System

Abstract

To provide a system.SOLUTION: A system includes: means for analyzing utterance content input by a user; means for comparing the analyzed utterance content with past utterance content of the user and a created document; and means for suggesting that contradictory content be corrected when content that contradicts the past utterance content and the created document is detected in the input utterance content.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 a description and related instruction sentence regarding 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] JP 2022-180282 A Summary of the Invention [Problem to be solved by the invention]

[0004] When a user posts something on a chat tool, there is a possibility that the user may post something that contradicts their own past statements or documents. However, it is difficult for a user to always keep track of all of their past statements and documents and to post statements that do not contradict them. [Means for solving the problem]

[0005] The present invention provides a means for receiving a user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, and a means for presenting the suggestions to the user, thereby enabling the user to easily check whether their own comments are consistent with past comments and created materials. [Brief description of the drawings]

[0006] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Diagram 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. FIG. [Diagram 3] FIG. 11 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Diagram 5] FIG. 13 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. 13 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 13 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. 2 is a sequence diagram showing a process flow of the data processing system in the first embodiment of the first form example. [Figure 12] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the application example 1 of the first embodiment. [Figure 13] FIG. 11 is a sequence diagram showing a process flow of the data processing system in the second embodiment of the second embodiment. [Figure 14] FIG. 13 is a sequence diagram showing a process flow of the data processing system in application example 2 of embodiment 2. [Figure 15] FIG. 13 is a sequence diagram showing a process flow of a data processing system in a third embodiment of the present invention. [Figure 16]FIG. 13 is a sequence diagram showing a process flow of the data processing system in the third application example of the third embodiment. [Figure 17] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in the first embodiment of the first form example when an emotion engine is combined. [Figure 18] FIG. 13 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1 of Form Example 1 when combined with an emotion engine. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0007] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0008] First, the terms used in the following description will be explained.

[0009] In the following embodiments, a signed processor (hereinafter simply referred to as a "processor") may be one arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be one 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), an APU (Accelerated Processing Unit), or a TPU (TENSOR PROCESSING UNIT (registered trademark)).

[0010] In the following embodiments, a signed RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0011] 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.

[0012] In the following embodiments, a communication I / F (Interface) with a code is an interface including a communication processor and an antenna. The communication I / F controls communication between multiple computers. An example of a communication standard applied to the communication I / F is a wireless communication standard including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0013] 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. In addition, in this specification, the same idea as "A and / or B" is also applied when three or more things are expressed by connecting them with "and / or."

[0014] [First embodiment]

[0015] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0016] 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.

[0017] 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 wide area network (WAN) and / or a local area network (LAN).

[0018] 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.

[0019] The reception device 38 includes a touch panel 38A and a microphone 38B, and receives user input. The touch panel 38A detects contact with a pointer (e.g., a pen or a finger) to receive user input by the touch of the pointer. 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.

[0020] 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 (e.g., voice and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs voice according to instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, an aperture, and a shutter, and an imaging element such as a Complementary Metal-Oxide-Semiconductor (CMOS) image sensor or a Charge Coupled Device (CCD) image sensor.

[0021] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54.

[0022] FIG. 2 shows an example of main functions of the data processing device 12 and the smart device 14.

[0023] As shown in Fig. 2, 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. The specific process program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific process program 56 from the storage 32, and executes the read specific process 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 process program 56 executed on the RAM 30.

[0024] 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.

[0025] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores a reception output program 60. The reception output program 60 is used together with the specific processing program 56 by the data processing system 10. The processor 46 reads out the reception output program 60 from the storage 50, and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0026] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0027] "Example 1"

[0028] In one embodiment of the present invention, the present invention is implemented as a plug-in for a chat tool. When a user makes a comment on the chat tool, the plug-in compares the content of the comment with past comments and created documents, and automatically checks for inconsistencies. Specifically, when a user inputs a new comment, the plug-in analyzes the content of the comment and compares it with past comments and created documents.

[0029] If a conflict is detected, the plugin will display a warning message to the user and suggest how to fix the conflict.

[0030] "Example 2"

[0031] Another embodiment of the present invention is implemented as an interactive tool that automatically checks for inconsistencies before a user inputs a new statement by comparing it with past statements and written materials. Specifically, before a user inputs a new statement, the tool analyzes past statements and written materials and checks for inconsistencies with the contents of the new statement. If an inconsistency is detected, the tool displays a warning message to the user and suggests that the user correct the inconsistent contents.

[0032] The process flow of each embodiment will be described below.

[0033] "Example 1"

[0034] Step 1: A user types a new message into the chat tool.

[0035] Step 2: The plugin analyzes the comment and compares it with previous comments and created materials.

[0036] Step 3: If any conflicts are detected, the plugin will display a warning message and suggest you fix the conflicts.

[0037] "Example 2"

[0038] Step 1: Before the user enters a new comment, the interactive tool analyzes past comments and written materials.

[0039] Step 2: Check whether the interactive tool is consistent with the content.

[0040] Step 3: If any conflicts are detected, the interactive tool will display a warning message and suggest ways to fix the conflicts.

[0041] Example 1

[0042] Next, a description will be given of Example 1 of the embodiment 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".

[0043] Conventional chat tools have the problem that users may post content that contradicts past statements or documents created, resulting in inconsistent information. This problem can cause confusion in project progress and communication. Therefore, a system that supports users to avoid contradicting past statements or documents created is required.

[0044] 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.

[0045] In this invention, the server includes means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the content of a new comment when the user inputs the comment, means for comparing the analyzed content of the comment with the past comments and created materials, and means for displaying a warning message to the user when a contradiction is detected and suggesting that the contradictory content be corrected. This makes it possible to maintain consistency in the content of user comments and prevent confusion in the progress of projects and communication.

[0046] "User" refers to an individual or organization that uses chat tools to make statements or create materials.

[0047] "Past comments" refers to records of comments a user previously made on the chat tool.

[0048] "Created Materials" refers to documents and data created by Users on chat tools or related platforms.

[0049] "Input" refers to the data or information a system receives for processing.

[0050] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0051] "Means" refers to a method or apparatus for achieving a specific function or purpose.

[0052] "Chat Tool" refers to software or platform that enables users to exchange messages in real time.

[0053] A "plugin" refers to an extension or module that is added to a chat tool.

[0054] An "interactive tool" refers to software or a platform that allows users and systems to communicate with each other.

[0055] "Managed online" means that data or information is stored in a manner that makes it accessible via the internet.

[0056] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[0057] "Analysis" refers to the process by which a system understands and extracts meaning from input data or information.

[0058] "Comparison" refers to the process in which the system compares new statements with past statements and created materials to detect differences.

[0059] "Warning message" refers to a message that the system displays to alert the user.

[0060] "Correction" refers to the act of a user changing statements or materials to resolve a contradiction.

[0061] This invention is a system that automatically checks for inconsistencies when a user makes a comment on a chat tool by comparing the comment with past comments and created documents. This system is implemented as a plug-in for the chat tool.

[0062] Generating a Program

[0063] The server generates a program implemented as a plug-in for the chat tool. This program has a function of automatically checking for inconsistencies by comparing the content of a user's comment with past comments and created documents when the user comments on the chat tool.

[0064] Program Processing

[0065] When a user types a new comment into the chat tool, the device sends the comment to the plugin, which then uses natural language processing (NLP) technology to analyze the comment. This analysis can be done using a generative AI model, such as OpenAI's (registered trademark) GPT-3 (registered trademark).

[0066] The analyzed content of the statement is compared with past statements and documents stored on the server. This comparison process uses MySQL (registered trademark) or PostgreSQL as a database management system (DBMS). The server compares the new statement with past statements and documents to check for any inconsistencies.

[0067] If a conflict is detected, the server sends that information back to the device, which then displays a warning message to the user with suggestions for correcting the conflict. The warning message includes the specific conflict and the suggested correction.

[0068] Examples

[0069] For example, if a user says, "The project is due next Friday," the plugin checks past statements and documentation. If a user has previously said, "The project is due this Friday," a contradiction is detected.

[0070] In this case, the user will see a warning message like this:

[0071] "There is a contradiction with your previous statement. Previously you said, 'The project deadline is this Friday.' Would you like to correct that?"

[0072] Examples of prompt statements

[0073] Examples of prompts to input to a generative AI model include:

[0074] "A user has recently stated, 'The project deadline is next Friday.' Please check for inconsistencies against previous statements and documentation."

[0075] In this way, the system helps users maintain consistency in what they say.

[0076] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0077] Step 1:

[0078] A user types a comment into a chat tool.

[0079] The user types a new statement into the chat tool's input field, for example, "The project is due next Friday," and the statement is sent to the next processing step.

[0080] Step 2:

[0081] The device sends the speech content to the plugin.

[0082] The terminal sends the user's input to the plugin. This transmission is done through the chat tool's internal API. The input is the user's input, and the output is the data sent to the plugin.

[0083] Step 3:

[0084] The plugin analyzes the content of the comment.

[0085] The plugin analyzes the received utterances using natural language processing (NLP) techniques. Specifically, it uses a generative AI model (e.g., OpenAI's GPT-3) to understand the meaning of the utterances. The input is the user's utterance, and the output is the analyzed utterance data.

[0086] Step 4:

[0087] The server compares the content of the statement with past statements and created materials.

[0088] The server compares the analyzed comment content with past comments and created documents stored in a database. MySQL or PostgreSQL is used as a database management system (DBMS) for this comparison process. The input is the analyzed comment content and past data, and the output is the comparison result.

[0089] Step 5:

[0090] The server detects the discrepancy.

[0091] The server compares new statements with past statements and documents to check for inconsistencies. For example, if someone previously said, "The project deadline is this Friday," a contradiction is detected. The input is the comparison result, and the output is information about whether there is a contradiction.

[0092] Step 6:

[0093] The server sends the conflicting information back to the terminal.

[0094] The server returns the detected conflict information to the terminal, including the specific conflicts and suggested corrections. The input is the conflict information, and the output is the data transmission to the terminal.

[0095] Step 7:

[0096] The device will display a warning message to the user.

[0097] The terminal displays a warning message to the user, for example, "There is a contradiction with your previous statement. You previously said, 'The project deadline is this Friday.' Would you like to correct it?" The input is the contradiction information from the server, and the output is the warning message to the user.

[0098] Step 8:

[0099] The user edits what was said.

[0100] The user checks the warning message and modifies the comment if necessary. The modified comment is sent to the plugin again, where the same check is performed. The input is the warning message, and the output is the modified comment.

[0101] (Application example 1)

[0102] Next, a description will be given of Application Example 1 of Form 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."

[0103] Work instructions and reports within the factory may contain information that contradicts previous statements or documents, which can lead to reduced work efficiency and safety. In addition, there is no way to immediately respond to contradictions, which can lead to problems becoming more serious.

[0104] 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. In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing work instructions and report contents in the factory in real time and checking for inconsistencies by comparing them with past instructions and report contents, and a means for generating a warning message and presenting it to the user when an inconsistency is detected. This maintains the consistency of work instructions and report contents in the factory, enabling efficient and safe work.

[0105] "User" means an individual or organization that uses the System.

[0106] "Past comments" are records of comments a user has made in the past.

[0107] "Created Materials" means documents and data records created by a User.

[0108] "Input" refers to data or information that is entered into a system.

[0109] "Content to be posted" refers to the new information or instructions that the user wishes to state.

[0110] "Contradiction" refers to a situation in which new statements are inconsistent with or in conflict with previous statements or documents.

[0111] "Suggest" refers to the system making suggestions or advice to the user.

[0112] "Present" refers to the system displaying information to the user.

[0113] "Factory" refers to the facility or location where manufacturing operations take place.

[0114] "Work instructions" are information that instructs the specific work content and procedures to be performed in a factory.

[0115] "Report content" refers to information about the progress and results of work.

[0116] "Real-time" refers to processing and display occurring immediately, without delay.

[0117] "Analysis" refers to the detailed examination of data or information to find meaning and patterns.

[0118] A "warning message" is a notification or alert intended to alert the user.

[0119] To implement this invention, a specific program must be installed on the robots and terminals used in the factory. This program has the function of receiving the user's past comments and documents as input, and suggesting whether there are any inconsistencies in the content to be posted in the future. It also includes a function of analyzing the work instructions and report contents in the factory in real time, and checking for inconsistencies by comparing them with past instructions and reports.

[0120] The server uses the following hardware and software:

[0121] Hardware: Factory robots

[0122] Software: Python (registered trademark), JSON files, natural language processing technology

[0123] The server first reads past comments and created documents from a JSON file. Then, when a new comment is entered, it checks for inconsistencies by comparing it with past comments and created documents. If a contradiction is detected, a warning message is generated and displayed to the user. If there is no contradiction, the new comment is added to the list of past comments and saved in a JSON file.

[0124] As a concrete example, consider instructions and reports for machine maintenance in a factory. If a past statement such as "Maintenance of machine A has started" is recorded, and a new statement such as "Maintenance of machine A has been completed" is entered, the system will compare these statements to check for inconsistencies.

[0125] Here are some example prompts using a generative AI model:

[0126] Previously stated: "Maintenance has begun on Machine A."

[0127] New statement: "Maintenance on Machine A has been completed."

[0128] Check for inconsistencies.

[0129] In this way, consistency in work instructions and reports within the factory can be maintained, enabling efficient and safe work.

[0130] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0131] Step 1:

[0132] The server reads past comments and created documents from a JSON file. Specifically, it opens a JSON file at a specified path from the file system and converts its contents into a data structure (list or dictionary). The input is the path to the JSON file, and the output is the data structure of past comments and created documents.

[0133] Step 2:

[0134] The user inputs a new utterance. The terminal receives the input from the user and sends the contents to the server. The input is the user's new utterance, and the output is the utterance data to be sent to the server.

[0135] Step 3:

[0136] The server compares the new statement with past statements and documents to check for inconsistencies. Specifically, it loops through the data structures of past statements and documents and compares them with the new statement. The input is the new statement and the data structures of past statements and documents, and the output is the presence or absence of inconsistencies and a list of contradictory statements.

[0137] Step 4:

[0138] The server generates a warning message when a contradiction is detected. Specifically, the server creates a warning message based on the list of contradictory statements and presents it to the user. The input is the list of contradictory statements, and the output is the warning message.

[0139] Step 5:

[0140] If no contradictions are detected, the server adds the new comment to the list of past comments and saves it in a JSON file. Specifically, the server adds the new comment to a data structure, converts the data structure to JSON format, and writes it to a file. The input is the new comment and the data structure of past comments and created materials, and the output is the updated JSON file.

[0141] Step 6:

[0142] The server notifies the user that the new comment has been added successfully. Specifically, it sends a message to the user's terminal to inform them that the comment has been added. The input is the result of adding the new comment, and the output is a notification message to the user.

[0143] Example 2

[0144] Next, a second embodiment of the second embodiment will be described. 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."

[0145] Conventional chat tools and dialogue tools lack the functionality to automatically check for inconsistencies with past comments or documents when a user makes a new comment. This means that users must manually check past comments and documents, which is inefficient and increases the risk of sending incorrect information. The present invention aims to solve these problems and enable users to send information efficiently and accurately.

[0146] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0147] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the past comments and created materials using natural language processing technology, a means for checking whether there are any inconsistencies between a new comment and the past comments and created materials using a generative AI model, and a means for displaying a warning message to the user and suggesting corrections when a inconsistency is detected. This enables the user to automatically check for inconsistencies with past comments and created materials when making a new comment, and to transmit information efficiently and accurately.

[0148] "User" means an individual or organization that uses the system to make statements or create materials.

[0149] "Past comments" are records of comments or statements a user has previously made through the system.

[0150] "Created Materials" means documents and data records created by Users through the System.

[0151] "Input" refers to data or information entered into a system.

[0152] "Suggestion" means that the system makes suggestions or advice to the user.

[0153] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[0154] A "generative AI model" is a model that uses artificial intelligence to generate and analyze data.

[0155] A "contradiction" is a state in which there is a discrepancy or inconsistency between a new statement and previous statements or documents.

[0156] A "warning message" is a message that the system displays to alert the user.

[0157] "Suggesting corrections" means that the system suggests to the user to correct the content of their statements or materials.

