be provided according to the viewer's emotional state, further improving the accuracy and efficiency of risk management.

[1068] Prompt Sentence Examples

[1069] "This video contains violent content, but what specifically is the problem?"

[1070] "Tell us your thoughts after watching: How did the video make you feel?"

[1071] In this way, it will be possible to provide risk assessment and support based on viewer feedback, which is expected to improve the quality of content and the viewer experience.

[1072] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1073] Step 1:

[1074] A user uploads viewing content.

[1075] Input: File path of the content you want to watch

[1076] Output: Content data stored in the database

[1077] Specific behavior:

[1078] Users upload content to the server via a dedicated interface from their device, and the server stores the content data and metadata (e.g., creation date and creator) in a secure database.

[1079] Step 2:

[1080] The server uses a generative AI module to perform an initial risk analysis of the content.

[1081] Input: Content data stored in the database

[1082] Output: Initial risk report with risk factors extracted

[1083] Specific behavior:

[1084] The server retrieves the content from the database and uses a generative AI module to analyze keywords and phrases in the content, identifying risk factors such as violent scenes or inappropriate language, and generating an initial risk report.

[1085] Step 3:

[1086] The server uses a generated AI module to perform detailed error detection and analysis.

[1087] Input: Initial Risk Report

[1088] Output: Detailed risk report identifying specific risks and errors

[1089] Specific behavior:

[1090] Based on the initial risk report, the server uses a generative AI module to perform further analysis, digging deeper into ambiguities and inconsistencies to identify specific issues and risks and generate a detailed risk report.

[1091] Step 4:

[1092] An emotion engine analyzes viewer feedback to identify emotional states.

[1093] Input: Audio and text feedback from viewers

[1094] Output: Emotional state data

[1095] Specific behavior:

[1096] Viewers provide their thoughts and opinions through a feedback interface, and the server uses an emotion engine to analyze the audio and text data of the feedback to identify the viewer's emotional state (e.g., anxiety, relief).

[1097] Step 5:

[1098] The server incorporates the emotional data into risk assessment.

[1099] Input: Emotional state data, detailed risk report

[1100] Output: Comprehensive risk assessment results reflecting emotional data

[1101] Specific behavior:

[1102] The server then incorporates the emotional state data obtained from the emotion engine into a detailed risk report, adding risk solutions and supplemental information if the viewer is feeling anxious, and generating a comprehensive risk assessment result.

[1103] Step 6:

[1104] Provide comprehensive risk assessment results and recommendations, and gather user feedback.

[1105] Input: Overall risk assessment results displayed on the user interface

[1106] Output: More feedback from the user

[1107] Specific behavior:

[1108] The server generates a comprehensive risk assessment result and notifies the viewer through a user interface. The viewer can check the results on the dashboard and enter any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[1109] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1110] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1111] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1112] [Third embodiment]

[1113] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1114] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1115] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1116] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1117] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1118] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1119] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1120] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1121] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1122] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1123] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1124] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1125] This invention relates to a risk management system for business documents that utilizes generative artificial intelligence. This system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. Specific embodiments of the system and details of the program processing are described below.

[1126] Overall system configuration

[1127] The system mainly consists of the following components:

[1128] Database server: stores business documents and risk data

[1129] Generative AI module: Performs document analysis and risk assessment

[1130] User Interface: A user interface for uploading documents and viewing analysis results.

[1131] Feedback module: To collect user feedback and reflect it in the system.

[1132] Program processing description

[1133] 1. Data input

[1134] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1135] The server receives the uploaded documents and stores them in a secure database, along with the document's metadata (creation date, author, tag information, etc.).

[1136] 2. Initial Risk Analysis

[1137] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[1138] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[1139] 3. Error Detection and Analysis

[1140] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[1141] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[1142] 4. Centralized risk management

[1143] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[1144] The server generates a comprehensive risk assessment result and presents it to the user in a user interface.

[1145] 5. Providing Feedback

[1146] Users can check the analysis results and risk assessments through a dashboard, and if necessary, enter feedback and request further analysis or corrections.

[1147] The server stores the received feedback in a database and reflects it in the next analysis.

[1148] 6. Applying a Custom Model

[1149] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1150] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[1151] Specific examples

[1152] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1153] 1. User uploads a contract

[1154] Users upload new supply contracts into the system, tag the contracts and select categories.

[1155] 2. The server receives and saves the contract

[1156] The server stores the contract in a secure database and generates metadata.

[1157] 3. Generative AI performs initial risk analysis

[1158] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1159] 4. Generative AI performs detailed analysis

[1160] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1161] 5. The server provides a comprehensive risk assessment

[1162] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1163] 6. User reviews the results and provides feedback

[1164] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1165] In this way, users can streamline risk assessment of contracts, recognize potential issues in advance, and take action.

[1166] The processing flow will be explained below.

[1167] Step 1:

[1168] Users upload business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1169] Step 2:

[1170] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1171] Step 3:

[1172] The server invokes a generative AI module to analyze the content of the stored documents.

[1173] Step 4:

[1174] Generative AI extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[1175] Step 5:

[1176] Generative AI analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadlines.

[1177] Step 6:

[1178] The server creates an initial risk report based on the risk factors and detailed issues provided by the generation AI and stores it in a database.

[1179] Step 7:

[1180] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[1181] Step 8:

[1182] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1183] Step 9:

[1184] Users can review the risk assessment results on the dashboard and enter any necessary corrections or additional feedback.

[1185] Step 10:

[1186] The server receives feedback from the user and reflects it in the database, which will be used in the next analysis.

[1187] Step 11:

[1188] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[1189] Step 12:

[1190] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs.

[1191] Step 13:

[1192] The server notifies the user of the results of applying the custom settings and provides improvements and recommendations.

[1193] Through these specific processing steps, the system quickly and accurately analyzes business documents, improving the efficiency and accuracy of risk management.

[1194] Example 1

[1195] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1196] This invention relates to a system that automatically extracts and analyzes risk factors in business documents that have been difficult to detect efficiently using conventional methods, and provides comprehensive risk assessments. Specifically, it improves business efficiency by providing a method for analyzing large volumes of business documents, identifying and assessing risk factors, and presenting appropriate risk information and recommended countermeasures to users. Another objective of this invention is to provide a means for supporting risk assessments specific to specific industries.

[1197] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1198] In this invention, the server includes: a means for users to upload business documents through a dedicated interface and store them in a secure database; a means for performing an initial risk analysis of the uploaded documents using a generative AI model and extracting keywords and phrases to identify risk factors; a means for performing contextual analysis of the identified risk factors using the generative AI model to detect and analyze specific errors; a means for performing a comprehensive risk assessment by comparing the results with a historical risk database and presenting a risk report to the user; a means for receiving feedback from the user, storing the feedback in a database and reflecting it in the system; and a means for receiving custom model settings and reflecting them in the generative AI model to perform a risk assessment specific to a specific industry. This makes it possible to efficiently extract and analyze risk factors from a large number of business documents and provide users with a comprehensive risk assessment and specific recommended measures.

[1199] "Business documents" refer to documents created and used by companies and organizations to carry out their business. Examples include contracts, work procedures, and reports.

[1200] A "generative AI model" refers to a type of artificial intelligence that uses machine learning and natural language processing to analyze text data and generate information.

[1201] "Risk factors" refer to keywords or phrases contained in business documents that could be potential risk factors, which could affect business operations.

[1202] "Initial risk analysis" refers to the initial analysis process using a generative AI model to extract keywords and phrases contained in business documents and identify risk factors.

[1203] "Contextual analysis" refers to the process of analyzing the text and context of documents containing identified risk factors to identify specific errors or issues.

[1204] "Database" refers to a system for systematically storing and managing business documents and risk-related information.

[1205] "Feedback" refers to comments and additional information provided by users based on the analysis results, which are reflected in improving the accuracy of the system and in the next analysis.

[1206] A "custom model" refers to a model that is configured and adjusted specifically for a specific industry or application, allowing the generative AI model to perform analysis tailored to those specific needs.

[1207] "Risk report" refers to a report that comprehensively summarizes the results of the initial risk analysis and context analysis, and is provided to the user.

[1208] "Dashboard" refers to the web application interface that allows users to view system analysis results, risk assessments, and feedback.

[1209] The business document risk management system of this invention is mainly composed of three entities: a server, a terminal, and a user. The entire system is realized by the main components such as a database server, a generation AI module, a user interface, and a feedback module.

[1210] System hardware and software examples

[1211] The system's database server is used to store business documents and risk data. This server should provide a secure and reliable data storage solution. Specifically, we recommend using an SQL database (e.g., MySQL, PostgreSQL) or a cloud storage service (e.g., Amazon S3).

[1212] The generative AI module includes models for document analysis and risk assessment. This module incorporates natural language processing (NLP) and machine learning (ML) techniques, such as generative AI models like GPT-3 and BERT.

[1213] The user interface is a tool that allows users to upload business documents and check the analysis results. This interface is implemented as a web application that can be accessed through a web browser and uses technologies such as HTML, CSS, and JavaScript.

[1214] The feedback module collects feedback from users and reflects it in the system. This module receives feedback data through form input or API submission and stores it in a database.

[1215] Specific operation of the system

[1216] Data Input

[1217] Users upload their company's business documents to the system through a dedicated interface. This involves clicking the document upload button and selecting the file. The server receives the uploaded document and stores it in a secure database. Document metadata (creation date, author, tag information) is also saved.

[1218] Initial Risk Analysis

[1219] The server applies a generative AI module to the stored documents. The generative AI model (e.g., GPT-3) extracts keywords and phrases from the documents and identifies potential risk factors. For example, in the case of a contract, important items such as "delivery date" and "quality standards" are analyzed.

[1220] Error Detection and Analysis

[1221] The generative AI module performs contextual analysis on the identified risk factors to identify specific errors or issues, such as vague quality requirements or unclear delivery deadlines, and reports the details of the risk to the user.

[1222] Centralized risk management

[1223] The server compares the latest risk report with the existing risk database to see if similar risks have occurred in the past, and then compiles the relevant information to make a comprehensive assessment and present it to the user via a user interface.

[1224] Providing feedback

[1225] Users can check the analysis results and risk assessment on the dashboard, and if necessary, enter feedback to request further analysis or corrections. The server stores the received feedback in a database and reflects it in the next analysis.