[0158] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0159] The present invention is a system that automatically checks for inconsistencies with past statements and created materials when a user makes a new statement, enabling the user to disseminate information efficiently and accurately. The system includes a server, a user terminal, and a generative AI model.

[0160] Hardware and Software Used

[0161] Hardware: Servers, user devices (PCs, smartphones, etc.)

[0162] Software: Generative AI models (e.g., GPT-4®), natural language processing libraries (e.g., spaCy, NLTK), database management systems (e.g., MySQL)

[0163] System Operation

[0164] 1. The user enters a new comment

[0165] The user uses the terminal to enter a new utterance into an input field of the interactive tool. For example, the user enters, "The start date for the new project is next Monday."

[0166] 2. The server retrieves past comments and documents

[0167] The server connects to the database management system and queries the user's past comments and documents. For example, a past comment such as "The project start date is the Monday after next" is retrieved from the database.

[0168] 3. The server performs natural language processing

[0169] The server uses natural language processing libraries to analyze the retrieved past utterances and written materials, specifically tokenizing sentences, tagging parts of speech, and analyzing dependencies.

[0170] 4. The server performs a conflict check

[0171] The server uses the generative AI model to check whether the new statement is consistent with previous statements or documents. For example, the generative AI model is given the following prompt: "The user has entered a new statement. Please check whether there are any inconsistencies with previous statements or documents. New statement: 'The start date of the new project is next Monday.' Previous statement: 'The start date of the project is the Monday after next.'"

[0172] 5. The server notifies the user of the result

[0173] If a conflict is detected, the server generates a warning message for the user and displays it on the terminal, for example "Isn't the start date of your new project on Monday after next?", and suggests ways to correct the conflict.

[0174] Examples

[0175] User input: "The start date for my new project is next Monday."

[0176] Previously stated: "The project start date is the Monday after next."

[0177] In this case, the system proceeds as follows.

[0178] 1. A user uses a terminal and types, "The start date for my new project is next Monday."

[0179] 2. The server retrieves the past statement "The project start date is the Monday after next" from the database.

[0180] 3. The server uses a natural language processing library to analyze past utterances.

[0181] 4. The server uses the generative AI model to check for inconsistencies between new and previous utterances.

[0182] 5. Because an inconsistency is detected, the server displays a warning message to the user asking "Isn't the start date of your new project on Monday after next?" and suggests a correction.

[0183] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[0184] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0185] Program processing flow

[0186] Step 1: User enters new utterance

[0187] What happens: The user uses the terminal to enter a new utterance into the input field of the interactive tool.

[0188] Input: The new utterance typed by the user (e.g., "The start date for the new project is next Monday.")

[0189] Output: The user's new utterance is sent to the system.

[0190] Step 2: The server retrieves past statements and documents

[0191] Specific operation: The server connects to a database management system and queries and retrieves the user's past comments and created materials.

[0192] Input: User ID or user identification

[0193] Data processing: Search the database for past statements and created materials and extract relevant data.

[0194] Output: Data from past statements and documents (e.g. "The project start date is the Monday after next week.")

[0195] Step 3: The server performs natural language processing

[0196] Specific operation: The server uses a natural language processing library to analyze the retrieved past statements and created materials.

[0197] Input: Data of past statements and created documents

[0198] Data processing: Sentence tokenization, part-of-speech tagging, dependency analysis, etc.

[0199] Output: Structured data of analyzed past statements and documents

[0200] Step 4: Server performs inconsistency check

[0201] How it works: The server uses a generative AI model to check for inconsistencies between new statements and previous statements or created materials.

[0202] Input: New statements, parsed past statements, and structured data of created documents

[0203] Data calculation: Input a prompt sentence into the generative AI model and perform inconsistency checks.

[0204] Example prompt: "The user has entered a new statement. Please check for inconsistencies with previous statements and documentation. New statement: 'The new project start date is next Monday.' Previous statement: 'The project start date is the Monday after next.'"

[0205] Output: Result on whether there is a conflict (e.g., conflict detected)

[0206] Step 5: The server notifies the user of the result

[0207] Specific Behavior: If a conflict is detected, the server generates a warning message for the user and displays it on the terminal.

[0208] Input: Result regarding whether or not there is a conflict

[0209] Data processing: Generate warning messages (e.g. "Isn't the start date of your new project on Monday after next?")

[0210] Output: A warning message that is displayed on the user's terminal.

[0211] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[0212] (Application example 2)

[0213] Next, a description will be given of Application Example 2 of Form 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."

[0214] Conventional dialogue tools and chat tools lack the functionality to automatically check for inconsistencies with past statements and prepared documents when a user makes a new statement. This makes it difficult to maintain consistency of important information, particularly in security services, and there is a risk of incorrect information being transmitted. The present invention aims to solve these problems and provide a system that automatically checks for inconsistencies with past statements and prepared documents when a user makes a new statement, and suggests corrections as necessary.

[0215] 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.

[0216] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for using a generative AI model to suggest corrections when a contradiction is detected. This makes it possible to automatically check for inconsistencies with past comments and created materials when the user makes a new comment, and to suggest corrections as necessary.

[0217] "User" means an individual or entity that uses the System to input statements or create materials.

[0218] "Past comments" are records of comments or statements a user has previously made within the system.

[0219] "Created Materials" refers to documents and data created or uploaded by a User within the System.

[0220] "Input" refers to the data or information that a system receives for processing.

[0221] "Content to be posted" refers to the new information or comments that a user intends to make within the system.

[0222] "Inconsistency" refers to a situation in which new statements or posts are inconsistent with or in conflict with previous statements or materials.

[0223] "Suggestions" refers to suggestions or advice that the system provides to the user.

[0224] A "generative AI model" refers to an algorithm or program that uses artificial intelligence techniques to analyze data and generate appropriate suggestions or revisions.

[0225] "Present" refers to the act of the system displaying information or suggestions to the user.

[0226] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for using a generative AI model to suggest corrections when an inconsistency is detected.

[0227] When a user inputs a new comment, the server retrieves past comments and written materials from the database and analyzes them. Natural language processing technology is used for the analysis to check for inconsistencies between the past comments and written materials and the new comment. If a contradiction is detected, a generative AI model is used to generate a correction proposal and present it to the user.

[0228] Specifically, the server performs the following process. First, when a user inputs a new statement, the statement is sent to the server. The server retrieves past statements and created documents from a database and analyzes this data using natural language processing technology. For example, OpenAI's API can be used for the analysis. If the analysis detects a contradiction between the new statement and past statements or created documents, the server uses a generative AI model to generate a correction proposal. This correction proposal is presented to the user as a warning message.

[0229] The hardware used includes a server and a user's terminal. The server is a computer equipped with a high-performance processor and a large amount of memory, and the user's terminal is a general device such as a smartphone or a PC. The software used includes libraries for implementing natural language processing technology (e.g., Python's difflib library) and APIs for executing generative AI models (e.g., OpenAI's API).

[0230] As a concrete example, consider the case of checking statements regarding security policies. Assume that a past statement recorded is "All passwords must be at least eight characters long." If the user attempts to newly state "All passwords must be at least six characters long," the server analyzes this new statement and detects any inconsistencies with the past statement. If a contradiction is detected, the generative AI model is used to generate a prompt sentence like the one below, and a correction suggestion is presented.

[0231] Example prompt:

[0232] New statement: All passwords must be at least six characters long.

[0233] Contradictory previous statement: All passwords must be at least eight characters long.

[0234] Generate suggestions for fixing the inconsistencies.

[0235] In this way, when a user makes a new statement, the system automatically checks for inconsistencies with previous statements and documents, and suggests corrections if necessary.

[0236] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0237] Step 1:

[0238] The user inputs a new comment. The input comment is sent from the user's terminal to the server. The input data is the text of the new comment.

[0239] Step 2:

[0240] The server retrieves the user's past comments and created documents from the database. The retrieved data is the text of the past comments and created documents.

[0241] Step 3:

[0242] The server uses natural language processing technology to analyze new comments and past comments and created documents. Specifically, it uses Python's difflib library to calculate the similarity between comments and documents. The input data is the text of the new comment and past comments and created documents, and the output data is the similarity score.

[0243] Step 4:

[0244] The server checks whether there are any inconsistencies between the new comment and past comments or documents based on the analysis results. If a inconsistency is detected, the information is recorded. The input data is the similarity score, and the output data is a flag indicating whether or not there is a inconsistency.

[0245] Step 5:

[0246] If a contradiction is detected, the server uses a generative AI model to generate a correction proposal. Specifically, it uses OpenAI's API to generate a prompt sentence and obtain a correction proposal. The input data is the text of past statements or created documents that contradict the new statement, and the output data is the text of the correction proposal.

[0247] Step 6:

[0248] The server presents the generated revision proposal to the user, which is displayed as a warning message on the user's terminal. The input data is the text of the revision proposal, and the output data is the warning message presented to the user.

[0249] Step 7:

[0250] The user checks the proposed revisions and revises the new utterance if necessary. The revised utterance is sent to the server again, and the same process is repeated. The input data is the text of the revised utterance, and the output data is the text of the final utterance.

[0251] Furthermore, an emotion engine that estimates the emotion of the user may be combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[0252] "Example 1"

[0253] As one embodiment of the present invention, a chat tool plug-in or interactive tool combined with an emotion engine is provided. The emotion engine analyzes the user's emotional state from the user's past comments and documents created. Specifically, it analyzes the text of messages posted by the user in the past and documents created by the user, and extracts emotional expressions and keywords suggesting emotions contained therein. Then, the user's emotional state is estimated based on this information. For example, if the user has used many positive expressions such as "fun" and "happy" in the past, the user's emotional state is estimated to be "joy" or "happiness." The emotional state estimated in this way is used as reference information when suggesting whether there are any inconsistencies in the content the user will post in the future.

[0254] "Example 2"

[0255] In another embodiment of the present invention, a system is provided in which an emotion engine analyzes a user's emotional state using natural language processing technology and emotion recognition technology. Specifically, natural language processing technology is used to analyze text, and expressions and keywords indicating emotions are extracted from the text. Next, emotion recognition technology is used to analyze the extracted expressions and keywords to estimate the user's emotional state. For example, if a user uses a lot of negative expressions such as "sad" and "disappointed," the user's emotional state is estimated to be "sadness" or "disappointment." The emotional state estimated in this way is used as reference information when suggesting whether the content the user is about to post is consistent with the user's emotional state.

[0256] The process flow of each embodiment will be described below.

[0257] "Example 1"

[0258] Step 1: A user types a message into the chat tool.

[0259] Step 2: The emotion engine analyzes the user's emotional state from their past statements and documents. Specifically, it analyzes the text of past messages and documents and extracts emotional expressions and keywords that suggest emotions.

[0260] Step 3: The emotion engine estimates the user's emotional state based on the extracted information.

[0261] Step 4: Based on the estimated emotional state, we suggest inconsistencies in the content the user will post next.

[0262] "Example 2"

[0263] Step 1: A user types a message into the chat tool.

[0264] Step 2: The emotion engine uses natural language processing techniques to analyze the text and extract expressions and keywords that indicate emotions.

[0265] Step 3: The emotion engine uses emotion recognition technology to analyze the extracted expressions and keywords and infer the user's emotional state.

[0266] Step 4: Based on the estimated emotional state, we suggest whether the content of the user's future post is consistent with that emotional state.

[0267] Example 1

[0268] Next, a description will be given of Example 1 of the embodiment 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".

[0269] Conventional chat tools and conversational tools lack the ability to automatically detect and warn users when their statements contradict previous statements or written materials. In addition, suggestions do not take into account the user's emotional state, making it difficult to provide appropriate advice that is in line with the user's intentions. This can lead to a lack of consistency in communication, leading to misunderstandings and confusion.

[0270] 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.

[0271] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there is any inconsistency in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the user's comments using natural language processing technology, a means for comparing the analyzed comments with the past comments and created materials stored in the database, a means for displaying a warning message to the user when a contradiction is detected, a means for analyzing the user's emotional state from the user's past comments and created materials, and a means for making suggestions based on the analyzed emotional state. This makes it possible to automatically check whether the user's comments are inconsistent with the past comments and created materials, and to provide appropriate warnings and suggestions. In addition, by making suggestions taking the user's emotional state into consideration, more consistent communication can be achieved.

[0272] "User" refers to an individual or organization that makes statements or creates materials using chat tools or interactive tools.

[0273] A "utterance" is a text message entered by a user through a chat or interactive tool.

[0274] "Created Materials" refers to documents and data files previously created by the User, including reports, memos, presentation materials, etc.

[0275] "Input" refers to the data or information a system receives for analysis or comparison.

[0276] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0277] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language, and includes tokenization, grammatical analysis, and semantic analysis.

[0278] A "database" refers to a collection of information that stores past statements and created materials and can be used for searching and comparison.

[0279] "Contradiction" refers to an inconsistency that occurs when a user's new statements do not match previous statements or materials they have created.

[0280] "Warning message" refers to a notification sent by the system to the user to inform them of a discrepancy and encourage them to correct it.

[0281] "Emotional state" refers to a user's psychological state inferred from their past statements and written materials, and may include joy, sadness, anger, etc.

[0282] "Analysis" refers to the process by which the system processes the data it receives and understands its meaning and structure.

[0283] "Comparison" refers to the process of comparing the analyzed statements with past statements and created materials to check for matches and inconsistencies.

[0284] The present invention is a system that automatically checks whether there are any inconsistencies with past statements or documents created when a user uses a chat tool or an interactive tool, and provides appropriate suggestions. A specific embodiment of this system will be described below.

[0285] System Configuration

[0286] This system is primarily composed of three elements: a server, a terminal, and a user. The server is responsible for analyzing what users say and comparing it with past comments and created materials. Terminals are devices that users use to operate chat tools and interactive tools, such as PCs and smartphones. Users are individuals or groups who use chat tools and interactive tools to make comments and create materials.

[0287] Hardware and Software Used

[0288] The server is a computer system equipped with a high-performance processor and a large amount of memory. The database uses a relational database management system (RDBMS) or a NoSQL database to store past statements and created materials. Specifically, MySQL is one such example.

[0289] Natural language processing technology is used for the analysis. Specifically, OpenAI's GPT-3 and Google's BERT are used as generative AI models. These models are used to tokenize the user's speech and analyze its grammar and meaning.

[0290] Data processing and calculation

[0291] When a user enters a new comment, the server receives the comment and analyzes it using natural language processing technology. The analyzed comment is compared with past comments and created materials stored in the database. This comparison is performed using ElasticSearch (registered trademark), a full-text search engine. Based on the comparison results, it is determined whether there are any inconsistencies, and if any are detected, a warning message is displayed to the user.

[0292] In addition, an emotion engine is used to analyze the user's emotional state from past comments and written materials. This analysis is carried out using tools such as IBM Watson's Natural Language Understanding. The results of the analysis are used to estimate the user's emotional state and are used as reference information when suggesting whether there are any inconsistencies in the content of future posts.

[0293] Examples

[0294] For example, consider a case where a user says in a chat tool, "A new project has started." If this statement contradicts a previous statement, "The project has not started yet," the server will warn, "This statement contradicts your previous statement. Do you want to correct it?"

[0295] Example prompt:

[0296] The user says "A new project has started", which contradicts their previous statement "The project hasn't started yet". Show a warning message and suggest a fix.

[0297] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[0298] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0299] Step 1:

[0300] The user types a new comment into the chat tool.

[0301] Input: User says "I've started a new project."

[0302] Output: The content of the comment is sent to the chat tool.

[0303] Specific behavior: The user types text into a chat window and clicks the send button.

[0304] Step 2:

[0305] The server receives the user's comments and analyzes them using natural language processing technology.

[0306] Input: User says "I've started a new project."

[0307] Output: Tokenization and grammar analysis of the parsed utterance.

[0308] How it works: The server uses a generative AI model (e.g., GPT-3) to tokenize the utterance and perform grammar analysis.

[0309] Step 3:

[0310] The server compares the analyzed content of the comments with past comments and created materials stored in a database.

[0311] Input: Analyzed statements and past statements and written materials in the database.

[0312] Output: Comparison result (conflicts or not).

[0313] Specific behavior: The server uses a full-text search engine (e.g., Elasticsearch) to search for matches to the past utterance "The project hasn't started yet."