[1226] Applying a Custom Model

[1227] On the device (such as a user's PC), custom settings specific to the industry are configured. For example, settings that emphasize quality control items for the manufacturing industry are configured. The server receives the custom settings and reflects them in the generated AI model. This enables risk assessment tailored to the needs of specific industries during analysis.

[1228] Examples of concrete examples and prompts

[1229] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1230] 1. User uploads a contract

[1231] Users upload new supply contracts into the system, tag the contracts and select categories.

[1232] 2. The server receives and saves the contract

[1233] The server stores the contract in a secure database and generates metadata.

[1234] 3. Generative AI performs initial risk analysis

[1235] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1236] 4. Generative AI performs detailed analysis

[1237] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1238] 5. The server provides a comprehensive risk assessment

[1239] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1240] 6. User reviews the results and provides feedback

[1241] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1242] In this way, users can streamline risk assessment of contracts and proactively identify and address potential issues. Specific examples of prompts include:

[1243] "I have uploaded a new supply contract. Please conduct an initial risk analysis and detailed analysis of this contract to identify risk factors related to delivery dates and quality standards."

[1244] Following this prompt, the system analyzes the contract, extracts risk factors, and provides feedback to the user, enabling quick and accurate risk management of business documents.

[1245] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1246] Step 1:

[1247] Uploading business documents

[1248] Users upload their company's business documents to the system through a dedicated interface. Specifically, users open a web browser, click the document upload button, and then specify the file in the file selection dialog or upload it by dragging and dropping.

[1249] Input: Business document file (e.g. PDF, Word)

[1250] Output: The uploaded document file is sent to the server.

[1251] Step 2:

[1252] Document storage and metadata registration

[1253] The server receives the uploaded document and stores it in a secure database, extracting the document's metadata (creation date, author, tag information) and registering it in the database.

[1254] Input: Uploaded document file

[1255] Data processing: file format check, metadata extraction

[1256] Output: Document files and metadata stored in a database

[1257] Step 3:

[1258] Initial Risk Analysis

[1259] The server applies a generative AI module (e.g., GPT-3) to the stored documents to extract keywords and phrases within the documents and identify potential risk factors. For example, in the case of a contract, it analyzes important items such as "delivery date" and "quality standards."

[1260] Input: Document files stored in the database

[1261] Data computation: Keyword extraction and risk factor identification using generative AI models

[1262] Output: Risk factor list

[1263] Step 4:

[1264] Detailed Risk Analysis

[1265] The generative AI module digs deeper into the identified risk factors and performs contextual analysis to identify specific issues or errors, leading to risk details such as "clarification of quality standards is required."

[1266] Input: Risk factor list, document text data

[1267] Data processing: Reanalysis of the context in which risk factors exist and identification of specific problems

[1268] Output: Detailed risk report

[1269] Step 5:

[1270] Centralized Risk Assessment

[1271] Based on the results of the generated AI, the server compares the latest risk report with the existing risk database to check whether similar risks have occurred in the past, and performs a comprehensive risk assessment and compiles the results into a risk report.

[1272] Inputs: Detailed risk reports, existing risk database

[1273] Data calculation: Database matching, calculation of comprehensive risk assessment

[1274] Output: Comprehensive risk report

[1275] Step 6:

[1276] Present risk assessment to users

[1277] The server presents a comprehensive risk report on a user interface, and the user can check the risk assessment results on a dashboard. Specifically, the risk report is displayed on the dashboard of a web application.

[1278] Input: Comprehensive Risk Report

[1279] Output: Risk report displayed in the user interface

[1280] Step 7:

[1281] User feedback

[1282] Users can check the analysis results and risk assessment on the dashboard, enter feedback as needed, and request further analysis or corrections.

[1283] Input: Risk report analysis results, feedback

[1284] Output: The received feedback is sent to the server

[1285] Step 8:

[1286] Feedback is saved and implemented

[1287] The server stores the feedback received from users in a database and reflects it in the next analysis. The feedback data is used to adjust the generative AI model.

[1288] Input: Feedback received

[1289] Data processing: saving feedback data and retuning the AI ​​model

[1290] Output: Updated AI model and feedback data

[1291] Step 9:

[1292] Configuring a Custom Model

[1293] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1294] Input: Custom setting item

[1295] Output: Custom settings are sent to the server

[1296] Step 10:

[1297] Reflecting custom settings

[1298] The server receives the custom configuration and feeds it into the generative AI model, which, upon analysis, tailors the risk assessment to specific industry needs.

[1299] Input: Custom setting item

[1300] Data processing: Adjusting the AI ​​model

[1301] Output: AI model with custom industry-specific settings

[1302] Through the above processing steps, this system can efficiently and accurately manage risks in business documents.

[1303] (Application example 1)

[1304] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1305] Traditionally, it has been extremely difficult to manually manage large volumes of business documents (e.g., delivery contracts, transportation regulations, customer requests) in logistics operations and identify risk factors. In particular, photographing documents, entering metadata, identifying and evaluating risk factors, and collecting and incorporating feedback were inefficient, often resulting in human error and time loss. This made it impossible to recognize potential risks in advance and take measures, leading to reduced operational efficiency and ultimately to a decline in service quality and financial losses.

[1306] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1307] In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database to perform a comprehensive risk assessment, means for presenting risk information and recommended countermeasures to users, means for collecting user feedback and reflecting it in the system, means for performing risk assessment specific to a specific industry by setting a custom model, terminal means for photographing business documents and inputting metadata, means for providing document analysis results on a dashboard, and means for users to input feedback based on the analysis results. This enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[1308] "Generative AI" refers to AI that has the ability to analyze and learn large amounts of text data and automatically generate new data and information.

[1309] "Business documents" are documents related to business within an organization or with external parties, including contracts, regulations, procedures, and customer requests.

[1310] "Risk factors" refer to elements that may cause potential risks, such as unclear clauses or uncertain elements contained in business documents.

[1311] A "risk database" is a database that accumulates and stores past business documents and risk assessment data, and is used to help with future risk assessments.

[1312] "Custom Model" refers to a risk assessment model that is customized to fit the needs of a specific industry or company.

[1313] "Feedback" refers to opinions and suggestions provided by users regarding analysis results and system usage, which are used to improve and update the system.

[1314] "Terminal" refers to a device (e.g., smartphone, tablet) used by a user to photograph, input, and upload business documents.

[1315] "Dashboard" refers to an interface that allows users to visually check analysis results and risk assessment information.

[1316] "Photographing" refers to taking an image of a business document using a digital camera or the camera function of a smartphone.

[1317] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings.

[1318] System Overview

[1319] The system of the present invention mainly comprises the following components:

[1320] 1. Server:

[1321] It includes a database server that stores business documents and risk data, a generative AI module, and a feedback module.

[1322] 2. Terminal:

[1323] A device (e.g., smartphone, tablet) that a user uses to capture, type, or upload business documents.

[1324] 3. User Interface:

[1325] A dashboard for users to view analysis results and risk assessments.

[1326] Program processing explanation

[1327] 1. Data input:

[1328] A user takes a photo of a business document using a terminal and inputs metadata (e.g., creation date and time, creator, tag information). The data is then sent from the terminal to the server.

[1329] The server stores the received business documents and metadata in a secure database.

[1330] 2. Initial risk analysis:

[1331] The server applies a generative AI module to the stored documents to extract risk factors. The generative AI analyzes keywords and phrases in the documents to identify potential risk factors (e.g., unclear clauses, delivery dates, quality standards).

[1332] 3. Risk Assessment and Dashboard View:

[1333] The server compares the extracted risk factors with a historical risk database to perform a comprehensive risk assessment, and the results are presented to the user through a user interface.

[1334] 4. Gather and incorporate feedback:

[1335] Users can check the analysis results displayed on the dashboard and enter feedback as needed, which is then sent to the system and stored on the server.

[1336] The server incorporates the collected feedback into the next risk analysis.

[1337] Hardware and software usage

[1338] Hardware:

[1339] Smartphone or tablet (used for taking photos and entering business documents)

[1340] Cloud server (used for data storage and analysis)

[1341] software:

[1342] Flask (Python web framework)

[1343] Generative AI module (used for document risk analysis)

[1344] Database module (used to store data)

[1345] Specific examples

[1346] For example, here's the scenario when a major logistics company wants to upload a new transportation contract into their system:

[1347] 1. Upload the contract:

[1348] Users take a photo of the shipping contract with their smartphone and upload it to the app, while also entering metadata such as the shipping date and the name of the carrier.

[1349] 2. Risk factor analysis:

[1350] The server uses a generative AI module to extract risk factors such as "transportation period" and "liability" from the contract.

[1351] 3. Providing risk assessments:

[1352] The extracted risk factors are displayed on the user's dashboard and compared with past risk data for evaluation.

[1353] 4. Enter your feedback:

[1354] The user checks the analysis results and inputs feedback such as "The terms of liability are unclear." The feedback is sent to the system and reflected in the next analysis.

[1355] Prompt Sentence Examples

[1356] "Analyze the risk factors in the transportation contract, especially those related to the duration of transportation and liability."

[1357] This invention enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[1358] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1359] Step 1:

[1360] Users take photos and upload business documents

[1361] The user takes a photo of a business document using a device (smartphone or tablet), enters the document's metadata (e.g., creation date and time, author, tag information) along with the image, and uploads it to the server via the app. This sends the business document image and its metadata to the server.

[1362] Step 2:

[1363] The server stores business documents and metadata

[1364] The server stores the received business document images and metadata in a secure database. The stored information includes the document content, creation date and time, author, tags, etc. This data storage ensures that the information necessary for analysis is secured.

[1365] Step 3:

[1366] The server applies the generated AI module to extract risk factors.

[1367] The server applies a generative AI module to the stored business document images. The generative AI module analyzes keywords and phrases (e.g., delivery dates, quality standards) within the document to extract potential risk factors. Based on the input document content, data processing and extraction of risk factors are performed. A list of risk factors is generated as the output.

[1368] Step 4:

[1369] The server compares the extracted risk factors with the risk database and performs a comprehensive evaluation.

[1370] The server compares the generated risk factor list with past risk data. The risk database stores past risk assessment data, and the server compares this data to evaluate the weighting and priority of risk factors. A comprehensive risk assessment is performed based on the input risk factor list, and the assessment results are output.