[0314] Step 4:

[0315] The server determines whether there is a conflict based on the comparison.

[0316] Input: The comparison result.

[0317] Output: The result of determining whether there is a contradiction.

[0318] Specific behavior: The server analyzes the comparison result and determines that "A new project has started" and "The project has not started yet" are inconsistent.

[0319] Step 5:

[0320] The server will display a warning message to the user if a discrepancy is detected.

[0321] Input: The result of the determination as to whether there is a contradiction or not.

[0322] Output: A warning message.

[0323] Specific behavior: The server displays a warning message to the user saying, "This statement contradicts a previous statement. Would you like to correct it?"

[0324] Step 6:

[0325] The server uses an emotion engine to analyze the user's emotional state from their past statements and written materials.

[0326] Input: Past statements and created materials.

[0327] Output: The parsed emotional state.

[0328] Specific operation: The server uses a sentiment analysis tool (e.g., IBM Watson) to extract positive sentiment from past statements such as "It was fun" and "I'm happy."

[0329] Step 7:

[0330] The server makes suggestions based on the analyzed emotional state.

[0331] Input: Parsed emotional state and current speech.

[0332] Output: Suggestions that take emotional state into account.

[0333] Specific behavior: The server suggests to the user, "Even though you've contradicted your previous statements, would you like to emphasize that the project is going well?"

[0334] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[0335] (Application example 1)

[0336] Next, a description will be given of Application Example 1 of Form 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."

[0337] In modern companies, improving the efficiency of internal communication and preventing information leaks are important issues. In particular, when communicating using chat tools, it is necessary to avoid inconsistencies with past statements and documents, to grasp the user's emotional state and detect abnormal behavior early, and to prevent the leaking of confidential information. However, it is difficult to realize these functions in a single system, and currently it is necessary to use multiple tools at the same time. This increases the burden on users and makes efficient operation difficult.

[0338] 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.

[0339] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing the user's emotional state and displaying a warning when an abnormal emotion is detected, and a means for detecting comments containing confidential information and displaying a warning. This makes it possible to avoid inconsistencies in internal communication, early detection of abnormal emotional states, and prevention of confidential information leaks.

[0340] "User" refers to any individual or entity that uses the System.

[0341] "Past comments" refers to the content of chats or messages a user has previously made.

[0342] "Created materials" refers to text data such as documents and reports created by users.

[0343] "Input" refers to data or information entered into a system.

[0344] "Contradiction" refers to a discrepancy or inconsistency that arises between past statements or prepared materials and new statements.

[0345] "Suggestions" refers to suggestions or advice that the system presents to the user.

[0346] "Emotional state" refers to a user's psychological state or mood.

[0347] "Analysis" refers to the process of analyzing data or information to extract meaning and patterns.

[0348] "Abnormal emotions" refers to negative emotions or abnormal psychological states that deviate from normal emotional states.

[0349] "Warning" refers to a message or notification that alerts the user.

[0350] "Confidential information" refers to important information that could cause harm if leaked to the outside.

[0351] "Discovery" refers to the process of finding specific conditions or patterns.

[0352] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the user's emotional state and displaying a warning if abnormal emotions are detected, and means for detecting comments that include confidential information and displaying a warning.

[0353] System Programming

[0354] This system is realized using the following hardware and software.

[0355] Hardware

[0356] Server: Manages the database and performs natural language processing and sentiment analysis.

[0357] User device: A smartphone or computer with a chat tool installed.

[0358] software

[0359] Natural Language Processing libraries: TfidfVectorizer, cosine_similarity (using Python's scikit-learn library)

[0360] Sentiment analysis library: TextBlob (Python library)

[0361] Database management system: A database for storing and managing past statements and created materials (e.g., MySQL, PostgreSQL)

[0362] Processing flow

[0363] 1. Receiving input:

[0364] The server retrieves the user's past comments and created materials from a database and receives new comments as input.

[0365] 2. Suggested contradictions:

[0366] The server uses TfidfVectorizer to vectorize new and past comments, and calculates cosine_similarity. If the similarity is low, it is determined that there is a contradiction and presents suggestions to the user.

[0367] 3. Emotional state analysis:

[0368] The server uses TextBlob to calculate the sentiment score of new utterances, and if the sentiment score is abnormal, it displays a warning to the user.

[0369] 4. Detecting Sensitive Information:

[0370] The server uses regular expressions to check whether new posts contain sensitive information, and if so, displays a warning to the user.

[0371] Examples

[0372] Suggested contradictions:

[0373] Previously stated: "Project X is progressing well."

[0374] New statement: "Progress on Project X is slow."

[0375] Result: The inconsistency is detected and the user is prompted with the suggestion, "This statement contradicts something you said previously. Would you like to fix it?"

[0376] Emotional state analysis:

[0377] Say: "I really hate this project."

[0378] Result: Negative sentiment is detected and the user is alerted with "Negative sentiment detected. Is something wrong?"

[0379] Sensitive information detection:

[0380] Say: "This is classified information."

[0381] Result: Because it contains sensitive information, the user will see a warning saying "This post contains sensitive information. Do you want to send this post?"

[0382] Examples of prompt statements

[0383] "Please create a program that checks for inconsistencies in what the user says against past statements and documents, and displays a warning if any inconsistencies are detected. Please also analyze the user's emotional state, and display a warning if abnormal emotions are detected. Furthermore, please add a function to display a warning if confidential information is included."

[0384] In this way, it is possible to avoid inconsistencies in internal communication, detect abnormal emotional states early, and prevent the leakage of confidential information.

[0385] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0386] Step 1:

[0387] The server retrieves the user's past comments and created materials from a database. This provides the past comments and created materials to the server as input. Specifically, a database query is executed to extract past comments and created materials related to the user ID.

[0388] Step 2:

[0389] The server receives the new comment entered by the user. This provides the new comment to the server as input. Specifically, the server receives an API request from the chat tool and obtains the new comment content.

[0390] Step 3:

[0391] The server vectorizes the new and past comments using TfidfVectorizer, which converts the text data into numerical vectors. Specifically, the new and past comments are input into TfidfVectorizer as a single corpus, and a TF-IDF vector is generated for each piece of text.

[0392] Step 4:

[0393] The server calculates cosine_similarity to evaluate the similarity between the new and past comments. This determines whether there is a contradiction. Specifically, it calculates the cosine similarity between the vector of the new comment and the vector of the past comment, and determines that there is a contradiction if the similarity is low.

[0394] Step 5:

[0395] If a contradiction is detected, the server presents a suggestion to the user, giving the user an opportunity to correct the contradiction. Specifically, the server displays a message through the chat tool interface saying, "This is inconsistent with your previous statement. Do you want to correct it?"

[0396] Step 6:

[0397] The server uses TextBlob to calculate the sentiment score of a new comment, which analyzes the sentiment state of the comment. Specifically, a new comment is input to TextBlob and a positive / negative sentiment score is obtained.

[0398] Step 7:

[0399] If the emotion score is abnormal, the server displays a warning to the user. This allows the user to become aware of the abnormal emotional state. Specifically, if the emotion score falls below a certain threshold, the server displays a message saying, "Negative emotion has been detected. Is there a problem?"

[0400] Step 8:

[0401] The server uses regular expressions to check whether new comments contain confidential information. This is how confidential information is detected. Specifically, the server searches new comments for keywords that indicate confidential information (e.g., "confidential" or "secret"), and if any are found, it determines that the new comment contains confidential information.

[0402] Step 9:

[0403] If confidential information is included, the server displays a warning to the user. This allows the user to prevent confidential information from being leaked. Specifically, the server displays a message saying, "Confidential information is included. Do you want to send this message?"

[0404] Example 2

[0405] Next, a second embodiment of the second embodiment will be described. 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."

[0406] In conventional dialogue and chat tools, there is a risk that users may post content that contradicts their past statements or documents they have created. In addition, since suggestions are made without taking into account the user's emotional state, advice that does not match the user's intentions or emotions may be provided. For this reason, there is a demand for a system that provides appropriate suggestions that take into account the user's past statements and emotional state.

[0407] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0408] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether the content to be posted from now on is consistent with inconsistencies, a means for presenting the suggestions to the user, a means for analyzing the user's comments and estimating the user's emotional state, and a means for suggesting whether the content to be posted from now on is consistent with the emotional state based on the estimated emotional state. This makes it possible to make appropriate suggestions taking into account the user's past comments and emotional state.

[0409] "User" refers to an individual or organization that uses the System to create and post comments and materials.

[0410] "Comments" refers to text data that users input and post through the system.

[0411] "Created Materials" refers to documents and data that a User has previously created and stored through the System.

[0412] "Input" refers to the data or information a system receives for processing.

[0413] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0414] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0415] "Emotional state" refers to the psychological state or emotion inferred from a user's statements.

[0416] "Contradiction" refers to a situation where a user's new statements are inconsistent with or contradict previous statements or materials created.

[0417] "Managed online" means that data and information are stored and managed in a manner that makes them accessible via the Internet.

[0418] "Analysis" refers to the process by which a system examines and evaluates data or information in detail.

[0419] "Inference" refers to the process by which a system predicts a state or outcome based on data and information.

[0420] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0421] The present invention is a system that receives a user's past comments and documents as input, and suggests whether the content to be posted in the future is consistent with the user's past comments and documents. Furthermore, the system has a function of analyzing the user's comments and estimating the user's emotional state, and then suggesting whether the content to be posted in the future is consistent with the user's emotional state based on the estimated emotional state.

[0422] Hardware and Software Used

[0423] The present system is implemented using the following hardware and software.

[0424] Server: Equipped with a database management system (DBMS), it stores and manages users' past comments and documents created by users. The server also provides the computing resources to execute natural language processing and emotion recognition technologies.

[0425] Terminal: The device (PC, smartphone, tablet, etc.) that a user uses to enter utterances.

[0426] Natural Language Processing libraries: Use Python's NLTK library and spaCy to tokenize and semantically analyze text.

[0427] Emotion Recognition Library: Analyzes and estimates the user’s emotional state using Python’s TextBlob and VADER Sentiment Analysis.

[0428] System Operation

[0429] 1. Check before user comments are entered

[0430] When a user enters a new comment, the server retrieves the user's past comments and created materials from the database.

[0431] The server uses natural language processing technology to analyze the content of past statements and documents, and check for any contradictions with new statements.

[0432] Example of how it works: When a user tries to type "The project is on track," the server retrieves the previous statement "The project is delayed" and detects the inconsistency.

[0433] 2. Detecting inconsistencies and displaying warning messages

[0434] If the server detects a discrepancy, it will display a warning message to the user.

[0435] Specific example of behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[0436] 3. Analyzing the user's emotional state

[0437] When a user enters a comment, the server uses natural language processing technology to analyze the text and extract expressions and keywords that indicate emotions.

[0438] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[0439] Example of specific operation: If a user enters "I'm sad that things haven't been going well at work recently," the server will extract the keyword "sad" and estimate the emotional state as "sad."

[0440] 4. Using Emotional States to Suggest

[0441] Based on the estimated emotional state, the system suggests whether the content the user will post next is consistent with that emotional state.

[0442] An example of how this works: When a user tries to type "This project is awesome," the server suggests "Do you really feel that way?"

[0443] Examples of prompt statements

[0444] Conflict check prompt

[0445] Compare the user's previous statements with their new statements and look for inconsistencies. Previous statement: "The project is delayed," New statement: "The project is on track."

[0446] Sentiment Analysis Prompt

[0447] Analyze user utterances and infer their emotional state. utterance: "I'm sad that things haven't been going well at work lately."

[0448] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0449] Step 1:

[0450] The user inputs a new comment.

[0451] Input: The user types text into an input field on the terminal.

[0452] What happens: The user types, "The project is going well."

[0453] Output: The entered text data is sent to the server.

[0454] Step 2:

[0455] The server retrieves the user's past comments and created materials from the database.

[0456] Input: A user's ID and a request for past comments.

[0457] What happens: The server runs a SQL query to retrieve the user's previous statement "The project is delayed."

[0458] Output: Data of past comments and created materials is returned to the server.

[0459] Step 3:

[0460] The server uses natural language processing technology to analyze the content of past statements and documents.

[0461] Input: Text data of past statements and created documents.

[0462] How it works: The server uses Python's NLTK library to tokenize past utterances and perform semantic analysis.

[0463] Output: Structured information of the parsed text data.

[0464] Step 4:

[0465] The server compares the new message with previous messages to check for inconsistencies.

[0466] Input: Text data of new utterances and structured information of previous utterances parsed.

[0467] What happens: The server compares the new statement "The project is on track" with the previous statement "The project is delayed" and detects the inconsistency.

[0468] Output: Information about whether there are any conflicts.

[0469] Step 5:

[0470] If the server detects a discrepancy, it displays a warning message to the user.

[0471] Input: Information about whether there is a conflict.

[0472] Specific behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[0473] Output: A warning message that is displayed on the user's terminal.

[0474] Step 6:

[0475] The server uses natural language processing technology to analyze the user's comments and extract expressions and keywords that indicate emotions.

[0476] Input: The text of the user's new utterance.

[0477] Specific operation: The server uses Python's TextBlob library to extract the keyword "sad" from the comment.

[0478] Output: Extracted sentiment expressions and keywords.

[0479] Step 7:

[0480] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[0481] Input: Extracted sentiment expressions and keywords.

[0482] What happens: The server uses VADER Sentiment Analysis to estimate the user's emotional state as "sad."

[0483] Output: The estimated emotional state.

[0484] Step 8:

[0485] Based on the estimated emotional state, the server suggests whether the content the user is about to post is consistent with that emotional state.

[0486] Input: Estimated emotional state and textual data of new utterances.

[0487] Specific behavior: The server responds to the user's new statement, "This project is awesome," by suggesting, "Do you really feel that way?"

[0488] Output: The suggestion message that is displayed on the user's device.

[0489] (Application example 2)

[0490] Next, a description will be given of Application Example 2 of Form 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."

[0491] Traditional chat tools and conversational tools lack the ability to automatically check whether the content a user posts is consistent with past statements or documents. They also lack the ability to analyze the user's emotional state and provide appropriate suggestions based on that emotion. This creates a risk that users will post inconsistent information or make posts that are not in line with their emotions.

[0492] 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.

[0493] In this invention, the server includes a means for receiving the user's past comments and documents as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state. This enables the user to transmit consistent information and post content appropriate to their emotions.

[0494] "User" means any individual or entity that uses the System to post statements or create materials.

[0495] "Past statements and created materials" refers to statements made by users previously on the system, as well as documents and data they created.

[0496] "Input" refers to the data or information a system receives for processing.

[0497] A "contradiction" is when new statements or materials are inconsistent with or in conflict with previous statements or materials.

[0498] A "suggestion" is a suggestion or advice that the system provides to the user.

[0499] "Presentation" refers to the act of the system displaying information to the user.

[0500] An "emotional state" is a state that indicates a user's current feelings or mood.

[0501] "Analysis" is the act of examining data or information in detail to understand its meaning and structure.

[0502] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[0503] "Managed online" means that data and information are stored in a manner that makes them accessible via the internet.

[0504] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state.

[0505] System Configuration

[0506] 1. Hardware:

[0507] Server: A server for storing and processing data.

[0508] Device: The smartphone or computer used by the user.

[0509] 2. Software:

[0510] Natural language processing library: Uses natural language processing technologies such as spaCy.

[0511] Sentiment Analysis Library: Uses sentiment analysis techniques such as TextBlob.

[0512] Program processing explanation

[0513] The server receives input from users' online records of past posts and written materials. It then uses natural language processing technology to compare new posts with past posts and written materials to check for inconsistencies. If a contradiction is detected, the server displays a warning message to the user and suggests correcting the inconsistent content.

[0514] In addition, the server uses emotion analysis technology to analyze the user's emotional state. Based on the analysis results, the server suggests posting content that is consistent with the user's emotional state. This enables users to post information that is consistent with their emotions and appropriate to their feelings.

[0515] Examples

[0516] For example, if a user has previously posted "Today was a very fun day," and then tries to post "Today is very sad," the server detects the inconsistency. In addition, because negative emotions are detected through emotion analysis, the server displays a warning message to the user and suggests that they review the content of their post.

[0517] Examples of prompt statements

[0518] Here is an example of a prompt to be input to the generative AI model:

[0519] A new post from a user reads: "I'm so sad today."

[0520] Previous post: "Today was so much fun."

[0521] Check whether this new post is inconsistent with previous posts and perform a sentiment analysis.

[0522] Using this prompt sentence, the generative AI model can make appropriate suggestions.

[0523] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0524] Step 1:

[0525] The server manages users' past comments and written materials online. These data are contents posted by users in the past and documents created by users, and are stored in a database. The input is the user's past comments and written materials, and the output is a list of these data.

[0526] Step 2:

[0527] The user inputs new post content into the terminal. The terminal sends this input to the server. The input is the new post content entered by the user, and the output is the new post content sent to the server.

[0528] Step 3:

[0529] The server uses natural language processing technology to analyze new posts and past comments and written materials. Specifically, it uses libraries such as spaCy to tokenize the text and perform semantic analysis. The input is the new post and past comments and written materials, and the output is the analysis results.

[0530] Step 4:

[0531] The server uses the analysis results to check whether there are any inconsistencies between the new post and past comments or documents. If a contradiction is detected, the server identifies the content and generates a warning message for the user. The input is the analysis results, and the output is the presence or absence of a contradiction and a warning message.

[0532] Step 5:

[0533] The server analyzes the emotional state of the new post using sentiment analysis techniques, specifically by calculating a sentiment score for the text using a library such as TextBlob. The input is the new post, and the output is the sentiment score.

[0534] Step 6:

[0535] The server estimates the user's emotional state based on the emotion score, and suggests posting content that is consistent with the emotional state. The input is the emotion score, and the output is the suggested content.

[0536] Step 7:

[0537] The server generates warning messages and suggestions and sends them to the user's terminal, which displays them to the user. The input is the warning messages and suggestions, and the output is the information displayed to the user.

[0538] Step 8:

[0539] The user checks the warning messages and suggestions from the server and modifies the new post content as necessary. The input is the information displayed to the user, and the output is the modified post content.

[0540] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the voice 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 voice data.

[0541] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search<URL: https: / / openai.com / blog / chatgpt> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to 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.

[0542] Another example of generative AI is Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[0543] In the above embodiment, an example was given in which the specific process was performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0544] [Second embodiment]

[0545] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0546] 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.

[0547] 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 wide area network (WAN) and / or a local area network (LAN).

[0548] 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.

[0549] 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 the voice according to instructions from the processor 46.

[0550] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a 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 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[0551] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[0552] Fig. 4 shows an example of 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.

[0553] 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.

[0554] 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.

[0555] 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.

[0556] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[0557] "Example 1"

[0558] As one embodiment of the present invention, the present invention is implemented as a plug-in for a chat tool. When a user makes a comment on the chat tool, the plug-in compares the comment content with past comments and created materials, and automatically checks for inconsistencies. Specifically, when a user inputs a new comment, the plug-in analyzes the comment content and compares it with past comments and created materials. If an inconsistency is detected, the plug-in displays a warning message to the user and suggests that the inconsistent content be corrected.

[0559] "Example 2"

[0560] Another embodiment of the present invention is implemented as an interactive tool that automatically checks for inconsistencies before a user inputs a new statement by comparing it with past statements and written materials. Specifically, before a user inputs a new statement, the tool analyzes past statements and written materials and checks for inconsistencies with the contents of the new statement. If an inconsistency is detected, the tool displays a warning message to the user and suggests that the user correct the inconsistent contents.

[0561] The process flow of each embodiment will be described below.

[0562] "Example 1"

[0563] Step 1: A user types a new message into the chat tool.

[0564] Step 2: The plugin analyzes the comment and compares it with previous comments and created materials.

[0565] Step 3: If any conflicts are detected, the plugin will display a warning message and suggest you fix the conflicts.

[0566] "Example 2"

[0567] Step 1: Before the user enters a new comment, the interactive tool analyzes past comments and written materials.

[0568] Step 2: Check whether the interactive tool is consistent with the content.

[0569] Step 3: If any conflicts are detected, the interactive tool will display a warning message and suggest ways to fix the conflicts.

[0570] Example 1

[0571] Next, a description will be given of Example 1 of Form 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".

[0572] Conventional chat tools have the problem that users may post content that contradicts past statements or documents created, resulting in inconsistent information. This problem can cause confusion in project progress and communication. Therefore, a system that supports users to avoid contradicting past statements or documents created is required.

[0573] 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.

[0574] In this invention, the server includes means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the content of a new comment when the user inputs the comment, means for comparing the analyzed content of the comment with the past comments and created materials, and means for displaying a warning message to the user when a contradiction is detected and suggesting that the contradictory content be corrected. This makes it possible to maintain consistency in the content of user comments and prevent confusion in the progress of projects and communication.

[0575] "User" refers to an individual or organization that uses chat tools to make statements or create materials.

[0576] "Past comments" refers to records of comments a user previously made on the chat tool.

[0577] "Created Materials" refers to documents and data created by Users on chat tools or related platforms.

[0578] "Input" refers to the data or information a system receives for processing.

[0579] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0580] "Means" refers to a method or apparatus for achieving a specific function or purpose.

[0581] "Chat Tool" refers to software or platform that enables users to exchange messages in real time.

[0582] A "plugin" refers to an extension or module that is added to a chat tool.

[0583] An "interactive tool" refers to software or a platform that allows users and systems to communicate with each other.

[0584] "Managed online" means that data or information is stored in a manner that makes it accessible via the internet.

[0585] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[0586] "Analysis" refers to the process by which a system understands and extracts meaning from input data or information.

[0587] "Comparison" refers to the process in which the system compares new statements with past statements and created materials to detect differences.

[0588] "Warning message" refers to a message that the system displays to alert the user.

[0589] "Correction" refers to the act of a user changing statements or materials to resolve a contradiction.

[0590] This invention is a system that automatically checks for inconsistencies when a user makes a comment on a chat tool by comparing the comment with past comments and created documents. This system is implemented as a plug-in for the chat tool.

[0591] Generating a Program

[0592] The server generates a program implemented as a plug-in for the chat tool. This program has a function of automatically checking for inconsistencies by comparing the content of a user's comment with past comments and created documents when the user comments on the chat tool.

[0593] Program Processing

[0594] When a user types a new comment into the chat tool, the device sends the comment to the plugin, which then uses natural language processing (NLP) techniques to analyze the comment. This analysis can be done using a generative AI model, such as OpenAI's GPT-3.

[0595] The analyzed content of the statement is compared with past statements and documents stored on the server. This comparison process uses MySQL or PostgreSQL as a database management system (DBMS). The server compares the new statement with past statements and documents to check for any inconsistencies.

[0596] If a conflict is detected, the server sends that information back to the device, which then displays a warning message to the user with suggestions for correcting the conflict. The warning message includes the specific conflict and the suggested correction.

[0597] Examples

[0598] For example, if a user says, "The project is due next Friday," the plugin checks past statements and documentation. If a user has previously said, "The project is due this Friday," a contradiction is detected.

[0599] In this case, the user will see a warning message like this:

[0600] "There is a contradiction with your previous statement. Previously you said, 'The project deadline is this Friday.' Would you like to correct that?"

[0601] Examples of prompt statements

[0602] Examples of prompts to input to a generative AI model include:

[0603] "A user has recently stated, 'The project deadline is next Friday.' Please check for inconsistencies against previous statements and documentation."

[0604] In this way, the system helps users maintain consistency in what they say.

[0605] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0606] Step 1:

[0607] A user types a comment into a chat tool.

[0608] The user types a new statement into the chat tool's input field, for example, "The project is due next Friday," and the statement is sent to the next processing step.

[0609] Step 2:

[0610] The device sends the speech content to the plugin.

[0611] The terminal sends the user's input to the plugin. This transmission is done through the chat tool's internal API. The input is the user's input, and the output is the data sent to the plugin.

[0612] Step 3:

[0613] The plugin analyzes the content of the comment.

[0614] The plugin analyzes the received utterances using natural language processing (NLP) techniques. Specifically, it uses a generative AI model (e.g., OpenAI's GPT-3) to understand the meaning of the utterances. The input is the user's utterance, and the output is the analyzed utterance data.

[0615] Step 4:

[0616] The server compares the content of the statement with past statements and created materials.

[0617] The server compares the analyzed comment content with past comments and created documents stored in a database. MySQL or PostgreSQL is used as a database management system (DBMS) for this comparison process. The input is the analyzed comment content and past data, and the output is the comparison result.

[0618] Step 5:

[0619] The server detects the discrepancy.

[0620] The server compares new statements with past statements and documents to check for inconsistencies. For example, if someone previously said, "The project deadline is this Friday," a contradiction is detected. The input is the comparison result, and the output is information about whether there is a contradiction.

[0621] Step 6:

[0622] The server sends the conflicting information back to the terminal.

[0623] The server returns the detected conflict information to the terminal, including the specific conflicts and suggested corrections. The input is the conflict information, and the output is the data transmission to the terminal.

[0624] Step 7:

[0625] The device will display a warning message to the user.

[0626] The terminal displays a warning message to the user, for example, "There is a contradiction with your previous statement. You previously said, 'The project deadline is this Friday.' Would you like to correct it?" The input is the contradiction information from the server, and the output is the warning message to the user.

[0627] Step 8:

[0628] The user edits what was said.

[0629] The user checks the warning message and modifies the comment if necessary. The modified comment is sent to the plugin again, where the same check is performed. The input is the warning message, and the output is the modified comment.

[0630] (Application example 1)

[0631] Next, a description will be given of Application Example 1 of Form 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".

[0632] Work instructions and reports within the factory may contain information that contradicts previous statements or documents, which can lead to reduced work efficiency and safety. In addition, there is no way to immediately respond to contradictions, which can lead to problems becoming more serious.

[0633] 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. In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing work instructions and report contents in the factory in real time and checking for inconsistencies by comparing them with past instructions and report contents, and a means for generating a warning message and presenting it to the user when an inconsistency is detected. This maintains the consistency of work instructions and report contents in the factory, enabling efficient and safe work.

[0634] "User" means an individual or organization that uses the System.

[0635] "Past comments" are records of comments a user has made in the past.

[0636] "Created Materials" means documents and data records created by a User.

[0637] "Input" refers to data or information that is entered into a system.

[0638] "Content to be posted" refers to the new information or instructions that the user wishes to state.

[0639] "Contradiction" refers to a situation in which new statements are inconsistent with or in conflict with previous statements or documents.

[0640] "Suggest" refers to the system making suggestions or advice to the user.

[0641] "Present" refers to the system displaying information to the user.

[0642] "Factory" refers to the facility or location where manufacturing operations take place.

[0643] "Work instructions" are information that instructs the specific work content and procedures to be performed in a factory.

[0644] "Report content" refers to information about the progress and results of work.

[0645] "Real-time" refers to processing and display occurring immediately, without delay.

[0646] "Analysis" refers to the detailed examination of data or information to find meaning and patterns.

[0647] A "warning message" is a notification or alert intended to alert the user.

[0648] To implement this invention, a specific program must be installed on the robots and terminals used in the factory. This program has the function of receiving the user's past comments and documents as input, and suggesting whether there are any inconsistencies in the content to be posted in the future. It also includes a function of analyzing the work instructions and report contents in the factory in real time, and checking for inconsistencies by comparing them with past instructions and reports.

[0649] The server uses the following hardware and software:

[0650] Hardware: Factory robots

[0651] Software: Python, JSON files, natural language processing technology

[0652] The server first reads past comments and created documents from a JSON file. Then, when a new comment is entered, it checks for inconsistencies by comparing it with past comments and created documents. If a contradiction is detected, a warning message is generated and displayed to the user. If there is no contradiction, the new comment is added to the list of past comments and saved in a JSON file.

[0653] As a concrete example, consider instructions and reports for machine maintenance in a factory. If a past statement such as "Maintenance of machine A has started" is recorded, and a new statement such as "Maintenance of machine A has been completed" is entered, the system will compare these statements to check for inconsistencies.

[0654] Here are some example prompts using a generative AI model:

[0655] Previously stated: "Maintenance has begun on Machine A."

[0656] New statement: "Maintenance on Machine A has been completed."

[0657] Check for inconsistencies.

[0658] In this way, consistency in work instructions and reports within the factory can be maintained, enabling efficient and safe work.

[0659] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0660] Step 1:

[0661] The server reads past comments and created documents from a JSON file. Specifically, it opens a JSON file at a specified path from the file system and converts its contents into a data structure (list or dictionary). The input is the path to the JSON file, and the output is the data structure of past comments and created documents.

[0662] Step 2:

[0663] The user inputs a new utterance. The terminal receives the input from the user and sends the contents to the server. The input is the user's new utterance, and the output is the utterance data to be sent to the server.

[0664] Step 3:

[0665] The server compares the new statement with past statements and documents to check for inconsistencies. Specifically, it loops through the data structures of past statements and documents and compares them with the new statement. The input is the new statement and the data structures of past statements and documents, and the output is the presence or absence of inconsistencies and a list of contradictory statements.

[0666] Step 4:

[0667] The server generates a warning message when a contradiction is detected. Specifically, the server creates a warning message based on the list of contradictory statements and presents it to the user. The input is the list of contradictory statements, and the output is the warning message.

[0668] Step 5:

[0669] If no contradictions are detected, the server adds the new comment to the list of past comments and saves it in a JSON file. Specifically, the server adds the new comment to a data structure, converts the data structure to JSON format, and writes it to a file. The input is the new comment and the data structure of past comments and created materials, and the output is the updated JSON file.

[0670] Step 6:

[0671] The server notifies the user that the new comment has been added successfully. Specifically, it sends a message to the user's terminal to inform them that the comment has been added. The input is the result of adding the new comment, and the output is a notification message to the user.

[0672] Example 2

[0673] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[0674] Conventional chat tools and dialogue tools lack the functionality to automatically check for inconsistencies with past comments or documents when a user makes a new comment. This means that users must manually check past comments and documents, which is inefficient and increases the risk of sending incorrect information. The present invention aims to solve these problems and enable users to send information efficiently and accurately.

[0675] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0676] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the past comments and created materials using natural language processing technology, a means for checking whether there are any inconsistencies between a new comment and the past comments and created materials using a generative AI model, and a means for displaying a warning message to the user and suggesting corrections when a inconsistency is detected. This enables the user to automatically check for inconsistencies with past comments and created materials when making a new comment, and to transmit information efficiently and accurately.

[0677] "User" means an individual or organization that uses the system to make statements or create materials.

[0678] "Past comments" are records of comments or statements a user has previously made through the system.

[0679] "Created Materials" means documents and data records created by Users through the System.

[0680] "Input" refers to data or information entered into a system.

[0681] "Suggestion" means that the system makes suggestions or advice to the user.

[0682] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[0683] A "generative AI model" is a model that uses artificial intelligence to generate and analyze data.

[0684] A "contradiction" is a state in which there is a discrepancy or inconsistency between a new statement and previous statements or documents.

[0685] A "warning message" is a message that the system displays to alert the user.

[0686] "Suggesting corrections" means that the system suggests to the user to correct the content of their statements or materials.

[0687] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0688] The present invention is a system that automatically checks for inconsistencies with past statements and created materials when a user makes a new statement, enabling the user to disseminate information efficiently and accurately. The system includes a server, a user terminal, and a generative AI model.

[0689] Hardware and Software Used

[0690] Hardware: Servers, user devices (PCs, smartphones, etc.)

[0691] Software: Generative AI models (e.g. GPT-4), Natural Language Processing libraries (e.g. spaCy, NLTK), Database Management Systems (e.g. MySQL)

[0692] System Operation

[0693] 1. The user enters a new comment

[0694] The user uses the terminal to enter a new utterance into an input field of the interactive tool. For example, the user enters, "The start date for the new project is next Monday."

[0695] 2. The server retrieves past comments and documents

[0696] The server connects to the database management system and queries the user's past comments and documents. For example, a past comment such as "The project start date is the Monday after next" is retrieved from the database.

[0697] 3. The server performs natural language processing

[0698] The server uses natural language processing libraries to analyze the retrieved past utterances and written materials, specifically tokenizing sentences, tagging parts of speech, and analyzing dependencies.