[1371] Step 5:

[1372] The server displays the risk assessment results on a dashboard.

[1373] The server displays the assessed risk factors and their details on a user interface (dashboard), where risk information and recommended measures are visually presented. The content displayed on the dashboard is generated and provided based on the input risk assessment results.

[1374] Step 6:

[1375] User enters feedback

[1376] The user checks the risk assessment results displayed on the dashboard and enters feedback as necessary. Specific examples (e.g., the liability conditions are unclear) are entered as feedback, and this is sent to the server.

[1377] Step 7:

[1378] The server collects feedback and applies it to the next analysis.

[1379] The server stores the received feedback in a database. The feedback is reflected in the next risk analysis and helps improve the analysis accuracy of the generative AI module. The database is updated based on the input feedback and reflected in the next analysis process.

[1380] In this way, the system enables efficient analysis and risk assessment of business documents through an automated risk management process.

[1381] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1382] This invention relates to a business document risk management system that utilizes generative artificial intelligence and an emotion engine. The system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. It also grasps the user's emotional state and reflects it in the risk assessment. A specific embodiment of the system and details of the program processing are described below.

[1383] Overall system configuration

[1384] The system mainly consists of the following components:

[1385] Database server: stores business documents and risk data

[1386] Generative AI module: Performs document analysis and risk assessment

[1387] Emotion Engine: Recognize and analyze the user's emotional state

[1388] User Interface: A user interface for uploading documents and viewing analysis results.

[1389] Feedback module: To collect user feedback and reflect it in the system.

[1390] Program processing description

[1391] 1. Data input

[1392] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1393] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1394] 2. Initial Risk Analysis

[1395] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[1396] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[1397] 3. Error Detection and Analysis

[1398] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[1399] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[1400] 4. Centralized risk management

[1401] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[1402] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1403] 5. Emotion recognition

[1404] The emotion engine analyzes emotions from the user's voice, text input, and facial expression data, and sends the results to the server.

[1405] For example, it recognizes the emotions (anxiety, relief, etc.) of users in real time when they review risk reports.

[1406] 6. Reflecting Emotional Data

[1407] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[1408] 7. Providing Feedback

[1409] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[1410] The server stores the received feedback in a database and reflects it in the next analysis.

[1411] 8. Applying Custom Models

[1412] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1413] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[1414] 9. Notification of risk assessment results and provision of countermeasures

[1415] The server notifies the user of the risk assessment results that reflect the custom settings and emotional data, and provides remediation measures and recommendations.

[1416] Specific examples

[1417] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1418] 1. User uploads a contract

[1419] Users upload new supply contracts into the system, tag the contracts and select categories.

[1420] 2. The server receives and saves the contract

[1421] The server stores the contract in a secure database and generates metadata.

[1422] 3. Generative AI performs initial risk analysis

[1423] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1424] 4. Generative AI performs detailed analysis

[1425] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1426] 5. The server provides a comprehensive risk assessment

[1427] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1428] 6. Emotion engine analyzes emotions

[1429] The emotion engine analyzes the user's facial expressions and voice when checking the risk report and determines that the user is feeling anxious.

[1430] 7. The server reflects the emotional data

[1431] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[1432] 8. User reviews the results and provides feedback

[1433] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1434] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[1435] The processing flow will be explained below.

[1436] Step 1:

[1437] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1438] Step 2:

[1439] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1440] Step 3:

[1441] The server invokes a generative AI module to analyze the content of the stored documents.

[1442] Step 4:

[1443] The generative AI module extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[1444] Step 5:

[1445] The generative AI module analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadline statements.

[1446] Step 6:

[1447] The server creates an initial risk report based on the risk information and issues obtained by the generative AI module and stores it in a database.

[1448] Step 7:

[1449] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[1450] Step 8:

[1451] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1452] Step 9:

[1453] The emotion engine collects the user's voice, text input, and facial expression data in real time through the user interface.

[1454] Step 10:

[1455] The emotion engine analyzes the collected data and recognizes the user's emotions (e.g., anxiety, relief).

[1456] Step 11:

[1457] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, if the user feels anxious, it may provide additional information or specific solutions to alleviate that anxiety.

[1458] Step 12:

[1459] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[1460] Step 13:

[1461] The server receives user feedback, records it in a database, and uses it in the next risk assessment.

[1462] Step 14:

[1463] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[1464] Step 15:

[1465] The server receives the custom configuration and feeds it into the generative AI, ensuring that the risk assessment meets industry-specific requirements.

[1466] Step 16:

[1467] The server generates a further optimized risk assessment result based on the custom settings and emotion data and notifies the user.

[1468] This allows users to efficiently and accurately assess the risks of business documents, enabling them to detect potential problems early and take measures.In addition, the emotion engine provides appropriate support according to the user's emotional state, improving the reliability of risk management and user satisfaction.

[1469] Example 2

[1470] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1471] Conventional risk management systems do not ensure sufficient accuracy and efficiency in risk assessment of business documents. Furthermore, they lack the functionality to reflect the user's emotional state in the risk assessment, which can prevent them from providing appropriate support. Furthermore, they lack flexibility when conducting risk assessments specific to specific industries.

[1472] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database and performing a comprehensive risk assessment, means for presenting risk information and recommended measures to the user, means for collecting feedback from the user and reflecting it in the system, means for performing a risk assessment specific to a specific industry by setting a custom model, and means for analyzing the user's emotional state and reflecting it in the risk assessment. This makes it possible to improve the accuracy and efficiency of risk assessment of business documents and provide appropriate support according to the user's emotional state.

[1473] "Generative AI" is AI that uses natural language processing and machine learning techniques to analyze text data and generate meaningful information.

[1474] "Business documents" refers to all documents created to carry out the daily work of a company or organization, including, for example, contracts and work procedures.

[1475] "Risk factors" refer to elements or items contained in documents or data that may cause problems or trouble.

[1476] A "risk database" is a database that collects information on risks that have occurred in the past and related data.

[1477] "Comprehensive risk assessment" is a comprehensive assessment based on extracted risk factors and compared with a past risk database.

[1478] "User" refers to the person who uses the system to upload business documents and receive the results of risk assessments.

[1479] "Feedback" refers to information such as evaluation information, improvements, and suggested corrections provided by users.

[1480] A "Custom Model" is an analytical model tailored to the unique needs and characteristics of a particular industry.

[1481] "Emotional state" refers to the psychological and emotional state a user is feeling at a particular moment.

[1482] An "emotion engine" is a technology or module that analyzes a user's voice, text, and facial expression data to recognize and evaluate their emotional state.

[1483] A "dashboard" refers to a graphical user interface that allows users to check system analysis results and risk assessment information at a glance.

[1484] "Error detection" refers to the process of identifying ambiguous statements and inconsistent items in business documents and pointing them out as specific problems.

[1485] The present invention relates to a risk management system for business documents that utilizes generative AI and an emotion engine. This system analyzes a large number of business documents, extracts risk factors, evaluates them, provides feedback, and grasps the user's emotional state, which is reflected in the risk assessment. Specific embodiments of the present invention are described below.

[1486] System Configuration

[1487] The system consists of the following main components:

[1488] Database server: Stores business documents and risk data.

[1489] Generative AI module: Performs document analysis and risk assessment.

[1490] Emotion Engine: Recognizes and analyzes the user's emotional state.

[1491] User Interface: The interface through which users can upload documents and view the analysis results.

[1492] Feedback module: Collects feedback from users and reflects it in the system.

[1493] Hardware and Software

[1494] Server: Hardware that stores and processes data. A high-performance server is desirable.

[1495] Generative AI: Software that uses natural language processing and machine learning techniques, such as models like BERT and GPT.

[1496] Emotion Engine: Software that analyzes a user's voice, text, and facial expression data to recognize their emotional state. For example, using the Emotion API or face recognition software.

[1497] Database: A system for storing risk data and business document metadata, such as a relational database like MySQL or PostgreSQL.

[1498] User interface: An interface for checking analysis results and feedback. These are primarily web-based applications, using JavaScript frameworks (React, Vue.js, etc.).

[1499] Emotion recognition and feedback reflection

[1500] The emotion engine analyzes facial, voice, and text data in real time as users review risk reports to recognize their emotional state (e.g., anxiety, relief).

[1501] The server then incorporates the emotion data into the analysis results and provides additional support information or solutions according to the user's emotional state. For example, if the user is feeling anxious, the server may provide more detailed explanations or additional materials.

[1502] Gathering and implementing feedback

[1503] Users can check the analysis results and risk assessments from the dashboard and enter any necessary corrections or additional feedback.

[1504] The server stores the feedback received from users in a database and reflects it in the next analysis, allowing the system to continuously improve and provide more accurate risk assessments.

[1505] Specific examples

[1506] For example, consider a scenario in which a large manufacturing company uploads a new supply contract to the system. The user uploads the new supply contract to the system, tags the contract, and selects a category. The server stores the contract in a secure database and generates metadata. The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors. The generation AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors. The server performs a comprehensive evaluation and notifies the user through the user interface. At this time, the emotion engine analyzes the user's emotions and provides appropriate additional information.

[1507] Prompt Sentence Examples

[1508] The generative AI model performs a risk assessment of a specific business document by providing prompts such as:

[1509] "Please identify the risk factors contained in this contract and conduct a specific risk assessment. Please analyze with particular emphasis on important items such as delivery dates, quality standards, and payment terms."

[1510] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[1511] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1512] Step 1: Data Input

[1513] Users upload their company's business documents to the system through a dedicated interface. Specifically, they click the "Upload" button on the user interface and select the business document file.

[1514] Input: Business document file, metadata (creation date and time, creator, tag information)

[1515] The server receives the uploaded documents and stores them in a secure database, along with metadata about the documents.

[1516] Output: Document files and metadata stored in a database

[1517] Step 2: Initial risk analysis

[1518] The server applies the generative AI module to documents stored in the database, using AI analysis algorithms to extract keywords and phrases within the documents and identify risk factors.

[1519] Input: Document files in the database

[1520] The server lists potential risk factors within the document based on the keywords and phrases extracted by the generative AI module.

[1521] Output: List of identified risk factors

[1522] Step 3: Error detection and analysis

[1523] The generative AI module digs deep into specific risk areas within a document, analyzes the context, and identifies specific issues, generating error messages such as "Delivery date unclear" or "Quality standards unclear."