[0699] 4. The server performs a conflict check

[0700] The server uses the generative AI model to check whether the new statement is consistent with previous statements or documents. For example, the generative AI model is given the following prompt: "The user has entered a new statement. Please check whether there are any inconsistencies with previous statements or documents. New statement: 'The start date of the new project is next Monday.' Previous statement: 'The start date of the project is the Monday after next.'"

[0701] 5. The server notifies the user of the result

[0702] If a conflict is detected, the server generates a warning message for the user and displays it on the terminal, for example "Isn't the start date of your new project on Monday after next?", and suggests ways to correct the conflict.

[0703] Examples

[0704] User input: "The start date for my new project is next Monday."

[0705] Previously stated: "The project start date is the Monday after next."

[0706] In this case, the system proceeds as follows.

[0707] 1. A user uses a terminal and types, "The start date for my new project is next Monday."

[0708] 2. The server retrieves the past statement "The project start date is the Monday after next" from the database.

[0709] 3. The server uses a natural language processing library to analyze past utterances.

[0710] 4. The server uses the generative AI model to check for inconsistencies between new and previous utterances.

[0711] 5. Because an inconsistency is detected, the server displays a warning message to the user asking "Isn't the start date of your new project on Monday after next?" and suggests a correction.

[0712] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[0713] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0714] Program processing flow

[0715] Step 1: User enters new utterance

[0716] What happens: The user uses the terminal to enter a new utterance into the input field of the interactive tool.

[0717] Input: The new utterance typed by the user (e.g., "The start date for the new project is next Monday.")

[0718] Output: The user's new utterance is sent to the system.

[0719] Step 2: The server retrieves past statements and documents

[0720] Specific operation: The server connects to a database management system and queries and retrieves the user's past comments and created materials.

[0721] Input: User ID or user identification

[0722] Data processing: Search the database for past statements and created materials and extract relevant data.

[0723] Output: Data from past statements and documents (e.g. "The project start date is the Monday after next week.")

[0724] Step 3: The server performs natural language processing

[0725] Specific operation: The server uses a natural language processing library to analyze the retrieved past statements and written materials.

[0726] Input: Data of past statements and created documents

[0727] Data processing: Sentence tokenization, part-of-speech tagging, dependency analysis, etc.

[0728] Output: Structured data of analyzed past statements and documents

[0729] Step 4: Server performs inconsistency check

[0730] How it works: The server uses a generative AI model to check for inconsistencies between new statements and previous statements or created materials.

[0731] Input: New statements, parsed past statements, and structured data of created documents

[0732] Data calculation: Input a prompt sentence into the generative AI model and perform inconsistency checks.

[0733] Example prompt: "The user has entered a new statement. Please check for inconsistencies with previous statements and documentation. New statement: 'The new project start date is next Monday.' Previous statement: 'The project start date is the Monday after next.'"

[0734] Output: Result on whether there is a conflict (e.g., conflict detected)

[0735] Step 5: The server notifies the user of the result

[0736] Specific Behavior: If a conflict is detected, the server generates a warning message for the user and displays it on the terminal.

[0737] Input: Result regarding whether or not there is a conflict

[0738] Data processing: Generate warning messages (e.g. "Isn't the start date of your new project on Monday after next?")

[0739] Output: A warning message that is displayed on the user's terminal.

[0740] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[0741] (Application example 2)

[0742] Next, a description will be given of Application Example 2 of Form 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".

[0743] Conventional dialogue tools and chat tools lack the functionality to automatically check for inconsistencies with past statements and prepared documents when a user makes a new statement. This makes it difficult to maintain consistency of important information, particularly in security services, and there is a risk of incorrect information being transmitted. The present invention aims to solve these problems and provide a system that automatically checks for inconsistencies with past statements and prepared documents when a user makes a new statement, and suggests corrections as necessary.

[0744] 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.

[0745] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for using a generative AI model to suggest corrections when a contradiction is detected. This makes it possible to automatically check for inconsistencies with past comments and created materials when the user makes a new comment, and to suggest corrections as necessary.

[0746] "User" means an individual or entity that uses the System to input statements or create materials.

[0747] "Past comments" are records of comments or statements a user has previously made within the system.

[0748] "Created Materials" refers to documents and data created or uploaded by a User within the System.

[0749] "Input" refers to the data or information that a system receives for processing.

[0750] "Content to be posted" refers to the new information or comments that a user intends to make within the system.

[0751] "Inconsistency" refers to a situation in which new statements or posts are inconsistent with or in conflict with previous statements or materials.

[0752] "Suggestions" refers to suggestions or advice that the system provides to the user.

[0753] A "generative AI model" refers to an algorithm or program that uses artificial intelligence techniques to analyze data and generate appropriate suggestions or revisions.

[0754] "Present" refers to the act of the system displaying information or suggestions to the user.

[0755] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for using a generative AI model to suggest corrections when an inconsistency is detected.

[0756] When a user inputs a new comment, the server retrieves past comments and written materials from the database and analyzes them. Natural language processing technology is used for the analysis to check for inconsistencies between the past comments and written materials and the new comment. If a contradiction is detected, a generative AI model is used to generate a correction proposal and present it to the user.

[0757] Specifically, the server performs the following process. First, when a user inputs a new statement, the statement is sent to the server. The server retrieves past statements and created documents from a database and analyzes this data using natural language processing technology. For example, OpenAI's API can be used for the analysis. If the analysis detects a contradiction between the new statement and past statements or created documents, the server uses a generative AI model to generate a correction proposal. This correction proposal is presented to the user as a warning message.

[0758] The hardware used includes a server and a user's terminal. The server is a computer equipped with a high-performance processor and a large amount of memory, and the user's terminal is a general device such as a smartphone or a PC. The software used includes libraries for implementing natural language processing technology (e.g., Python's difflib library) and APIs for executing generative AI models (e.g., OpenAI's API).

[0759] As a concrete example, consider the case of checking statements regarding security policies. Assume that a past statement recorded is "All passwords must be at least eight characters long." If the user attempts to newly state "All passwords must be at least six characters long," the server analyzes this new statement and detects any inconsistencies with the past statement. If a contradiction is detected, the generative AI model is used to generate a prompt sentence like the one below, and a correction suggestion is presented.

[0760] Example prompt:

[0761] New statement: All passwords must be at least six characters long.

[0762] Previous contradictory statement: All passwords must be at least eight characters long.

[0763] Generate suggestions for fixing the inconsistencies.

[0764] In this way, when a user makes a new statement, the system automatically checks for inconsistencies with previous statements and written materials, and suggests corrections if necessary.

[0765] The flow of the specific process in the application example 2 will be described with reference to FIG.

[0766] Step 1:

[0767] The user inputs a new comment. The input comment is sent from the user's terminal to the server. The input data is the text of the new comment.

[0768] Step 2:

[0769] The server retrieves the user's past comments and created documents from the database. The retrieved data is the text of the past comments and created documents.

[0770] Step 3:

[0771] The server uses natural language processing technology to analyze new comments and past comments and created documents. Specifically, it uses Python's difflib library to calculate the similarity between comments and documents. The input data is the text of the new comment and past comments and created documents, and the output data is the similarity score.

[0772] Step 4:

[0773] The server checks whether there are any inconsistencies between the new comment and past comments or documents based on the analysis results. If a inconsistency is detected, the information is recorded. The input data is the similarity score, and the output data is a flag indicating whether or not there is a inconsistency.

[0774] Step 5:

[0775] If a contradiction is detected, the server uses a generative AI model to generate a correction proposal. Specifically, it uses OpenAI's API to generate a prompt sentence and obtain a correction proposal. The input data is the text of past statements or created documents that contradict the new statement, and the output data is the text of the correction proposal.

[0776] Step 6:

[0777] The server presents the generated revision proposal to the user, which is displayed as a warning message on the user's terminal. The input data is the text of the revision proposal, and the output data is the warning message presented to the user.

[0778] Step 7:

[0779] The user checks the proposed revisions and revises the new utterance if necessary. The revised utterance is sent to the server again, and the same process is repeated. The input data is the text of the revised utterance, and the output data is the text of the final utterance.

[0780] In addition, an emotion engine that estimates the emotion of the user may be further combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[0781] "Example 1"

[0782] As one embodiment of the present invention, a chat tool plug-in or interactive tool combined with an emotion engine is provided. The emotion engine analyzes the user's emotional state from the user's past comments and documents created. Specifically, it analyzes the text of messages posted by the user in the past and documents created by the user, and extracts emotional expressions and keywords suggesting emotions contained therein. Then, the user's emotional state is estimated based on this information. For example, if the user has used many positive expressions such as "fun" and "happy" in the past, the user's emotional state is estimated to be "joy" or "happiness." The emotional state estimated in this way is used as reference information when suggesting whether there are any inconsistencies in the content the user will post in the future.

[0783] "Example 2"

[0784] In another embodiment of the present invention, a system is provided in which an emotion engine analyzes a user's emotional state using natural language processing technology and emotion recognition technology. Specifically, natural language processing technology is used to analyze text, and expressions and keywords indicating emotions are extracted from the text. Next, emotion recognition technology is used to analyze the extracted expressions and keywords to estimate the user's emotional state. For example, if a user uses a lot of negative expressions such as "sad" and "disappointed," the user's emotional state is estimated to be "sadness" or "disappointment." The emotional state estimated in this way is used as reference information when suggesting whether the content the user is about to post is consistent with the user's emotional state.

[0785] The process flow of each embodiment will be described below.

[0786] "Example 1"

[0787] Step 1: User types a message into the chat tool.

[0788] Step 2: The emotion engine analyzes the user's emotional state from their past statements and documents. Specifically, it analyzes the text of past messages and documents and extracts emotional expressions and keywords that suggest emotions.

[0789] Step 3: The emotion engine estimates the user's emotional state based on the extracted information.

[0790] Step 4: Based on the estimated emotional state, we suggest inconsistencies in the content the user will post next.

[0791] "Example 2"

[0792] Step 1: User types a message into the chat tool.

[0793] Step 2: The emotion engine uses natural language processing techniques to analyze the text and extract expressions and keywords that indicate emotions.

[0794] Step 3: The emotion engine uses emotion recognition technology to analyze the extracted expressions and keywords and infer the user's emotional state.

[0795] Step 4: Based on the estimated emotional state, we suggest whether the content of the user's future post is consistent with that emotional state.

[0796] Example 1

[0797] Next, a description will be given of Example 1 of Form 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".

[0798] Conventional chat tools and conversational tools lack the ability to automatically detect and warn users when their statements contradict previous statements or written materials. In addition, suggestions do not take into account the user's emotional state, making it difficult to provide appropriate advice that is in line with the user's intentions. This can lead to a lack of consistency in communication, leading to misunderstandings and confusion.

[0799] 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.

[0800] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there is any inconsistency in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the user's comments using natural language processing technology, a means for comparing the analyzed comments with the past comments and created materials stored in the database, a means for displaying a warning message to the user when a contradiction is detected, a means for analyzing the user's emotional state from the user's past comments and created materials, and a means for making suggestions based on the analyzed emotional state. This makes it possible to automatically check whether the user's comments are inconsistent with the past comments and created materials, and to provide appropriate warnings and suggestions. In addition, by making suggestions taking the user's emotional state into consideration, more consistent communication can be achieved.

[0801] "User" refers to an individual or organization that makes statements or creates materials using chat tools or interactive tools.

[0802] A "utterance" is a text message entered by a user through a chat or interactive tool.

[0803] "Created Materials" refers to documents and data files previously created by the User, including reports, memos, presentation materials, etc.

[0804] "Input" refers to the data or information a system receives for analysis or comparison.

[0805] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0806] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language, and includes tokenization, grammatical analysis, and semantic analysis.

[0807] A "database" refers to a collection of information that stores past statements and created materials and can be used for searching and comparison.

[0808] "Contradiction" refers to an inconsistency that occurs when a user's new statements do not match previous statements or materials they have created.

[0809] "Warning message" refers to a notification sent by the system to the user to inform them of a discrepancy and encourage them to correct it.

[0810] "Emotional state" refers to a user's psychological state inferred from their past statements and written materials, and may include joy, sadness, anger, etc.

[0811] "Analysis" refers to the process by which the system processes the data it receives and understands its meaning and structure.

[0812] "Comparison" refers to the process of comparing the analyzed statements with past statements and created materials to check for matches and inconsistencies.

[0813] The present invention is a system that automatically checks whether there are any inconsistencies with past statements or documents created when a user uses a chat tool or an interactive tool, and provides appropriate suggestions. A specific embodiment of this system will be described below.

[0814] System Configuration

[0815] This system is primarily composed of three elements: a server, a terminal, and a user. The server is responsible for analyzing what users say and comparing it with past comments and created materials. Terminals are devices that users use to operate chat tools and interactive tools, such as PCs and smartphones. Users are individuals or groups who use chat tools and interactive tools to make comments and create materials.

[0816] Hardware and Software Used

[0817] The server is a computer system equipped with a high-performance processor and a large amount of memory. The database uses a relational database management system (RDBMS) or a NoSQL database to store past statements and created materials. Specifically, MySQL is one such example.

[0818] The analysis uses natural language processing technology. Specifically, OpenAI's GPT-3 and Google's BERT are used as generative AI models. These models are used to tokenize the user's speech and analyze its grammar and meaning.

[0819] Data processing and calculation

[0820] When a user enters a new comment, the server receives the comment and analyzes it using natural language processing technology. The analyzed comment is compared with past comments and created materials stored in the database. This comparison is made using Elasticsearch, a full-text search engine. Based on the comparison results, it is determined whether there are any inconsistencies, and if any are detected, a warning message is displayed to the user.

[0821] In addition, an emotion engine is used to analyze the user's emotional state from their past comments and written materials, using tools such as IBM Watson's Natural Language Understanding. The results of the analysis are used to estimate the user's emotional state and are used as reference information when suggesting consistency in future posts.

[0822] Examples

[0823] For example, consider a case where a user says in a chat tool, "A new project has started." If this statement contradicts a previous statement, "The project has not started yet," the server will warn, "This statement contradicts your previous statement. Do you want to correct it?"

[0824] Example prompt:

[0825] The user says "A new project has started", which contradicts their previous statement "The project hasn't started yet". Show a warning message and suggest a fix.

[0826] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[0827] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0828] Step 1:

[0829] The user types a new comment into the chat tool.

[0830] Input: User says "I've started a new project."

[0831] Output: The content of the comment is sent to the chat tool.

[0832] Specific behavior: The user types text into a chat window and clicks the send button.

[0833] Step 2:

[0834] The server receives the user's comments and analyzes them using natural language processing technology.

[0835] Input: User says "I've started a new project."

[0836] Output: Tokenization and grammar analysis of the parsed utterance.

[0837] How it works: The server uses a generative AI model (e.g., GPT-3) to tokenize the utterance and perform grammar analysis.

[0838] Step 3:

[0839] The server compares the analyzed content of the comments with past comments and created materials stored in a database.

[0840] Input: Analyzed statements and past statements and written materials in the database.

[0841] Output: Comparison result (conflicts or not).

[0842] Specific behavior: The server uses a full-text search engine (e.g., Elasticsearch) to search for matches to the past utterance "The project hasn't started yet."

[0843] Step 4:

[0844] The server determines whether there is a conflict based on the comparison.

[0845] Input: The comparison result.

[0846] Output: The result of determining whether there is a contradiction.

[0847] Specific behavior: The server analyzes the comparison result and determines that "A new project has started" and "The project has not started yet" are inconsistent.

[0848] Step 5:

[0849] The server will display a warning message to the user if a discrepancy is detected.

[0850] Input: The result of the determination as to whether there is a contradiction or not.

[0851] Output: A warning message.

[0852] Specific behavior: The server displays a warning message to the user saying, "This statement contradicts a previous statement. Would you like to correct it?"

[0853] Step 6:

[0854] The server uses an emotion engine to analyze the user's emotional state from their past statements and written materials.

[0855] Input: Past statements and created materials.

[0856] Output: The parsed emotional state.

[0857] Specific operation: The server uses a sentiment analysis tool (e.g., IBM Watson) to extract positive sentiment from past statements such as "It was fun" and "I'm happy."