[1524] Input: List of identified risk factors, document files in the database

[1525] The generative AI module analyzes the context of the document and identifies specific error locations.

[1526] Output: Detailed risk report with specific error messages

[1527] Step 4: Centralized risk management

[1528] The server compares the risk report created by the generative AI module with an existing risk database, checking whether similar risks have occurred in the past and adding any relevant information.

[1529] Input: Detailed risk report, existing risk database

[1530] The server generates a comprehensive risk assessment and stores the information in a database.

[1531] Output: Comprehensive risk assessment results

[1532] Step 5: Emotion Recognition

[1533] The emotion engine analyzes the user's voice, text input, and facial expression data to recognize their emotional state. For example, it captures the user's facial expressions with a camera and uses a facial expression recognition algorithm to determine their emotions.

[1534] Input: User's voice data, text data, facial expression data

[1535] The emotion engine analyzes this data and sends the results to the server.

[1536] Output: Parsed emotion data

[1537] Step 6: Reflecting emotional data

[1538] The server then converts the emotional data from the emotion engine into a detailed risk report, providing more specific explanations and additional resources if the user is feeling anxious, for example.

[1539] Inputs: Analyzed sentiment data, detailed risk reports

[1540] The server adds additional support information to the risk report according to the emotion data.

[1541] Output: Final risk report reflecting sentiment data

[1542] Step 7: Provide feedback

[1543] Users can review the final risk report from the dashboard and enter feedback such as comments and correction requests.

[1544] Input: User feedback data

[1545] The server stores the feedback received from the user in a database and reflects it in the next risk assessment.

[1546] Output: Feedback data stored in a database

[1547] Step 8: Applying the custom model

[1548] Terminals (such as users' PCs) are custom configured for specific industries, such as manufacturing, to strictly evaluate quality standards.

[1549] Input: User's custom setting data

[1550] The server receives the custom settings and reflects them in the generated AI.

[1551] Output: A generative AI model with your custom settings

[1552] Step 9: Communicate risk assessment results and provide countermeasures

[1553] The server notifies the user of the final risk assessment result, which reflects the custom settings and emotion data, by displaying the result on a dashboard through the user interface.

[1554] Inputs: Custom Settings, Emotion Data, Final Risk Report

[1555] The server notifies the user of the risk assessment results to provide appropriate remediation measures and recommendations.

[1556] Output: Notified risk assessment results and countermeasures

[1557] (Application example 2)

[1558] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1559] Modern content distribution services require automatic analysis of the content being viewed by viewers to verify its safety and appropriateness. However, conventional systems lack the means to efficiently analyze large amounts of content, making it difficult to provide effective feedback that reflects viewers' emotional state. Furthermore, content improvement and risk assessment based on viewer feedback are also inadequate. To solve these problems, an efficient content analysis system that combines an emotion engine and generative AI is needed.

[1560] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1561] In this invention, the server includes means for analyzing risk factors of viewing content using generative artificial intelligence, means for analyzing viewer feedback using an emotion engine to identify an emotional state, and means for providing recommendations to the viewer based on the emotion data and the risk assessment results, thereby making it possible to automatically evaluate the safety and appropriateness of viewing content and provide feedback that reflects the viewer's emotional state.

[1562] "Generative artificial intelligence" refers to artificial intelligence techniques used to automatically analyze data and extract and assess risk factors.

[1563] "Risk factors" are elements that may have an impact on users, such as potential problems, inappropriate information, or violent expressions in the viewing content.

[1564] "Content" is a general term for various digital media provided to viewers, such as video, audio, images, and text.

[1565] A "viewer" is a user who uses a content distribution service to view content.

[1566] An "emotion engine" is a technology that analyzes the emotional state of a user from their feedback and reactions and reflects that in the system.

[1567] "Feedback" refers to information such as impressions, opinions, and evaluations provided by viewers regarding the content they watch.

[1568] A "dashboard" is an interface that allows users and viewers to check analysis results and risk assessments.

[1569] "Comprehensive risk assessment" is a procedure for assessing the safety and appropriateness of content based on analyzed risk factors and emotional data.

[1570] "Recommendations" are specific suggestions or advice provided by the system to improve viewing content or ensure viewer safety.

[1571] This invention is a system that improves the safety and quality of content distribution services by analyzing risk factors of viewing content and providing feedback based on the viewer's emotional state. Specific embodiments and program processing of this system are described below.

[1572] Overall system configuration

[1573] The system mainly consists of the following components:

[1574] Database server: Stores content data and risk data.

[1575] Generative AI module: Analyzes and evaluates risk factors of viewed content.

[1576] Emotion Engine: Analyzes viewer feedback and identifies their emotional state.

[1577] User Interface: The interface for viewers to provide feedback and see the analysis results.

[1578] Feedback module: Generates and provides feedback to viewers based on sentiment data and risk assessment.

[1579] Program processing description

[1580] 1. Data input

[1581] Users upload content to the system from their own devices via a dedicated interface. The server receives the uploaded content and stores it in a secure database, along with the content's metadata (creation date, creator, etc.).

[1582] 2. Initial Risk Analysis

[1583] The server then applies a generative AI module to the content stored in the database. The generative AI extracts keywords and phrases from the content and identifies potential risk factors, such as violent scenes or inappropriate language in a video.

[1584] 3. Error Detection and Analysis

[1585] The generative AI module drills down into specific risk areas and analyzes the context to identify specific issues, which then generates a specific risk report and provides risk details to the user.

[1586] 4. Emotion recognition

[1587] The emotion engine analyzes voice, text input, and facial expression data from viewers and sends the results to the server. For example, it can recognize the viewer's emotions (anxiety, relief, etc.) in real time when watching content.

[1588] 5. Reflecting Emotional Data

[1589] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[1590] 6. Providing Feedback

[1591] The user can check the analysis results and risk assessment through the user interface and input any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[1592] Specific examples

[1593] For example, consider a scenario where a video streaming company uploads a new documentary to their system:

[1594] 1. User uploads a video

[1595] Users upload new documentary videos to the system, tag the videos and select categories.

[1596] 2. The server receives and saves the video

[1597] The server stores the videos in a secure database and generates metadata.

[1598] 3. Generative AI performs initial risk analysis

[1599] The generative AI analyzes important items in the video, such as "violent scenes" and "inappropriate expressions," and extracts risk factors.

[1600] 4. Generative AI performs detailed analysis

[1601] Generative AI detects ambiguities and inconsistencies and identifies specific risks and errors, such as "This scene contains violent content."

[1602] 5. The server provides a comprehensive risk assessment

[1603] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1604] 6. Emotion engine analyzes emotions

[1605] The emotion engine analyzes the viewer's facial expressions and voice when watching a video and determines whether the viewer is feeling anxious.

[1606] 7. The server reflects the emotional data

[1607] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[1608] 8. User reviews the results and provides feedback

[1609] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1610] In this way, viewers can efficiently assess the risk of distributed content, recognize potential problems in advance, and take measures. In addition, by using the emotion engine, appropriate support can be provided according to the viewer's emotional state, further improving the accuracy and efficiency of risk management.

[1611] Prompt Sentence Examples

[1612] "This video contains violent content, but what specifically is the problem?"

[1613] "Tell us your thoughts after watching: How did the video make you feel?"

[1614] In this way, it will be possible to provide risk assessment and support based on viewer feedback, which is expected to improve the quality of content and the viewer experience.

[1615] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1616] Step 1:

[1617] A user uploads viewing content.

[1618] Input: File path of the content you want to watch

[1619] Output: Content data stored in the database

[1620] Specific behavior:

[1621] Users upload content to the server via a dedicated interface from their device, and the server stores the content data and metadata (e.g., creation date and creator) in a secure database.

[1622] Step 2:

[1623] The server uses a generative AI module to perform an initial risk analysis of the content.

[1624] Input: Content data stored in the database

[1625] Output: Initial risk report with risk factors extracted

[1626] Specific behavior:

[1627] The server retrieves the content from the database and uses a generative AI module to analyze keywords and phrases in the content, identifying risk factors such as violent scenes or inappropriate language, and generating an initial risk report.

[1628] Step 3:

[1629] The server uses a generated AI module to perform detailed error detection and analysis.

[1630] Input: Initial Risk Report

[1631] Output: Detailed risk report identifying specific risks and errors

[1632] Specific behavior:

[1633] Based on the initial risk report, the server uses a generative AI module to perform further analysis, digging deeper into ambiguities and inconsistencies to identify specific issues and risks and generate a detailed risk report.

[1634] Step 4:

[1635] An emotion engine analyzes viewer feedback to identify emotional states.

[1636] Input: Audio and text feedback from viewers

[1637] Output: Emotional state data

[1638] Specific behavior:

[1639] Viewers provide their thoughts and opinions through a feedback interface, and the server uses an emotion engine to analyze the audio and text data of the feedback to identify the viewer's emotional state (e.g., anxiety, relief).

[1640] Step 5:

[1641] The server incorporates the emotional data into risk assessment.

[1642] Input: Emotional state data, detailed risk report

[1643] Output: Comprehensive risk assessment results reflecting emotional data

[1644] Specific behavior:

[1645] The server then incorporates the emotional state data obtained from the emotion engine into a detailed risk report, adding risk solutions and supplemental information if the viewer is feeling anxious, and generating a comprehensive risk assessment result.

[1646] Step 6:

[1647] Provide comprehensive risk assessment results and recommendations, and gather user feedback.

[1648] Input: Overall risk assessment results displayed on the user interface

[1649] Output: More feedback from the user

[1650] Specific behavior:

[1651] The server generates a comprehensive risk assessment result and notifies the viewer through a user interface. The viewer can check the results on the dashboard and enter any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[1652] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1653] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1654] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1655] [Fourth embodiment]

[1656] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1657] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1658] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1659] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1660] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1661] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1662] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1663] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1664] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1665] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1666] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1667] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1668] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1669] This invention relates to a risk management system for business documents that utilizes generative artificial intelligence. This system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. Specific embodiments of the system and details of the program processing are described below.

[1670] Overall system configuration

[1671] The system mainly consists of the following components:

[1672] Database server: stores business documents and risk data

[1673] Generative AI module: Performs document analysis and risk assessment

[1674] User Interface: A user interface for uploading documents and viewing analysis results.