[0858] Step 7:

[0859] The server makes suggestions based on the analyzed emotional state.

[0860] Input: Parsed emotional state and current speech.

[0861] Output: Suggestions that take emotional state into account.

[0862] Specific behavior: The server suggests to the user, "Even though you've contradicted your previous statements, would you like to emphasize that the project is going well?"

[0863] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[0864] (Application example 1)

[0865] Next, a description will be given of Application Example 1 of Form 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".

[0866] In modern companies, improving the efficiency of internal communication and preventing information leaks are important issues. In particular, when communicating using chat tools, it is necessary to avoid inconsistencies with past statements and documents, to grasp the user's emotional state and detect abnormal behavior early, and to prevent the leaking of confidential information. However, it is difficult to realize these functions in a single system, and currently it is necessary to use multiple tools at the same time. This increases the burden on users and makes efficient operation difficult.

[0867] 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.

[0868] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing the user's emotional state and displaying a warning when an abnormal emotion is detected, and a means for detecting comments containing confidential information and displaying a warning. This makes it possible to avoid inconsistencies in internal communication, early detection of abnormal emotional states, and prevention of confidential information leaks.

[0869] "User" refers to any individual or entity that uses the System.

[0870] "Past comments" refers to the content of chats or messages a user has previously made.

[0871] "Created materials" refers to text data such as documents and reports created by users.

[0872] "Input" refers to data or information entered into a system.

[0873] "Contradiction" refers to a discrepancy or inconsistency that arises between past statements or prepared materials and new statements.

[0874] "Suggestions" refers to suggestions or advice that the system presents to the user.

[0875] "Emotional state" refers to a user's psychological state or mood.

[0876] "Analysis" refers to the process of analyzing data or information to extract meaning and patterns.

[0877] "Abnormal emotions" refers to negative emotions or abnormal psychological states that deviate from normal emotional states.

[0878] "Warning" refers to a message or notification that alerts the user.

[0879] "Confidential information" refers to important information that could cause harm if leaked to the outside.

[0880] "Discovery" refers to the process of finding specific conditions or patterns.

[0881] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the user's emotional state and displaying a warning if abnormal emotions are detected, and means for detecting comments that include confidential information and displaying a warning.

[0882] System Programming

[0883] This system is realized using the following hardware and software.

[0884] Hardware

[0885] Server: Manages the database and performs natural language processing and sentiment analysis.

[0886] User device: A smartphone or computer with a chat tool installed.

[0887] software

[0888] Natural Language Processing libraries: TfidfVectorizer, cosine_similarity (using Python's scikit-learn library)

[0889] Sentiment analysis library: TextBlob (Python library)

[0890] Database management system: A database for storing and managing past statements and created materials (e.g., MySQL, PostgreSQL)

[0891] Processing flow

[0892] 1. Receiving input:

[0893] The server retrieves the user's past comments and created materials from a database and receives new comments as input.

[0894] 2. Suggested contradictions:

[0895] The server uses TfidfVectorizer to vectorize new and past comments, and calculates cosine_similarity. If the similarity is low, it is determined that there is a contradiction and presents suggestions to the user.

[0896] 3. Emotional state analysis:

[0897] The server uses TextBlob to calculate the sentiment score of new utterances, and if the sentiment score is abnormal, it displays a warning to the user.

[0898] 4. Detecting Sensitive Information:

[0899] The server uses regular expressions to check whether new posts contain sensitive information, and if so, displays a warning to the user.

[0900] Examples

[0901] Suggested contradictions:

[0902] Previously stated: "Project X is progressing well."

[0903] New statement: "Progress on Project X is slow."

[0904] Result: The inconsistency is detected and the user is prompted with the suggestion, "This statement contradicts something you said previously. Would you like to fix it?"

[0905] Emotional state analysis:

[0906] Say: "I really hate this project."

[0907] Result: Negative sentiment is detected and the user is alerted with "Negative sentiment detected. Is something wrong?"

[0908] Sensitive information detection:

[0909] Say: "This is classified information."

[0910] Result: Because it contains sensitive information, the user will see a warning saying "This post contains sensitive information. Do you want to send this post?"

[0911] Examples of prompt statements

[0912] "Please create a program that checks for inconsistencies in what the user says against past statements and documents, and displays a warning if any inconsistencies are detected. Please also analyze the user's emotional state, and display a warning if abnormal emotions are detected. Furthermore, please add a function to display a warning if confidential information is included."

[0913] In this way, it is possible to avoid inconsistencies in internal communication, detect abnormal emotional states early, and prevent the leakage of confidential information.

[0914] The flow of the specific process in the application example 1 will be described with reference to FIG.

[0915] Step 1:

[0916] The server retrieves the user's past comments and created materials from a database. This provides the past comments and created materials to the server as input. Specifically, a database query is executed to extract past comments and created materials related to the user ID.

[0917] Step 2:

[0918] The server receives the new comment entered by the user. This provides the new comment to the server as input. Specifically, the server receives an API request from the chat tool and obtains the new comment content.

[0919] Step 3:

[0920] The server vectorizes the new and past comments using TfidfVectorizer, which converts the text data into numerical vectors. Specifically, the new and past comments are input into TfidfVectorizer as a single corpus, and a TF-IDF vector is generated for each piece of text.

[0921] Step 4:

[0922] The server calculates cosine_similarity to evaluate the similarity between the new and past comments. This determines whether there is a contradiction. Specifically, it calculates the cosine similarity between the vector of the new comment and the vector of the past comment, and determines that there is a contradiction if the similarity is low.

[0923] Step 5:

[0924] If a contradiction is detected, the server presents a suggestion to the user, giving the user an opportunity to correct the contradiction. Specifically, the server displays a message through the chat tool interface saying, "This is inconsistent with your previous statement. Do you want to correct it?"

[0925] Step 6:

[0926] The server uses TextBlob to calculate the sentiment score of a new comment, which analyzes the sentiment state of the comment. Specifically, a new comment is input to TextBlob and a positive / negative sentiment score is obtained.

[0927] Step 7:

[0928] If the emotion score is abnormal, the server displays a warning to the user. This allows the user to become aware of the abnormal emotional state. Specifically, if the emotion score falls below a certain threshold, the server displays a message saying, "Negative emotion has been detected. Is there a problem?"

[0929] Step 8:

[0930] The server uses regular expressions to check whether new comments contain confidential information. This is how confidential information is detected. Specifically, the server searches new comments for keywords that indicate confidential information (e.g., "confidential" or "secret"), and if any are found, it determines that the new comment contains confidential information.

[0931] Step 9:

[0932] If confidential information is included, the server displays a warning to the user. This allows the user to prevent confidential information from being leaked. Specifically, the server displays a message saying, "Confidential information is included. Do you want to send this message?"

[0933] Example 2

[0934] Next, a description will be given of Example 2 of the embodiment 2. In the following description, the data processing device 12 is referred to as a "server" and the smart glasses 214 are referred to as a "terminal".

[0935] In conventional dialogue and chat tools, there is a risk that users may post content that contradicts their past statements or documents they have created. In addition, since suggestions are made without taking into account the user's emotional state, advice that does not match the user's intentions or emotions may be provided. For this reason, there is a demand for a system that provides appropriate suggestions that take into account the user's past statements and emotional state.

[0936] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0937] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether the content to be posted from now on is consistent with inconsistencies, a means for presenting the suggestions to the user, a means for analyzing the user's comments and estimating the user's emotional state, and a means for suggesting whether the content to be posted from now on is consistent with the emotional state based on the estimated emotional state. This makes it possible to make appropriate suggestions taking into account the user's past comments and emotional state.

[0938] "User" refers to an individual or organization that uses the System to create and post comments and materials.

[0939] "Comments" refers to text data that users input and post through the system.

[0940] "Created Materials" refers to documents and data that a User has previously created and stored through the System.

[0941] "Input" refers to the data or information a system receives for processing.

[0942] "Suggestion" refers to suggestions or advice that the system provides to the user.

[0943] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[0944] "Emotional state" refers to the psychological state or emotion inferred from a user's statements.

[0945] "Contradiction" refers to a situation where a user's new statements are inconsistent with or contradict previous statements or materials created.

[0946] "Managed online" means that data and information are stored and managed in a manner that makes them accessible via the Internet.

[0947] "Analysis" refers to the process by which a system examines and evaluates data or information in detail.

[0948] "Inference" refers to the process by which a system predicts a state or outcome based on data and information.

[0949] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0950] The present invention is a system that receives a user's past comments and documents as input, and suggests whether the content to be posted in the future is consistent with the user's past comments and documents. Furthermore, the system has a function of analyzing the user's comments and estimating the user's emotional state, and then suggesting whether the content to be posted in the future is consistent with the user's emotional state based on the estimated emotional state.

[0951] Hardware and Software Used

[0952] The present system is implemented using the following hardware and software.

[0953] Server: Equipped with a database management system (DBMS), it stores and manages users' past comments and documents created by users. The server also provides the computing resources to execute natural language processing and emotion recognition technologies.

[0954] Terminal: The device (PC, smartphone, tablet, etc.) that a user uses to enter utterances.

[0955] Natural Language Processing libraries: Use Python's NLTK library and spaCy to tokenize and semantically analyze text.

[0956] Emotion Recognition Library: Analyzes and estimates the user’s emotional state using Python’s TextBlob and VADER Sentiment Analysis.

[0957] System Operation

[0958] 1. Check before user comments are entered

[0959] When a user enters a new comment, the server retrieves the user's past comments and created materials from the database.

[0960] The server uses natural language processing technology to analyze the content of past statements and documents, and check for any contradictions with new statements.

[0961] Example of how it works: When a user tries to type "The project is on track," the server retrieves the previous statement "The project is delayed" and detects the inconsistency.

[0962] 2. Detecting inconsistencies and displaying warning messages

[0963] If the server detects a discrepancy, it will display a warning message to the user.

[0964] Specific example of behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[0965] 3. Analyzing the user's emotional state

[0966] When a user enters a comment, the server uses natural language processing technology to analyze the text and extract expressions and keywords that indicate emotions.

[0967] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[0968] Example of specific operation: If a user enters "I'm sad that things haven't been going well at work recently," the server will extract the keyword "sad" and estimate the emotional state as "sad."

[0969] 4. Using Emotional States to Suggest

[0970] Based on the estimated emotional state, the system suggests whether the content the user will post next is consistent with that emotional state.

[0971] An example of how this works: When a user tries to type "This project is awesome," the server suggests "Do you really feel that way?"

[0972] Examples of prompt statements

[0973] Conflict check prompt

[0974] Compare the user's previous statements with their new statements and check for inconsistencies. Previous statement: "The project is delayed," New statement: "The project is on track."

[0975] Sentiment Analysis Prompt

[0976] Analyze user utterances and infer their emotional state. utterance: "I'm sad that things haven't been going well at work lately."

[0977] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0978] Step 1:

[0979] The user inputs a new comment.

[0980] Input: The user types text into an input field on the terminal.

[0981] What happens: The user types, "The project is going well."

[0982] Output: The entered text data is sent to the server.

[0983] Step 2:

[0984] The server retrieves the user's past comments and created materials from the database.

[0985] Input: A user's ID and a request for past comments.

[0986] What happens: The server runs a SQL query to retrieve the user's previous statement "The project is delayed."

[0987] Output: Data of past comments and created materials is returned to the server.

[0988] Step 3:

[0989] The server uses natural language processing technology to analyze the content of past statements and documents.

[0990] Input: Text data of past statements and created documents.

[0991] How it works: The server uses Python's NLTK library to tokenize past utterances and perform semantic analysis.

[0992] Output: Structured information of the parsed text data.

[0993] Step 4:

[0994] The server compares the new message with previous messages to check for inconsistencies.

[0995] Input: Text data of new utterances and structured information of previous utterances parsed.

[0996] What happens: The server compares the new statement "The project is on track" with the previous statement "The project is delayed" and detects the inconsistency.

[0997] Output: Information about whether there are any conflicts.

[0998] Step 5:

[0999] If the server detects a discrepancy, it displays a warning message to the user.

[1000] Input: Information about whether there is a conflict.

[1001] Specific behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[1002] Output: A warning message that is displayed on the user's terminal.

[1003] Step 6:

[1004] The server uses natural language processing technology to analyze the user's comments and extract expressions and keywords that indicate emotions.

[1005] Input: The text of the user's new utterance.

[1006] Specific operation: The server uses Python's TextBlob library to extract the keyword "sad" from the comment.

[1007] Output: Extracted sentiment expressions and keywords.

[1008] Step 7:

[1009] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[1010] Input: Extracted sentiment expressions and keywords.

[1011] What happens: The server uses VADER Sentiment Analysis to estimate the user's emotional state as "sad."

[1012] Output: The estimated emotional state.

[1013] Step 8:

[1014] Based on the estimated emotional state, the server suggests whether the content the user is about to post is consistent with that emotional state.

[1015] Input: Estimated emotional state and textual data of new utterances.

[1016] Specific behavior: The server responds to the user's new statement, "This project is awesome," by suggesting, "Do you really feel that way?"

[1017] Output: The suggestion message that is displayed on the user's device.

[1018] (Application example 2)

[1019] Next, a description will be given of Application Example 2 of Form 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".

[1020] Traditional chat tools and conversational tools lack the ability to automatically check whether the content a user posts is consistent with past statements or documents. They also lack the ability to analyze the user's emotional state and provide appropriate suggestions based on that emotion. This creates a risk that users will post inconsistent information or make posts that are not in line with their emotions.

[1021] 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.

[1022] In this invention, the server includes a means for receiving the user's past comments and documents as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state. This enables the user to transmit consistent information and post content appropriate to their emotions.

[1023] "User" means any individual or entity that uses the System to post statements or create materials.

[1024] "Past statements and created materials" refers to statements made by users previously on the system, as well as documents and data they created.

[1025] "Input" refers to the data or information a system receives for processing.

[1026] A "contradiction" is when a new statement or production is inconsistent with or contradicts an earlier statement or production.

[1027] A "suggestion" is a suggestion or advice that the system provides to the user.

[1028] "Presentation" refers to the act of the system displaying information to the user.

[1029] An "emotional state" is a state that indicates a user's current feelings or mood.

[1030] "Analysis" is the act of examining data or information in detail to understand its meaning and structure.

[1031] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[1032] "Managed online" means that data and information are stored in a manner that makes them accessible via the internet.

[1033] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state.

[1034] System Configuration

[1035] 1. Hardware:

[1036] Server: A server for storing and processing data.

[1037] Device: The smartphone or computer used by the user.

[1038] 2. Software:

[1039] Natural language processing library: Uses natural language processing technologies such as spaCy.

[1040] Sentiment Analysis Library: Uses sentiment analysis techniques such as TextBlob.

[1041] Program processing explanation

[1042] The server receives input from users' online records of past posts and written materials. It then uses natural language processing technology to compare new posts with past posts and written materials to check for inconsistencies. If a contradiction is detected, the server displays a warning message to the user and suggests correcting the inconsistent content.

[1043] In addition, the server uses emotion analysis technology to analyze the user's emotional state. Based on the analysis results, the server suggests posting content that is consistent with the user's emotional state. This enables users to post information that is consistent with their emotions and appropriate to their feelings.

[1044] Examples

[1045] For example, if a user has previously posted "Today was a very fun day," and then tries to post "Today is very sad," the server detects the inconsistency. In addition, because negative emotions are detected through emotion analysis, the server displays a warning message to the user and suggests that they review the content of their post.

[1046] Examples of prompt statements

[1047] Here is an example of a prompt to be input to the generative AI model:

[1048] A new post from a user reads: "I'm so sad today."

[1049] Previous post: "Today was so much fun."

[1050] Check whether this new post is inconsistent with previous posts and perform a sentiment analysis.

[1051] Using this prompt sentence, the generative AI model can make appropriate suggestions.

[1052] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1053] Step 1:

[1054] The server manages users' past comments and written materials online. These data are contents posted by users in the past and documents created by users, and are stored in a database. The input is the user's past comments and written materials, and the output is a list of these data.

[1055] Step 2:

[1056] The user inputs new post content into the terminal. The terminal sends this input to the server. The input is the new post content entered by the user, and the output is the new post content sent to the server.