[1675] Feedback module: To collect user feedback and reflect it in the system.

[1676] Program processing description

[1677] 1. Data input

[1678] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1679] The server receives the uploaded documents and stores them in a secure database, along with the document's metadata (creation date, author, tag information, etc.).

[1680] 2. Initial Risk Analysis

[1681] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[1682] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[1683] 3. Error Detection and Analysis

[1684] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[1685] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[1686] 4. Centralized risk management

[1687] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[1688] The server generates a comprehensive risk assessment result and presents it to the user in a user interface.

[1689] 5. Providing Feedback

[1690] Users can check the analysis results and risk assessments through a dashboard, and if necessary, enter feedback and request further analysis or corrections.

[1691] The server stores the received feedback in a database and reflects it in the next analysis.

[1692] 6. Applying a Custom Model

[1693] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1694] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[1695] Specific examples

[1696] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1697] 1. User uploads a contract

[1698] Users upload new supply contracts into the system, tag the contracts and select categories.

[1699] 2. The server receives and saves the contract

[1700] The server stores the contract in a secure database and generates metadata.

[1701] 3. Generative AI performs initial risk analysis

[1702] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1703] 4. Generative AI performs detailed analysis

[1704] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1705] 5. The server provides a comprehensive risk assessment

[1706] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1707] 6. User reviews the results and provides feedback

[1708] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1709] In this way, users can streamline risk assessment of contracts, recognize potential issues in advance, and take action.

[1710] The processing flow will be explained below.

[1711] Step 1:

[1712] Users upload business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1713] Step 2:

[1714] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1715] Step 3:

[1716] The server invokes a generative AI module to analyze the content of the stored documents.

[1717] Step 4:

[1718] Generative AI extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[1719] Step 5:

[1720] Generative AI analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadlines.

[1721] Step 6:

[1722] The server creates an initial risk report based on the risk factors and detailed issues provided by the generation AI and stores it in a database.

[1723] Step 7:

[1724] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[1725] Step 8:

[1726] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1727] Step 9:

[1728] Users can review the risk assessment results on the dashboard and enter any necessary corrections or additional feedback.

[1729] Step 10:

[1730] The server receives feedback from the user and reflects it in the database, which will be used in the next analysis.

[1731] Step 11:

[1732] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[1733] Step 12:

[1734] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs.

[1735] Step 13:

[1736] The server notifies the user of the results of applying the custom settings and provides improvements and recommendations.

[1737] Through these specific processing steps, the system quickly and accurately analyzes business documents, improving the efficiency and accuracy of risk management.

[1738] Example 1

[1739] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1740] This invention relates to a system that automatically extracts and analyzes risk factors in business documents that have been difficult to detect efficiently using conventional methods, and provides comprehensive risk assessments. Specifically, it improves business efficiency by providing a method for analyzing large volumes of business documents, identifying and assessing risk factors, and presenting appropriate risk information and recommended countermeasures to users. Another objective of this invention is to provide a means for supporting risk assessments specific to specific industries.

[1741] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1742] In this invention, the server includes: a means for users to upload business documents through a dedicated interface and store them in a secure database; a means for performing an initial risk analysis of the uploaded documents using a generative AI model and extracting keywords and phrases to identify risk factors; a means for performing contextual analysis of the identified risk factors using the generative AI model to detect and analyze specific errors; a means for performing a comprehensive risk assessment by comparing the results with a historical risk database and presenting a risk report to the user; a means for receiving feedback from the user, storing the feedback in a database and reflecting it in the system; and a means for receiving custom model settings and reflecting them in the generative AI model to perform a risk assessment specific to a specific industry. This makes it possible to efficiently extract and analyze risk factors from a large number of business documents and provide users with a comprehensive risk assessment and specific recommended measures.

[1743] "Business documents" refer to documents created and used by companies and organizations to carry out their business. Examples include contracts, work procedures, and reports.

[1744] A "generative AI model" refers to a type of artificial intelligence that uses machine learning and natural language processing to analyze text data and generate information.

[1745] "Risk factors" refer to keywords or phrases contained in business documents that could be potential risk factors, which could affect business operations.

[1746] "Initial risk analysis" refers to the initial analysis process using a generative AI model to extract keywords and phrases contained in business documents and identify risk factors.

[1747] "Contextual analysis" refers to the process of analyzing the text and context of documents containing identified risk factors to identify specific errors or issues.

[1748] "Database" refers to a system for systematically storing and managing business documents and risk-related information.

[1749] "Feedback" refers to comments and additional information provided by users based on the analysis results, which are reflected in improving the accuracy of the system and in the next analysis.

[1750] A "custom model" refers to a model that is configured and adjusted specifically for a specific industry or application, allowing the generative AI model to perform analysis tailored to those specific needs.

[1751] "Risk report" refers to a report that comprehensively summarizes the results of the initial risk analysis and context analysis, and is provided to the user.

[1752] "Dashboard" refers to the web application interface that allows users to view system analysis results, risk assessments, and feedback.

[1753] The business document risk management system of this invention is mainly composed of three entities: a server, a terminal, and a user. The entire system is realized by the main components such as a database server, a generation AI module, a user interface, and a feedback module.

[1754] System hardware and software examples

[1755] The system's database server is used to store business documents and risk data. This server should provide a secure and reliable data storage solution. Specifically, we recommend using an SQL database (e.g., MySQL, PostgreSQL) or a cloud storage service (e.g., Amazon S3).

[1756] The generative AI module includes models for document analysis and risk assessment. This module incorporates natural language processing (NLP) and machine learning (ML) techniques, such as generative AI models like GPT-3 and BERT.

[1757] The user interface is a tool that allows users to upload business documents and check the analysis results. This interface is implemented as a web application that can be accessed through a web browser and uses technologies such as HTML, CSS, and JavaScript.

[1758] The feedback module collects feedback from users and reflects it in the system. This module receives feedback data through form input or API submission and stores it in a database.

[1759] Specific operation of the system

[1760] Data Input

[1761] Users upload their company's business documents to the system through a dedicated interface. This involves clicking the document upload button and selecting the file. The server receives the uploaded document and stores it in a secure database. Document metadata (creation date, author, tag information) is also saved.

[1762] Initial Risk Analysis

[1763] The server applies a generative AI module to the stored documents. The generative AI model (e.g., GPT-3) extracts keywords and phrases from the documents and identifies potential risk factors. For example, in the case of a contract, important items such as "delivery date" and "quality standards" are analyzed.

[1764] Error Detection and Analysis

[1765] The generative AI module performs contextual analysis on the identified risk factors to identify specific errors or issues, such as vague quality requirements or unclear delivery deadlines, and reports the details of the risk to the user.

[1766] Centralized risk management

[1767] The server compares the latest risk report with the existing risk database to see if similar risks have occurred in the past, and then compiles the relevant information to make a comprehensive assessment and present it to the user via a user interface.

[1768] Providing feedback

[1769] Users can check the analysis results and risk assessment on the dashboard, and if necessary, enter feedback to request further analysis or corrections. The server stores the received feedback in a database and reflects it in the next analysis.

[1770] Applying a Custom Model

[1771] On the device (such as a user's PC), custom settings specific to the industry are configured. For example, settings that emphasize quality control items for the manufacturing industry are configured. The server receives the custom settings and reflects them in the generated AI model. This enables risk assessment tailored to the needs of specific industries during analysis.

[1772] Examples of concrete examples and prompts

[1773] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1774] 1. User uploads a contract

[1775] Users upload new supply contracts into the system, tag the contracts and select categories.

[1776] 2. The server receives and saves the contract

[1777] The server stores the contract in a secure database and generates metadata.

[1778] 3. Generative AI performs initial risk analysis

[1779] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1780] 4. Generative AI performs detailed analysis

[1781] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1782] 5. The server provides a comprehensive risk assessment

[1783] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1784] 6. User reviews the results and provides feedback

[1785] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1786] In this way, users can streamline risk assessment of contracts and proactively identify and address potential issues. Specific examples of prompts include:

[1787] "I have uploaded a new supply contract. Please conduct an initial risk analysis and detailed analysis of this contract to identify risk factors related to delivery dates and quality standards."

[1788] Following this prompt, the system analyzes the contract, extracts risk factors, and provides feedback to the user, enabling quick and accurate risk management of business documents.

[1789] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1790] Step 1:

[1791] Uploading business documents

[1792] Users upload their company's business documents to the system through a dedicated interface. Specifically, users open a web browser, click the document upload button, and then specify the file in the file selection dialog or upload it by dragging and dropping.

[1793] Input: Business document file (e.g. PDF, Word)

[1794] Output: The uploaded document file is sent to the server.

[1795] Step 2:

[1796] Document storage and metadata registration

[1797] The server receives the uploaded document and stores it in a secure database, extracting the document's metadata (creation date, author, tag information) and registering it in the database.

[1798] Input: Uploaded document file

[1799] Data processing: file format check, metadata extraction

[1800] Output: Document files and metadata stored in a database

[1801] Step 3:

[1802] Initial Risk Analysis

[1803] The server applies a generative AI module (e.g., GPT-3) to the stored documents to extract keywords and phrases within the documents and identify potential risk factors. For example, in the case of a contract, it analyzes important items such as "delivery date" and "quality standards."

[1804] Input: Document files stored in the database

[1805] Data computation: Keyword extraction and risk factor identification using generative AI models

[1806] Output: Risk factor list

[1807] Step 4:

[1808] Detailed Risk Analysis

[1809] The generative AI module digs deeper into the identified risk factors and performs contextual analysis to identify specific issues or errors, leading to risk details such as "clarification of quality standards is required."

[1810] Input: Risk factor list, document text data

[1811] Data processing: Reanalysis of the context in which risk factors exist and identification of specific problems

[1812] Output: Detailed risk report

[1813] Step 5:

[1814] Centralized Risk Assessment

[1815] Based on the results of the generated AI, the server compares the latest risk report with the existing risk database to check whether similar risks have occurred in the past, and performs a comprehensive risk assessment and compiles the results into a risk report.

[1816] Inputs: Detailed risk reports, existing risk database

[1817] Data calculation: Database matching, calculation of comprehensive risk assessment

[1818] Output: Comprehensive risk report

[1819] Step 6:

[1820] Present risk assessment to users

[1821] The server presents a comprehensive risk report on a user interface, and the user can check the risk assessment results on a dashboard. Specifically, the risk report is displayed on the dashboard of a web application.