[1057] Step 3:

[1058] The server uses natural language processing technology to analyze new posts and past comments and written materials. Specifically, it uses libraries such as spaCy to tokenize the text and perform semantic analysis. The input is the new post and past comments and written materials, and the output is the analysis results.

[1059] Step 4:

[1060] The server uses the analysis results to check whether there are any inconsistencies between the new post and past comments or documents. If a contradiction is detected, the server identifies the content and generates a warning message for the user. The input is the analysis results, and the output is the presence or absence of a contradiction and a warning message.

[1061] Step 5:

[1062] The server analyzes the emotional state of the new post using sentiment analysis techniques, specifically by calculating a sentiment score for the text using a library such as TextBlob. The input is the new post, and the output is the sentiment score.

[1063] Step 6:

[1064] The server estimates the user's emotional state based on the emotion score, and suggests posting content that is consistent with the emotional state. The input is the emotion score, and the output is the suggested content.

[1065] Step 7:

[1066] The server generates warning messages and suggestions and sends them to the user's terminal, which displays them to the user. The input is the warning messages and suggestions, and the output is the information displayed to the user.

[1067] Step 8:

[1068] The user checks the warning messages and suggestions from the server and modifies the new post content as necessary. The input is the information displayed to the user, and the output is the modified post content.

[1069] 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 a voice indicating a user input for the result of the specific processing. The control unit 46A transmits the 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.

[1070] 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> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to 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.

[1071] Another example of generative AI is Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1072] 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 smart glasses 214.

[1073] [Third embodiment]

[1074] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1075] 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.

[1076] 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 wide area network (WAN) and / or a local area network (LAN).

[1077] 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.

[1078] 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 the voice according to instructions from the processor 46.

[1079] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a 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 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[1080] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[1081] Fig. 6 shows an example of 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.

[1082] 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.

[1083] 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.

[1084] In the headset type terminal 314, 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.

[1085] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1086] "Example 1"

[1087] As one embodiment of the present invention, the present invention is implemented as a plug-in for a chat tool. When a user makes a comment on the chat tool, the plug-in compares the comment content with past comments and created materials, and automatically checks for inconsistencies. Specifically, when a user inputs a new comment, the plug-in analyzes the comment content and compares it with past comments and created materials. If an inconsistency is detected, the plug-in displays a warning message to the user and suggests that the inconsistent content be corrected.

[1088] "Example 2"

[1089] Another embodiment of the present invention is implemented as an interactive tool that automatically checks for inconsistencies before a user inputs a new statement by comparing it with past statements and written materials. Specifically, before a user inputs a new statement, the tool analyzes past statements and written materials and checks for inconsistencies with the contents of the new statement. If an inconsistency is detected, the tool displays a warning message to the user and suggests that the user correct the inconsistent contents.

[1090] The process flow of each embodiment will be described below.

[1091] "Example 1"

[1092] Step 1: A user types a new message into the chat tool.

[1093] Step 2: The plugin analyzes the comment and compares it with previous comments and created materials.

[1094] Step 3: If any conflicts are detected, the plugin will display a warning message and suggest you fix the conflicts.

[1095] "Example 2"

[1096] Step 1: Before the user enters a new comment, the interactive tool analyzes past comments and written materials.

[1097] Step 2: Check whether the interactive tool is consistent with the content.

[1098] Step 3: If any conflicts are detected, the interactive tool will display a warning message and suggest ways to fix the conflicts.

[1099] Example 1

[1100] Next, a description will be given of Example 1 of the embodiment 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".

[1101] Conventional chat tools have the problem that users may post content that contradicts past statements or documents created, resulting in inconsistent information. This problem can cause confusion in project progress and communication. Therefore, a system that supports users to avoid contradicting past statements or documents created is required.

[1102] 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.

[1103] In this invention, the server includes means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the content of a new comment when the user inputs the comment, means for comparing the analyzed content of the comment with the past comments and created materials, and means for displaying a warning message to the user when a contradiction is detected and suggesting that the contradictory content be corrected. This makes it possible to maintain consistency in the content of user comments and prevent confusion in the progress of projects and communication.

[1104] "User" refers to an individual or organization that uses chat tools to make statements or create materials.

[1105] "Past comments" refers to records of comments a user previously made on the chat tool.

[1106] "Created Materials" refers to documents and data created by Users on chat tools or related platforms.

[1107] "Input" refers to the data or information a system receives for processing.

[1108] "Suggestion" refers to suggestions or advice that the system provides to the user.

[1109] "Means" refers to a method or apparatus for achieving a specific function or purpose.

[1110] "Chat Tool" refers to software or platform that enables users to exchange messages in real time.

[1111] A "plugin" refers to an extension or module that is added to a chat tool.

[1112] An "interactive tool" refers to software or a platform that allows users and systems to communicate with each other.

[1113] "Managed online" means that data or information is stored in a manner that makes it accessible via the internet.

[1114] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[1115] "Analysis" refers to the process by which a system understands and extracts meaning from input data or information.

[1116] "Comparison" refers to the process in which the system compares new statements with past statements and created materials to detect differences.

[1117] "Warning message" refers to a message that the system displays to alert the user.

[1118] "Correction" refers to the act of a user changing statements or materials to resolve a contradiction.

[1119] This invention is a system that automatically checks for inconsistencies when a user makes a comment on a chat tool by comparing the comment with past comments and created documents. This system is implemented as a plug-in for the chat tool.

[1120] Generating a Program

[1121] The server generates a program implemented as a plug-in for the chat tool. This program has a function of automatically checking for inconsistencies by comparing the content of a user's comment with past comments and created documents when the user comments on the chat tool.

[1122] Program Processing

[1123] When a user types a new comment into the chat tool, the device sends the comment to the plugin, which then uses natural language processing (NLP) techniques to analyze the comment. This analysis can be done using a generative AI model, such as OpenAI's GPT-3.

[1124] The analyzed content of the statement is compared with past statements and documents stored on the server. This comparison process uses MySQL or PostgreSQL as a database management system (DBMS). The server compares the new statement with past statements and documents to check for any inconsistencies.

[1125] If a conflict is detected, the server sends that information back to the device, which then displays a warning message to the user with suggestions for correcting the conflict. The warning message includes the specific conflict and the suggested correction.

[1126] Examples

[1127] For example, if a user says, "The project is due next Friday," the plugin checks past statements and documentation. If a user has previously said, "The project is due this Friday," a contradiction is detected.

[1128] In this case, the user will see a warning message like this:

[1129] "There is a contradiction with your previous statement. Previously you said, 'The project deadline is this Friday.' Would you like to correct that?"

[1130] Examples of prompt statements

[1131] Examples of prompts to input to a generative AI model include:

[1132] "A user has recently stated, 'The project deadline is next Friday.' Please check for inconsistencies against previous statements and documentation."

[1133] In this way, the system helps users maintain consistency in what they say.

[1134] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1135] Step 1:

[1136] A user types a comment into a chat tool.

[1137] The user types a new statement into the chat tool's input field, for example, "The project is due next Friday," and the statement is sent to the next processing step.

[1138] Step 2:

[1139] The device sends the speech content to the plugin.

[1140] The terminal sends the user's input to the plugin. This transmission is done through the chat tool's internal API. The input is the user's input, and the output is the data sent to the plugin.

[1141] Step 3:

[1142] The plugin analyzes the content of the comment.

[1143] The plugin analyzes the received utterances using natural language processing (NLP) techniques. Specifically, it uses a generative AI model (e.g., OpenAI's GPT-3) to understand the meaning of the utterances. The input is the user's utterance, and the output is the analyzed utterance data.

[1144] Step 4:

[1145] The server compares the content of the statement with past statements and created materials.

[1146] The server compares the analyzed comment content with past comments and created documents stored in a database. MySQL or PostgreSQL is used as a database management system (DBMS) for this comparison process. The input is the analyzed comment content and past data, and the output is the comparison result.

[1147] Step 5:

[1148] The server detects the discrepancy.

[1149] The server compares new statements with past statements and documents to check for inconsistencies. For example, if someone previously said, "The project deadline is this Friday," a contradiction is detected. The input is the comparison result, and the output is information about whether there is a contradiction.

[1150] Step 6:

[1151] The server sends the conflicting information back to the terminal.

[1152] The server returns the detected conflict information to the terminal, including the specific conflicts and suggested corrections. The input is the conflict information, and the output is the data transmission to the terminal.

[1153] Step 7:

[1154] The device will display a warning message to the user.

[1155] The terminal displays a warning message to the user, for example, "There is a contradiction with your previous statement. You previously said, 'The project deadline is this Friday.' Would you like to correct it?" The input is the contradiction information from the server, and the output is the warning message to the user.

[1156] Step 8:

[1157] The user edits what was said.

[1158] The user checks the warning message and modifies the comment if necessary. The modified comment is sent to the plugin again, where the same check is performed. The input is the warning message, and the output is the modified comment.

[1159] (Application example 1)

[1160] Next, a description will be given of Application Example 1 of Form 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".

[1161] Work instructions and reports within the factory may contain information that contradicts previous statements or documents, which can lead to reduced work efficiency and safety. In addition, there is no way to immediately respond to contradictions, which can lead to problems becoming more serious.

[1162] 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. In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing work instructions and report contents in the factory in real time and checking for inconsistencies by comparing them with past instructions and report contents, and a means for generating a warning message and presenting it to the user when an inconsistency is detected. This maintains the consistency of work instructions and report contents in the factory, enabling efficient and safe work.

[1163] "User" means an individual or organization that uses the System.

[1164] "Past comments" are records of comments a user has made in the past.

[1165] "Created Materials" means documents and data records created by a User.

[1166] "Input" refers to data or information that is entered into a system.

[1167] "Content to be posted" refers to the new information or instructions that the user wishes to state.

[1168] "Contradiction" refers to a situation in which new statements are inconsistent with or in conflict with previous statements or documents.

[1169] "Suggest" refers to the system making suggestions or advice to the user.

[1170] "Present" refers to the system displaying information to the user.

[1171] "Factory" refers to the facility or location where manufacturing operations take place.

[1172] "Work instructions" are information that instructs the specific work content and procedures to be performed in a factory.

[1173] "Report content" refers to information about the progress and results of work.

[1174] "Real-time" refers to processing and display occurring immediately, without delay.

[1175] "Analysis" refers to the detailed examination of data or information to find meaning and patterns.

[1176] A "warning message" is a notification or alert intended to alert the user.

[1177] To implement this invention, a specific program must be installed on the robots and terminals used in the factory. This program has the function of receiving the user's past comments and documents as input, and suggesting whether there are any inconsistencies in the content to be posted in the future. It also includes a function of analyzing the work instructions and report contents in the factory in real time, and checking for inconsistencies by comparing them with past instructions and reports.

[1178] The server uses the following hardware and software:

[1179] Hardware: Factory robots

[1180] Software: Python, JSON files, natural language processing technology

[1181] The server first reads past comments and created documents from a JSON file. Then, when a new comment is entered, it checks for inconsistencies by comparing it with past comments and created documents. If a contradiction is detected, a warning message is generated and displayed to the user. If there is no contradiction, the new comment is added to the list of past comments and saved in a JSON file.

[1182] As a concrete example, consider instructions and reports for machine maintenance in a factory. If a past statement such as "Maintenance of machine A has started" is recorded, and a new statement such as "Maintenance of machine A has been completed" is entered, the system will compare these statements to check for inconsistencies.

[1183] Here are some example prompts using a generative AI model:

[1184] Previously stated: "Maintenance has begun on Machine A."

[1185] New statement: "Maintenance on Machine A has been completed."

[1186] Check for inconsistencies.

[1187] In this way, consistency in work instructions and reports within the factory can be maintained, enabling efficient and safe work.

[1188] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1189] Step 1:

[1190] The server reads past comments and created documents from a JSON file. Specifically, it opens a JSON file at a specified path from the file system and converts its contents into a data structure (list or dictionary). The input is the path to the JSON file, and the output is the data structure of past comments and created documents.

[1191] Step 2:

[1192] The user inputs a new utterance. The terminal receives the input from the user and sends the contents to the server. The input is the user's new utterance, and the output is the utterance data to be sent to the server.

[1193] Step 3:

[1194] The server compares the new statement with past statements and documents to check for inconsistencies. Specifically, it loops through the data structures of past statements and documents and compares them with the new statement. The input is the new statement and the data structures of past statements and documents, and the output is the presence or absence of inconsistencies and a list of contradictory statements.

[1195] Step 4:

[1196] The server generates a warning message when a contradiction is detected. Specifically, the server creates a warning message based on the list of contradictory statements and presents it to the user. The input is the list of contradictory statements, and the output is the warning message.

[1197] Step 5:

[1198] If no contradictions are detected, the server adds the new comment to the list of past comments and saves it in a JSON file. Specifically, the server adds the new comment to a data structure, converts the data structure to JSON format, and writes it to a file. The input is the new comment and the data structure of past comments and created materials, and the output is the updated JSON file.

[1199] Step 6:

[1200] The server notifies the user that the new comment has been added successfully. Specifically, it sends a message to the user's terminal to inform them that the comment has been added. The input is the result of adding the new comment, and the output is a notification message to the user.

[1201] Example 2

[1202] Next, a second embodiment of the second form will be described. 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".

[1203] Conventional chat tools and dialogue tools lack the functionality to automatically check for inconsistencies with past comments or documents when a user makes a new comment. This means that users must manually check past comments and documents, which is inefficient and increases the risk of sending incorrect information. The present invention aims to solve these problems and enable users to send information efficiently and accurately.

[1204] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1205] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the past comments and created materials using natural language processing technology, a means for checking whether there are any inconsistencies between a new comment and the past comments and created materials using a generative AI model, and a means for displaying a warning message to the user and suggesting corrections when a inconsistency is detected. This enables the user to automatically check for inconsistencies with past comments and created materials when making a new comment, and to transmit information efficiently and accurately.

[1206] "User" means an individual or organization that uses the system to make statements or create materials.

[1207] "Past comments" are records of comments or statements a user has previously made through the system.

[1208] "Created Materials" means documents and data records created by Users through the System.

[1209] "Input" refers to data or information entered into a system.

[1210] "Suggestion" means that the system makes suggestions or advice to the user.

[1211] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1212] A "generative AI model" is a model that uses artificial intelligence to generate and analyze data.

[1213] A "contradiction" is a state in which there is a discrepancy or inconsistency between a new statement and previous statements or documents.

[1214] A "warning message" is a message that the system displays to alert the user.

[1215] "Suggesting corrections" means that the system suggests to the user to correct the content of their statements or materials.

[1216] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1217] The present invention is a system that automatically checks for inconsistencies with past statements and created materials when a user makes a new statement, enabling the user to disseminate information efficiently and accurately. The system includes a server, a user terminal, and a generative AI model.

[1218] Hardware and Software Used

[1219] Hardware: Servers, user devices (PCs, smartphones, etc.)

[1220] Software: Generative AI models (e.g. GPT-4), Natural Language Processing libraries (e.g. spaCy, NLTK), Database Management Systems (e.g. MySQL)

[1221] System Operation

[1222] 1. The user enters a new comment

[1223] The user uses the terminal to enter a new utterance into an input field of the interactive tool. For example, the user enters, "The start date for the new project is next Monday."

[1224] 2. The server retrieves past comments and documents

[1225] The server connects to the database management system and queries the user's past comments and documents. For example, a past comment such as "The project start date is the Monday after next" is retrieved from the database.

[1226] 3. The server performs natural language processing

[1227] The server uses natural language processing libraries to analyze the retrieved past utterances and written materials, specifically tokenizing sentences, tagging parts of speech, and analyzing dependencies.

[1228] 4. The server performs a conflict check

[1229] The server uses the generative AI model to check whether the new statement is consistent with previous statements or documents. For example, the generative AI model is given the following prompt: "The user has entered a new statement. Please check whether there are any inconsistencies with previous statements or documents. New statement: 'The start date of the new project is next Monday.' Previous statement: 'The start date of the project is the Monday after next.'"

[1230] 5. The server notifies the user of the result

[1231] If a conflict is detected, the server generates a warning message for the user and displays it on the terminal, for example "Isn't the start date of your new project on Monday after next?", and suggests ways to correct the conflict.