[1822] Input: Comprehensive Risk Report

[1823] Output: Risk report displayed in the user interface

[1824] Step 7:

[1825] User feedback

[1826] Users can check the analysis results and risk assessment on the dashboard, enter feedback as needed, and request further analysis or corrections.

[1827] Input: Risk report analysis results, feedback

[1828] Output: The received feedback is sent to the server

[1829] Step 8:

[1830] Feedback is saved and implemented

[1831] The server stores the feedback received from users in a database and reflects it in the next analysis. The feedback data is used to adjust the generative AI model.

[1832] Input: Feedback received

[1833] Data processing: saving feedback data and retuning the AI ​​model

[1834] Output: Updated AI model and feedback data

[1835] Step 9:

[1836] Configuring a Custom Model

[1837] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1838] Input: Custom setting item

[1839] Output: Custom settings are sent to the server

[1840] Step 10:

[1841] Reflecting custom settings

[1842] The server receives the custom configuration and feeds it into the generative AI model, which, upon analysis, tailors the risk assessment to specific industry needs.

[1843] Input: Custom setting item

[1844] Data processing: Adjusting the AI ​​model

[1845] Output: AI model with custom industry-specific settings

[1846] Through the above processing steps, this system can efficiently and accurately manage risks in business documents.

[1847] (Application example 1)

[1848] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1849] Traditionally, it has been extremely difficult to manually manage large volumes of business documents (e.g., delivery contracts, transportation regulations, customer requests) in logistics operations and identify risk factors. In particular, photographing documents, entering metadata, identifying and evaluating risk factors, and collecting and incorporating feedback were inefficient, often resulting in human error and time loss. This made it impossible to recognize potential risks in advance and take measures, leading to reduced operational efficiency and ultimately to a decline in service quality and financial losses.

[1850] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1851] In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database to perform a comprehensive risk assessment, means for presenting risk information and recommended countermeasures to users, means for collecting user feedback and reflecting it in the system, means for performing risk assessment specific to a specific industry by setting a custom model, terminal means for photographing business documents and inputting metadata, means for providing document analysis results on a dashboard, and means for users to input feedback based on the analysis results. This enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[1852] "Generative AI" refers to AI that has the ability to analyze and learn large amounts of text data and automatically generate new data and information.

[1853] "Business documents" are documents related to business within an organization or with external parties, including contracts, regulations, procedures, and customer requests.

[1854] "Risk factors" refer to elements that may cause potential risks, such as unclear clauses or uncertain elements contained in business documents.

[1855] A "risk database" is a database that accumulates and stores past business documents and risk assessment data, and is used to help with future risk assessments.

[1856] "Custom Model" refers to a risk assessment model that is customized to fit the needs of a specific industry or company.

[1857] "Feedback" refers to opinions and suggestions provided by users regarding analysis results and system usage, which are used to improve and update the system.

[1858] "Terminal" refers to a device (e.g., smartphone, tablet) used by a user to photograph, input, and upload business documents.

[1859] "Dashboard" refers to an interface that allows users to visually check analysis results and risk assessment information.

[1860] "Photographing" refers to taking an image of a business document using a digital camera or the camera function of a smartphone.

[1861] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings.

[1862] System Overview

[1863] The system of the present invention mainly comprises the following components:

[1864] 1. Server:

[1865] It includes a database server that stores business documents and risk data, a generative AI module, and a feedback module.

[1866] 2. Terminal:

[1867] A device (e.g., smartphone, tablet) that a user uses to capture, type, or upload business documents.

[1868] 3. User Interface:

[1869] A dashboard for users to view analysis results and risk assessments.

[1870] Program processing explanation

[1871] 1. Data input:

[1872] A user takes a photo of a business document using a terminal and inputs metadata (e.g., creation date and time, creator, tag information). The data is then sent from the terminal to the server.

[1873] The server stores the received business documents and metadata in a secure database.

[1874] 2. Initial risk analysis:

[1875] The server applies a generative AI module to the stored documents to extract risk factors. The generative AI analyzes keywords and phrases in the documents to identify potential risk factors (e.g., unclear clauses, delivery dates, quality standards).

[1876] 3. Risk Assessment and Dashboard View:

[1877] The server compares the extracted risk factors with a historical risk database to perform a comprehensive risk assessment, and the results are presented to the user through a user interface.

[1878] 4. Gather and incorporate feedback:

[1879] Users can check the analysis results displayed on the dashboard and enter feedback as needed, which is then sent to the system and stored on the server.

[1880] The server incorporates the collected feedback into the next risk analysis.

[1881] Hardware and software usage

[1882] Hardware:

[1883] Smartphone or tablet (used for taking photos and entering business documents)

[1884] Cloud server (used for data storage and analysis)

[1885] software:

[1886] Flask (Python web framework)

[1887] Generative AI module (used for document risk analysis)

[1888] Database module (used to store data)

[1889] Specific examples

[1890] For example, here's the scenario when a major logistics company wants to upload a new transportation contract into their system:

[1891] 1. Upload the contract:

[1892] Users take a photo of the shipping contract with their smartphone and upload it to the app, while also entering metadata such as the shipping date and the name of the carrier.

[1893] 2. Risk factor analysis:

[1894] The server uses a generative AI module to extract risk factors such as "transportation period" and "liability" from the contract.

[1895] 3. Providing risk assessments:

[1896] The extracted risk factors are displayed on the user's dashboard and compared with past risk data for evaluation.

[1897] 4. Enter your feedback:

[1898] The user checks the analysis results and inputs feedback such as "The terms of liability are unclear." The feedback is sent to the system and reflected in the next analysis.

[1899] Prompt Sentence Examples

[1900] "Analyze the risk factors in the transportation contract, especially those related to the duration of transportation and liability."

[1901] This invention enables efficient document risk management in logistics operations, early detection of potential risks, and the implementation of accurate countermeasures.

[1902] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1903] Step 1:

[1904] Users take photos and upload business documents

[1905] The user takes a photo of a business document using a device (smartphone or tablet), enters the document's metadata (e.g., creation date and time, author, tag information) along with the image, and uploads it to the server via the app. This sends the business document image and its metadata to the server.

[1906] Step 2:

[1907] The server stores business documents and metadata

[1908] The server stores the received business document images and metadata in a secure database. The stored information includes the document content, creation date and time, author, tags, etc. This data storage ensures that the information necessary for analysis is secured.

[1909] Step 3:

[1910] The server applies the generated AI module to extract risk factors.

[1911] The server applies a generative AI module to the stored business document images. The generative AI module analyzes keywords and phrases (e.g., delivery dates, quality standards) within the document to extract potential risk factors. Based on the input document content, data processing and extraction of risk factors are performed. A list of risk factors is generated as the output.

[1912] Step 4:

[1913] The server compares the extracted risk factors with the risk database and performs a comprehensive evaluation.

[1914] The server compares the generated risk factor list with past risk data. The risk database stores past risk assessment data, and the server compares this data to evaluate the weighting and priority of risk factors. A comprehensive risk assessment is performed based on the input risk factor list, and the assessment results are output.

[1915] Step 5:

[1916] The server displays the risk assessment results on a dashboard.

[1917] The server displays the assessed risk factors and their details on a user interface (dashboard), where risk information and recommended measures are visually presented. The content displayed on the dashboard is generated and provided based on the input risk assessment results.

[1918] Step 6:

[1919] User enters feedback

[1920] The user checks the risk assessment results displayed on the dashboard and enters feedback as necessary. Specific examples (e.g., the liability conditions are unclear) are entered as feedback, and this is sent to the server.

[1921] Step 7:

[1922] The server collects feedback and applies it to the next analysis.

[1923] The server stores the received feedback in a database. The feedback is reflected in the next risk analysis and helps improve the analysis accuracy of the generative AI module. The database is updated based on the input feedback and reflected in the next analysis process.

[1924] In this way, the system enables efficient analysis and risk assessment of business documents through an automated risk management process.

[1925] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1926] This invention relates to a business document risk management system that utilizes generative artificial intelligence and an emotion engine. The system analyzes a large number of business documents, extracts risk factors, evaluates them, and provides feedback. It also grasps the user's emotional state and reflects it in the risk assessment. A specific embodiment of the system and details of the program processing are described below.

[1927] Overall system configuration

[1928] The system mainly consists of the following components:

[1929] Database server: stores business documents and risk data

[1930] Generative AI module: Performs document analysis and risk assessment

[1931] Emotion Engine: Recognize and analyze the user's emotional state

[1932] User Interface: A user interface for uploading documents and viewing analysis results.

[1933] Feedback module: To collect user feedback and reflect it in the system.

[1934] Program processing description

[1935] 1. Data input

[1936] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1937] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1938] 2. Initial Risk Analysis

[1939] The server applies a generative AI module to the stored documents, which extracts keywords and phrases within the documents and identifies potential risk factors.

[1940] For example, important items such as "delivery date" and "quality standards" included in the contract are analyzed, and risk factors are listed.

[1941] 3. Error Detection and Analysis

[1942] The generative AI module digs deep into specific risk areas within a document, analyzing the context to identify specific issues.

[1943] For example, it detects ambiguous quality requirements and unclear delivery deadline descriptions and reports details of the risks to the user.

[1944] 4. Centralized risk management

[1945] The server compares the latest risk report with the existing risk database, checking whether similar risks have occurred in the past and compiling relevant information.

[1946] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1947] 5. Emotion recognition

[1948] The emotion engine analyzes emotions from the user's voice, text input, and facial expression data, and sends the results to the server.

[1949] For example, it recognizes the emotions (anxiety, relief, etc.) of users in real time when they review risk reports.

[1950] 6. Reflecting Emotional Data

[1951] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[1952] 7. Providing Feedback

[1953] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[1954] The server stores the received feedback in a database and reflects it in the next analysis.

[1955] 8. Applying Custom Models

[1956] Customizing the device (user's PC, etc.) to suit the industry. For example, configuring the device to prioritize quality control items for the manufacturing industry.

[1957] The server receives the custom configuration and feeds it into the generative AI, which then tailors the risk assessment to specific industry needs during analysis.

[1958] 9. Notification of risk assessment results and provision of countermeasures

[1959] The server notifies the user of the risk assessment results that reflect the custom settings and emotional data, and provides remediation measures and recommendations.

[1960] Specific examples

[1961] For example, consider a scenario where a large manufacturing company uploads a new supply contract into their system:

[1962] 1. User uploads a contract

[1963] Users upload new supply contracts into the system, tag the contracts and select categories.

[1964] 2. The server receives and saves the contract

[1965] The server stores the contract in a secure database and generates metadata.

[1966] 3. Generative AI performs initial risk analysis

[1967] The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors.

[1968] 4. Generative AI performs detailed analysis

[1969] Generative AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors, such as "quality standards need to be clarified."

[1970] 5. The server provides a comprehensive risk assessment

[1971] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[1972] 6. Emotion engine analyzes emotions

[1973] The emotion engine analyzes the user's facial expressions and voice when checking the risk report and determines that the user is feeling anxious.

[1974] 7. The server reflects the emotional data

[1975] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[1976] 8. User reviews the results and provides feedback

[1977] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[1978] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[1979] The processing flow will be explained below.

[1980] Step 1:

[1981] Users upload their company's business documents (e.g., work procedures, contracts, etc.) to the system through a dedicated interface.

[1982] Step 2:

[1983] The server receives the uploaded documents and stores them in a secure database, along with document metadata (creation date, author, tag information, etc.).

[1984] Step 3:

[1985] The server invokes a generative AI module to analyze the content of the stored documents.

[1986] Step 4:

[1987] The generative AI module extracts keywords and phrases from documents to identify potential risk factors, such as listing important items like "delivery date" and "quality standards."

[1988] Step 5:

[1989] The generative AI module analyzes risk factors in more detail and understands the context to identify specific issues, such as vague quality requirements or unclear delivery deadline statements.

[1990] Step 6:

[1991] The server creates an initial risk report based on the risk information and issues obtained by the generative AI module and stores it in a database.

[1992] Step 7:

[1993] The server compares the latest risk report with a historical risk database to see if similar risks or errors have occurred.

[1994] Step 8:

[1995] The server generates a comprehensive risk assessment result and notifies the user through a user interface.

[1996] Step 9:

[1997] The emotion engine collects the user's voice, text input, and facial expression data in real time through the user interface.

[1998] Step 10:

[1999] The emotion engine analyzes the collected data and recognizes the user's emotions (e.g., anxiety, relief).

[2000] Step 11:

[2001] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, if the user feels anxious, it may provide additional information or specific solutions to alleviate that anxiety.

[2002] Step 12:

[2003] Users can review the analysis results and risk assessments through a dashboard and enter any necessary corrections or additional feedback.

[2004] Step 13:

[2005] The server receives user feedback, records it in a database, and uses it in the next risk assessment.

[2006] Step 14:

[2007] The device can be configured with industry-specific custom settings, such as those that prioritize quality control items for the manufacturing industry.

[2008] Step 15:

[2009] The server receives the custom configuration and feeds it into the generative AI, ensuring that the risk assessment meets industry-specific requirements.

[2010] Step 16:

[2011] The server generates a further optimized risk assessment result based on the custom settings and emotion data and notifies the user.

[2012] This allows users to efficiently and accurately assess the risks of business documents, enabling them to detect potential problems early and take measures.In addition, the emotion engine provides appropriate support according to the user's emotional state, improving the reliability of risk management and user satisfaction.

[2013] Example 2

[2014] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2015] Conventional risk management systems do not ensure sufficient accuracy and efficiency in risk assessment of business documents. Furthermore, they lack the functionality to reflect the user's emotional state in the risk assessment, which can prevent them from providing appropriate support. Furthermore, they lack flexibility when conducting risk assessments specific to specific industries.

[2016] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for analyzing a large number of business documents using generative artificial intelligence and extracting risk factors, means for comparing the results with a past risk database and performing a comprehensive risk assessment, means for presenting risk information and recommended measures to the user, means for collecting feedback from the user and reflecting it in the system, means for performing a risk assessment specific to a specific industry by setting a custom model, and means for analyzing the user's emotional state and reflecting it in the risk assessment. This makes it possible to improve the accuracy and efficiency of risk assessment of business documents and provide appropriate support according to the user's emotional state.

[2017] "Generative AI" is AI that uses natural language processing and machine learning techniques to analyze text data and generate meaningful information.

[2018] "Business documents" refers to all documents created to carry out the daily work of a company or organization, including, for example, contracts and work procedures.

[2019] "Risk factors" refer to elements or items contained in documents or data that may cause problems or trouble.

[2020] A "risk database" is a database that collects information on risks that have occurred in the past and related data.

[2021] "Comprehensive risk assessment" is a comprehensive assessment based on extracted risk factors and compared with a past risk database.

[2022] "User" refers to the person who uses the system to upload business documents and receive the results of risk assessments.

[2023] "Feedback" refers to information such as evaluation information, improvements, and suggested corrections provided by users.

[2024] A "Custom Model" is an analytical model tailored to the unique needs and characteristics of a particular industry.

[2025] "Emotional state" refers to the psychological and emotional state a user is feeling at a particular moment.

[2026] An "emotion engine" is a technology or module that analyzes a user's voice, text, and facial expression data to recognize and evaluate their emotional state.

[2027] A "dashboard" refers to a graphical user interface that allows users to check system analysis results and risk assessment information at a glance.

[2028] "Error detection" refers to the process of identifying ambiguous statements and inconsistent items in business documents and pointing them out as specific problems.

[2029] The present invention relates to a risk management system for business documents that utilizes generative AI and an emotion engine. This system analyzes a large number of business documents, extracts risk factors, evaluates them, provides feedback, and grasps the user's emotional state, which is reflected in the risk assessment. Specific embodiments of the present invention are described below.

[2030] System Configuration

[2031] The system consists of the following main components:

[2032] Database server: Stores business documents and risk data.

[2033] Generative AI module: Performs document analysis and risk assessment.

[2034] Emotion Engine: Recognizes and analyzes the user's emotional state.

[2035] User Interface: The interface through which users can upload documents and view the analysis results.

[2036] Feedback module: Collects feedback from users and reflects it in the system.

[2037] Hardware and Software

[2038] Server: Hardware that stores and processes data. A high-performance server is desirable.

[2039] Generative AI: Software that uses natural language processing and machine learning techniques, such as models like BERT and GPT.

[2040] Emotion Engine: Software that analyzes a user's voice, text, and facial expression data to recognize their emotional state. For example, using the Emotion API or face recognition software.

[2041] Database: A system for storing risk data and business document metadata, such as a relational database like MySQL or PostgreSQL.

[2042] User interface: An interface for checking analysis results and feedback. These are primarily web-based applications, using JavaScript frameworks (React, Vue.js, etc.).

[2043] Emotion recognition and feedback reflection

[2044] The emotion engine analyzes facial, voice, and text data in real time as users review risk reports to recognize their emotional state (e.g., anxiety, relief).

[2045] The server then incorporates the emotion data into the analysis results and provides additional support information or solutions according to the user's emotional state. For example, if the user is feeling anxious, the server may provide more detailed explanations or additional materials.

[2046] Gathering and implementing feedback

[2047] Users can check the analysis results and risk assessments from the dashboard and enter any necessary corrections or additional feedback.

[2048] The server stores the feedback received from users in a database and reflects it in the next analysis, allowing the system to continuously improve and provide more accurate risk assessments.

[2049] Specific examples

[2050] For example, consider a scenario in which a large manufacturing company uploads a new supply contract to the system. The user uploads the new supply contract to the system, tags the contract, and selects a category. The server stores the contract in a secure database and generates metadata. The generation AI analyzes important items in the contract, such as "delivery date" and "quality standards," and extracts risk factors. The generation AI detects ambiguous clauses and inconsistencies and identifies specific risks and errors. The server performs a comprehensive evaluation and notifies the user through the user interface. At this time, the emotion engine analyzes the user's emotions and provides appropriate additional information.

[2051] Prompt Sentence Examples

[2052] The generative AI model performs a risk assessment of a specific business document by providing prompts such as:

[2053] "Please identify the risk factors contained in this contract and conduct a specific risk assessment. Please analyze with particular emphasis on important items such as delivery dates, quality standards, and payment terms."

[2054] In this way, users can streamline risk assessment of contracts, recognize potential problems in advance, and take measures. By using the emotion engine, it is possible to provide appropriate support according to the user's emotional state, further improving the accuracy and efficiency of risk management.

[2055] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2056] Step 1: Data Input

[2057] Users upload their company's business documents to the system through a dedicated interface. Specifically, they click the "Upload" button on the user interface and select the business document file.

[2058] Input: Business document file, metadata (creation date and time, creator, tag information)

[2059] The server receives the uploaded documents and stores them in a secure database, along with metadata about the documents.

[2060] Output: Document files and metadata stored in a database

[2061] Step 2: Initial risk analysis

[2062] The server applies the generative AI module to documents stored in the database, using AI analysis algorithms to extract keywords and phrases within the documents and identify risk factors.

[2063] Input: Document files in the database

[2064] The server lists potential risk factors within the document based on the keywords and phrases extracted by the generative AI module.

[2065] Output: List of identified risk factors

[2066] Step 3: Error detection and analysis

[2067] The generative AI module digs deep into specific risk areas within a document, analyzes the context, and identifies specific issues, generating error messages such as "Delivery date unclear" or "Quality standards unclear."

[2068] Input: List of identified risk factors, document files in the database

[2069] The generative AI module analyzes the context of the document and identifies specific error locations.

[2070] Output: Detailed risk report with specific error messages

[2071] Step 4: Centralized risk management

[2072] The server compares the risk report created by the generative AI module with an existing risk database, checking whether similar risks have occurred in the past and adding any relevant information.

[2073] Input: Detailed risk report, existing risk database

[2074] The server generates a comprehensive risk assessment and stores the information in a database.

[2075] Output: Comprehensive risk assessment results

[2076] Step 5: Emotion Recognition

[2077] The emotion engine analyzes the user's voice, text input, and facial expression data to recognize their emotional state. For example, it captures the user's facial expressions with a camera and uses a facial expression recognition algorithm to determine their emotions.

[2078] Input: User's voice data, text data, facial expression data

[2079] The emotion engine analyzes this data and sends the results to the server.

[2080] Output: Parsed emotion data

[2081] Step 6: Reflecting emotional data

[2082] The server then converts the emotional data from the emotion engine into a detailed risk report, providing more specific explanations and additional resources if the user is feeling anxious, for example.

[2083] Inputs: Analyzed sentiment data, detailed risk reports

[2084] The server adds additional support information to the risk report according to the emotion data.

[2085] Output: Final risk report reflecting sentiment data

[2086] Step 7: Provide feedback

[2087] Users can review the final risk report from the dashboard and enter feedback such as comments and correction requests.

[2088] Input: User feedback data

[2089] The server stores the feedback received from the user in a database and reflects it in the next risk assessment.

[2090] Output: Feedback data stored in a database

[2091] Step 8: Applying the custom model

[2092] Terminals (such as users' PCs) are custom configured for specific industries, such as manufacturing, to strictly evaluate quality standards.

[2093] Input: User's custom setting data

[2094] The server receives the custom settings and reflects them in the generated AI.

[2095] Output: A generative AI model with your custom settings

[2096] Step 9: Communicate risk assessment results and provide countermeasures

[2097] The server notifies the user of the final risk assessment result, which reflects the custom settings and emotion data, by displaying the result on a dashboard through the user interface.

[2098] Inputs: Custom Settings, Emotion Data, Final Risk Report

[2099] The server notifies the user of the risk assessment results to provide appropriate remediation measures and recommendations.

[2100] Output: Notified risk assessment results and countermeasures

[2101] (Application example 2)

[2102] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2103] Modern content distribution services require automatic analysis of the content being viewed by viewers to verify its safety and appropriateness. However, conventional systems lack the means to efficiently analyze large amounts of content, making it difficult to provide effective feedback that reflects viewers' emotional state. Furthermore, content improvement and risk assessment based on viewer feedback are also inadequate. To solve these problems, an efficient content analysis system that combines an emotion engine and generative AI is needed.

[2104] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2105] In this invention, the server includes means for analyzing risk factors of viewing content using generative artificial intelligence, means for analyzing viewer feedback using an emotion engine to identify an emotional state, and means for providing recommendations to the viewer based on the emotion data and the risk assessment results, thereby making it possible to automatically evaluate the safety and appropriateness of viewing content and provide feedback that reflects the viewer's emotional state.

[2106] "Generative artificial intelligence" refers to artificial intelligence techniques used to automatically analyze data and extract and assess risk factors.

[2107] "Risk factors" are elements that may have an impact on users, such as potential problems, inappropriate information, or violent expressions in the viewing content.

[2108] "Content" is a general term for various digital media provided to viewers, such as video, audio, images, and text.

[2109] A "viewer" is a user who uses a content distribution service to view content.

[2110] An "emotion engine" is a technology that analyzes the emotional state of a user from their feedback and reactions and reflects that in the system.

[2111] "Feedback" refers to information such as impressions, opinions, and evaluations provided by viewers regarding the content they watch.

[2112] A "dashboard" is an interface that allows users and viewers to check analysis results and risk assessments.

[2113] "Comprehensive risk assessment" is a procedure for assessing the safety and appropriateness of content based on analyzed risk factors and emotional data.

[2114] "Recommendations" are specific suggestions or advice provided by the system to improve viewing content or ensure viewer safety.

[2115] This invention is a system that improves the safety and quality of content distribution services by analyzing risk factors of viewing content and providing feedback based on the viewer's emotional state. Specific embodiments and program processing of this system are described below.

[2116] Overall system configuration

[2117] The system mainly consists of the following components:

[2118] Database server: Stores content data and risk data.

[2119] Generative AI module: Analyzes and evaluates risk factors of viewed content.

[2120] Emotion Engine: Analyzes viewer feedback and identifies their emotional state.

[2121] User Interface: The interface for viewers to provide feedback and see the analysis results.

[2122] Feedback module: Generates and provides feedback to viewers based on sentiment data and risk assessment.

[2123] Program processing description

[2124] 1. Data input

[2125] Users upload content to the system from their own devices via a dedicated interface. The server receives the uploaded content and stores it in a secure database, along with the content's metadata (creation date, creator, etc.).

[2126] 2. Initial Risk Analysis

[2127] The server then applies a generative AI module to the content stored in the database. The generative AI extracts keywords and phrases from the content and identifies potential risk factors, such as violent scenes or inappropriate language in a video.

[2128] 3. Error Detection and Analysis

[2129] The generative AI module drills down into specific risk areas and analyzes the context to identify specific issues, which then generates a specific risk report and provides risk details to the user.

[2130] 4. Emotion recognition

[2131] The emotion engine analyzes voice, text input, and facial expression data from viewers and sends the results to the server. For example, it can recognize the viewer's emotions (anxiety, relief, etc.) in real time when watching content.

[2132] 5. Reflecting Emotional Data

[2133] The server incorporates the emotional data obtained from the emotion engine into the risk assessment, for example, providing additional risk solutions or supplemental information if the user feels anxious.

[2134] 6. Providing Feedback

[2135] The user can check the analysis results and risk assessment through the user interface and input any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[2136] Specific examples

[2137] For example, consider a scenario where a video streaming company uploads a new documentary to their system:

[2138] 1. User uploads a video

[2139] Users upload new documentary videos to the system, tag the videos and select categories.

[2140] 2. The server receives and saves the video

[2141] The server stores the videos in a secure database and generates metadata.

[2142] 3. Generative AI performs initial risk analysis

[2143] The generative AI analyzes important items in the video, such as "violent scenes" and "inappropriate expressions," and extracts risk factors.

[2144] 4. Generative AI performs detailed analysis

[2145] Generative AI detects ambiguities and inconsistencies and identifies specific risks and errors, such as "This scene contains violent content."

[2146] 5. The server provides a comprehensive risk assessment

[2147] The server performs a comprehensive evaluation of the generated risk report, compares it with past risk data, and notifies the user.

[2148] 6. Emotion engine analyzes emotions

[2149] The emotion engine analyzes the viewer's facial expressions and voice when watching a video and determines whether the viewer is feeling anxious.

[2150] 7. The server reflects the emotional data

[2151] The server provides the user with additional risk solutions and supplemental information that reflects the emotional data.

[2152] 8. User reviews the results and provides feedback

[2153] Users can check the analysis results on the dashboard and provide any necessary corrections or additional feedback to be reflected in the next analysis.

[2154] In this way, viewers can efficiently assess the risk of distributed content, recognize potential problems in advance, and take measures. In addition, by using the emotion engine, appropriate support can be provided according to the viewer's emotional state, further improving the accuracy and efficiency of risk management.

[2155] Prompt Sentence Examples

[2156] "This video contains violent content, but what specifically is the problem?"

[2157] "Tell us your thoughts after watching: How did the video make you feel?"

[2158] In this way, it will be possible to provide risk assessment and support based on viewer feedback, which is expected to improve the quality of content and the viewer experience.

[2159] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2160] Step 1:

[2161] A user uploads viewing content.

[2162] Input: File path of the content you want to watch

[2163] Output: Content data stored in the database

[2164] Specific behavior:

[2165] Users upload content to the server via a dedicated interface from their device, and the server stores the content data and metadata (e.g., creation date and creator) in a secure database.

[2166] Step 2:

[2167] The server uses a generative AI module to perform an initial risk analysis of the content.

[2168] Input: Content data stored in the database

[2169] Output: Initial risk report with risk factors extracted

[2170] Specific behavior:

[2171] The server retrieves the content from the database and uses a generative AI module to analyze keywords and phrases in the content, identifying risk factors such as violent scenes or inappropriate language, and generating an initial risk report.

[2172] Step 3:

[2173] The server uses a generated AI module to perform detailed error detection and analysis.

[2174] Input: Initial Risk Report

[2175] Output: Detailed risk report identifying specific risks and errors

[2176] Specific behavior:

[2177] Based on the initial risk report, the server uses a generative AI module to perform further analysis, digging deeper into ambiguities and inconsistencies to identify specific issues and risks and generate a detailed risk report.

[2178] Step 4:

[2179] An emotion engine analyzes viewer feedback to identify emotional states.

[2180] Input: Audio and text feedback from viewers

[2181] Output: Emotional state data

[2182] Specific behavior:

[2183] Viewers provide their thoughts and opinions through a feedback interface, and the server uses an emotion engine to analyze the audio and text data of the feedback to identify the viewer's emotional state (e.g., anxiety, relief).

[2184] Step 5:

[2185] The server incorporates the emotional data into risk assessment.

[2186] Input: Emotional state data, detailed risk report

[2187] Output: Comprehensive risk assessment results reflecting emotional data

[2188] Specific behavior:

[2189] The server then incorporates the emotional state data obtained from the emotion engine into a detailed risk report, adding risk solutions and supplemental information if the viewer is feeling anxious, and generating a comprehensive risk assessment result.

[2190] Step 6:

[2191] Provide comprehensive risk assessment results and recommendations, and gather user feedback.

[2192] Input: Overall risk assessment results displayed on the user interface

[2193] Output: More feedback from the user

[2194] Specific behavior:

[2195] The server generates a comprehensive risk assessment result and notifies the viewer through a user interface. The viewer can check the results on the dashboard and enter any necessary corrections or additional feedback. The server stores the received feedback in a database and reflects it in the next analysis.

[2196] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2197] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2198] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2199] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2200] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2201] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2202] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2203] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2204] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2205] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2206] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2207] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2208] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2209] 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.

[2210] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2211] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2212] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2213] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2214] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2215] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, a...

Claims

1. A means for analyzing a large amount of business documents and extracting risk factors using generative artificial intelligence; A means of conducting a comprehensive risk assessment by comparing it with a past risk database; means for presenting risk information and recommended measures to a user; A means of collecting user feedback and incorporating it into the system; A means to configure custom models to assess risk specific to a particular industry; A system including:

2. means for performing an initial risk analysis of the document using generative artificial intelligence to generate a risk report; a means of using generative artificial intelligence to perform detailed error detection and analysis; A means for performing a comprehensive risk assessment based on generative artificial intelligence and presenting it to a user in a dashboard; The system of claim 1 , comprising:

3. A means to configure custom risk models specific to a particular industry and prioritize risks using generative artificial intelligence; A means for uploading business documents from a user terminal; The system of claim 1 , comprising:

Images