[1232] Examples

[1233] User input: "The start date for my new project is next Monday."

[1234] Previously stated: "The project start date is the Monday after next."

[1235] In this case, the system proceeds as follows.

[1236] 1. A user uses a terminal and types, "The start date for my new project is next Monday."

[1237] 2. The server retrieves the past statement "The project start date is the Monday after next" from the database.

[1238] 3. The server uses a natural language processing library to analyze past utterances.

[1239] 4. The server uses the generative AI model to check for inconsistencies between new and previous utterances.

[1240] 5. Because an inconsistency is detected, the server displays a warning message to the user asking "Isn't the start date of your new project on Monday after next?" and suggests a correction.

[1241] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[1242] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1243] Program processing flow

[1244] Step 1: User enters new utterance

[1245] What happens: The user uses the terminal to enter a new utterance into the input field of the interactive tool.

[1246] Input: The new utterance typed by the user (e.g., "The start date for the new project is next Monday.")

[1247] Output: The user's new utterance is sent to the system.

[1248] Step 2: The server retrieves past statements and documents

[1249] Specific operation: The server connects to a database management system and queries and retrieves the user's past comments and created materials.

[1250] Input: User ID or user identification

[1251] Data processing: Search the database for past statements and created materials and extract relevant data.

[1252] Output: Data from past statements and documents (e.g. "The project start date is the Monday after next week.")

[1253] Step 3: The server performs natural language processing

[1254] Specific operation: The server uses a natural language processing library to analyze the retrieved past statements and written materials.

[1255] Input: Data of past statements and created documents

[1256] Data processing: Sentence tokenization, part-of-speech tagging, dependency analysis, etc.

[1257] Output: Structured data of analyzed past statements and documents

[1258] Step 4: Server performs inconsistency check

[1259] How it works: The server uses a generative AI model to check for inconsistencies between new statements and previous statements or created materials.

[1260] Input: New statements, parsed past statements, and structured data of created documents

[1261] Data calculation: Input a prompt sentence into the generative AI model and perform inconsistency checks.

[1262] Example prompt: "The user has entered a new statement. Please check for inconsistencies with previous statements and documentation. New statement: 'The new project start date is next Monday.' Previous statement: 'The project start date is the Monday after next.'"

[1263] Output: Result on whether there is a conflict (e.g., conflict detected)

[1264] Step 5: The server notifies the user of the result

[1265] Specific Behavior: If a conflict is detected, the server generates a warning message for the user and displays it on the terminal.

[1266] Input: Result regarding whether or not there is a conflict

[1267] Data processing: Generate warning messages (e.g. "Isn't the start date of your new project on Monday after next?")

[1268] Output: A warning message that is displayed on the user's terminal.

[1269] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[1270] (Application example 2)

[1271] Next, a description will be given of Application Example 2 of Form 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".

[1272] Conventional dialogue tools and chat tools lack the functionality to automatically check for inconsistencies with past statements and prepared documents when a user makes a new statement. This makes it difficult to maintain consistency of important information, particularly in security services, and there is a risk of incorrect information being transmitted. The present invention aims to solve these problems and provide a system that automatically checks for inconsistencies with past statements and prepared documents when a user makes a new statement, and suggests corrections as necessary.

[1273] 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.

[1274] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for using a generative AI model to suggest corrections when a contradiction is detected. This makes it possible to automatically check for inconsistencies with past comments and created materials when the user makes a new comment, and to suggest corrections as necessary.

[1275] "User" means an individual or entity that uses the System to input statements or create materials.

[1276] "Past comments" are records of comments or statements a user has previously made within the system.

[1277] "Created Materials" refers to documents and data created or uploaded by a User within the System.

[1278] "Input" refers to the data or information that a system receives for processing.

[1279] "Content to be posted" refers to the new information or comments that a user intends to make within the system.

[1280] "Inconsistency" refers to a situation in which new statements or posts are inconsistent with or in conflict with previous statements or materials.

[1281] "Suggestions" refers to suggestions or advice that the system provides to the user.

[1282] A "generative AI model" refers to an algorithm or program that uses artificial intelligence techniques to analyze data and generate appropriate suggestions or revisions.

[1283] "Present" refers to the act of the system displaying information or suggestions to the user.

[1284] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for using a generative AI model to suggest corrections when an inconsistency is detected.

[1285] When a user inputs a new comment, the server retrieves past comments and written materials from the database and analyzes them. Natural language processing technology is used for the analysis to check for inconsistencies between the past comments and written materials and the new comment. If a contradiction is detected, a generative AI model is used to generate a correction proposal and present it to the user.

[1286] Specifically, the server performs the following process. First, when a user inputs a new statement, the statement is sent to the server. The server retrieves past statements and created documents from a database and analyzes this data using natural language processing technology. For example, OpenAI's API can be used for the analysis. If the analysis detects a contradiction between the new statement and past statements or created documents, the server uses a generative AI model to generate a correction proposal. This correction proposal is presented to the user as a warning message.

[1287] The hardware used includes a server and a user's terminal. The server is a computer equipped with a high-performance processor and a large amount of memory, and the user's terminal is a general device such as a smartphone or a PC. The software used includes libraries for implementing natural language processing technology (e.g., Python's difflib library) and APIs for executing generative AI models (e.g., OpenAI's API).

[1288] As a concrete example, consider the case of checking statements regarding security policies. Assume that a past statement recorded is "All passwords must be at least eight characters long." If the user attempts to newly state "All passwords must be at least six characters long," the server analyzes this new statement and detects any inconsistencies with the past statement. If a contradiction is detected, the generative AI model is used to generate a prompt sentence like the one below, and a correction suggestion is presented.

[1289] Example prompt:

[1290] New statement: All passwords must be at least six characters long.

[1291] Previous contradictory statement: All passwords must be at least eight characters long.

[1292] Generate suggestions for fixing the inconsistencies.

[1293] In this way, when a user makes a new statement, the system automatically checks for inconsistencies with previous statements and written materials, and suggests corrections if necessary.

[1294] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1295] Step 1:

[1296] The user inputs a new comment. The input comment is sent from the user's terminal to the server. The input data is the text of the new comment.

[1297] Step 2:

[1298] The server retrieves the user's past comments and created documents from the database. The retrieved data is the text of the past comments and created documents.

[1299] Step 3:

[1300] The server uses natural language processing technology to analyze new comments and past comments and created documents. Specifically, it uses Python's difflib library to calculate the similarity between comments and documents. The input data is the text of the new comment and past comments and created documents, and the output data is the similarity score.

[1301] Step 4:

[1302] The server checks whether there are any inconsistencies between the new comment and past comments or documents based on the analysis results. If a inconsistency is detected, the information is recorded. The input data is the similarity score, and the output data is a flag indicating whether or not there is a inconsistency.

[1303] Step 5:

[1304] If a contradiction is detected, the server uses a generative AI model to generate a correction proposal. Specifically, it uses OpenAI's API to generate a prompt sentence and obtain a correction proposal. The input data is the text of past statements or created documents that contradict the new statement, and the output data is the text of the correction proposal.

[1305] Step 6:

[1306] The server presents the generated revision proposal to the user, which is displayed as a warning message on the user's terminal. The input data is the text of the revision proposal, and the output data is the warning message presented to the user.

[1307] Step 7:

[1308] The user checks the proposed revisions and revises the new utterance if necessary. The revised utterance is sent to the server again, and the same process is repeated. The input data is the text of the revised utterance, and the output data is the text of the final utterance.

[1309] In addition, an emotion engine that estimates the emotion of the user may be further combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[1310] "Example 1"

[1311] As one embodiment of the present invention, a chat tool plug-in or interactive tool combined with an emotion engine is provided. The emotion engine analyzes the user's emotional state from the user's past comments and documents created. Specifically, it analyzes the text of messages posted by the user in the past and documents created by the user, and extracts emotional expressions and keywords suggesting emotions contained therein. Then, the user's emotional state is estimated based on this information. For example, if the user has used many positive expressions such as "fun" and "happy" in the past, the user's emotional state is estimated to be "joy" or "happiness." The emotional state estimated in this way is used as reference information when suggesting whether there are any inconsistencies in the content the user will post in the future.

[1312] "Example 2"

[1313] In another embodiment of the present invention, a system is provided in which an emotion engine analyzes a user's emotional state using natural language processing technology and emotion recognition technology. Specifically, natural language processing technology is used to analyze text, and expressions and keywords indicating emotions are extracted from the text. Next, emotion recognition technology is used to analyze the extracted expressions and keywords to estimate the user's emotional state. For example, if a user uses a lot of negative expressions such as "sad" and "disappointed," the user's emotional state is estimated to be "sadness" or "disappointment." The emotional state estimated in this way is used as reference information when suggesting whether the content the user is about to post is consistent with the user's emotional state.

[1314] The process flow of each embodiment will be described below.

[1315] "Example 1"

[1316] Step 1: A user types a message into the chat tool.

[1317] Step 2: The emotion engine analyzes the user's emotional state from their past statements and documents. Specifically, it analyzes the text of past messages and documents and extracts emotional expressions and keywords that suggest emotions.

[1318] Step 3: The emotion engine estimates the user's emotional state based on the extracted information.

[1319] Step 4: Based on the estimated emotional state, we suggest inconsistencies in the content the user will post next.

[1320] "Example 2"

[1321] Step 1: A user types a message into the chat tool.

[1322] Step 2: The emotion engine uses natural language processing techniques to analyze the text and extract expressions and keywords that indicate emotions.

[1323] Step 3: The emotion engine uses emotion recognition technology to analyze the extracted expressions and keywords and infer the user's emotional state.

[1324] Step 4: Based on the estimated emotional state, we suggest whether the content of the user's future post is consistent with that emotional state.

[1325] Example 1

[1326] Next, a description will be given of Example 1 of the embodiment 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".

[1327] Conventional chat tools and conversational tools lack the ability to automatically detect and warn users when their statements contradict previous statements or written materials. In addition, suggestions do not take into account the user's emotional state, making it difficult to provide appropriate advice that is in line with the user's intentions. This can lead to a lack of consistency in communication, leading to misunderstandings and confusion.

[1328] 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.

[1329] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there is any inconsistency in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the user's comments using natural language processing technology, a means for comparing the analyzed comments with the past comments and created materials stored in the database, a means for displaying a warning message to the user when a contradiction is detected, a means for analyzing the user's emotional state from the user's past comments and created materials, and a means for making suggestions based on the analyzed emotional state. This makes it possible to automatically check whether the user's comments are inconsistent with the past comments and created materials, and to provide appropriate warnings and suggestions. In addition, by making suggestions taking the user's emotional state into consideration, more consistent communication can be achieved.

[1330] "User" refers to an individual or organization that makes statements or creates materials using chat tools or interactive tools.

[1331] A "utterance" is a text message entered by a user through a chat or interactive tool.

[1332] "Created Materials" refers to documents and data files previously created by the User, including reports, memos, presentation materials, etc.

[1333] "Input" refers to the data or information a system receives for analysis or comparison.

[1334] "Suggestion" refers to suggestions or advice that the system provides to the user.

[1335] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language, and includes tokenization, grammatical analysis, and semantic analysis.

[1336] A "database" refers to a collection of information that stores past statements and created materials and can be used for searching and comparison.

[1337] "Contradiction" refers to an inconsistency that occurs when a user's new statements do not match previous statements or materials they have created.

[1338] "Warning message" refers to a notification sent by the system to the user to inform them of a discrepancy and encourage them to correct it.

[1339] "Emotional state" refers to a user's psychological state inferred from their past statements and written materials, and may include joy, sadness, anger, etc.

[1340] "Analysis" refers to the process by which the system processes the data it receives and understands its meaning and structure.

[1341] "Comparison" refers to the process of comparing the analyzed statements with past statements and created materials to check for matches and inconsistencies.

[1342] The present invention is a system that automatically checks whether there are any inconsistencies with past statements or documents created when a user uses a chat tool or an interactive tool, and provides appropriate suggestions. A specific embodiment of this system will be described below.

[1343] System Configuration

[1344] This system is primarily composed of three elements: a server, a terminal, and a user. The server is responsible for analyzing what users say and comparing it with past comments and created materials. Terminals are devices that users use to operate chat tools and interactive tools, such as PCs and smartphones. Users are individuals or groups who use chat tools and interactive tools to make comments and create materials.

[1345] Hardware and Software Used

[1346] The server is a computer system equipped with a high-performance processor and a large amount of memory. The database uses a relational database management system (RDBMS) or a NoSQL database to store past statements and created materials. Specifically, MySQL is one such example.

[1347] The analysis uses natural language processing technology. Specifically, OpenAI's GPT-3 and Google's BERT are used as generative AI models. These models are used to tokenize the user's speech and analyze its grammar and meaning.

[1348] Data processing and calculation

[1349] When a user enters a new comment, the server receives the comment and analyzes it using natural language processing technology. The analyzed comment is compared with past comments and created materials stored in the database. This comparison is made using Elasticsearch, a full-text search engine. Based on the comparison results, it is determined whether there are any inconsistencies, and if any are detected, a warning message is displayed to the user.

[1350] In addition, an emotion engine is used to analyze the user's emotional state from their past comments and written materials, using tools such as IBM Watson's Natural Language Understanding. The results of the analysis are used to estimate the user's emotional state and are used as reference information when suggesting consistency in future posts.

[1351] Examples

[1352] For example, if a user says in a chat tool, "A new project has started," if this statement contradicts a previous statement, "The project hasn't started yet," the server will warn, "This statement contradicts your previous statement. Do you want to correct it?"

[1353] Example prompt:

[1354] The user says "A new project has started", which contradicts their previous statement "The project hasn't started yet". Show a warning message and suggest a fix.

[1355] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[1356] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1357] Step 1:

[1358] The user types a new comment into the chat tool.

[1359] Input: User says "I've started a new project."

[1360] Output: The content of the comment is sent to the chat tool.

[1361] Specific behavior: The user types text into a chat window and clicks the send button.

[1362] Step 2:

[1363] The server receives the user's comments and analyzes them using natural language processing technology.

[1364] Input: User says "I've started a new project."

[1365] Output: Tokenization and grammar analysis of the parsed utterance.

[1366] How it works: The server uses a generative AI model (e.g., GPT-3) to tokenize the utterance and perform grammar analysis.

[1367] Step 3:

[1368] The server compares the analyzed content of the comments with past comments and created materials stored in a database.

[1369] Input: Analyzed statements and past statements and written materials in the database.

[1370] Output: Comparison result (conflicts or not).

[1371] Specific behavior: The server uses a full-text search engine (e.g., Elasticsearch) to search for matches to the past utterance "The project hasn't started yet."

[1372] Step 4:

[1373] The server determines whether there is a conflict based on the comparison.

[1374] Input: The comparison result.

[1375] Output: The result of determining whether there is a contradiction.

[1376] Specific behavior: The server analyzes the comparison result and determines that "A new project has started" and "The project has not started yet" are inconsistent.

[1377] Step 5:

[1378] The server will display a warning message to the user if a discrepancy is detected.

[1379] Input: The result of the determination as to whether there is a contradiction or not.

[1380] Output: A warning message.

[1381] Specific behavior: The server displays a warning message to the user saying, "This statement contradicts a previous statement. Would you like to correct it?"

[1382] Step 6:

[1383] The server uses an emotion engine to analyze the user's emotional state from their past statements and written materials.

[1384] Input: Past statements and created materials.

[1385] Output: The parsed emotional state.

[1386] Specific operation: The server uses a sentiment analysis tool (e.g., IBM Watson) to extract positive sentiment from past statements such as "It was fun" and "I'm happy."

[1387] Step 7:

[1388] The server makes suggestions based on the analyzed emotional state.

[1389] Input: Parsed emotional state and current speech.

[1390] Output: Suggestions that take emotional state into account.

[1391] Specific behavior: The server suggests to the user, "Even though you've contradicted your previous statements, would you like to emphasize that the project is going well?"

[1392] In this way, the system analyzes the content of the user's comments, compares them with past comments and written materials to detect inconsistencies, and displays warning messages if necessary. By combining it with an emotion engine, it is possible to make suggestions that take into account the user's emotional state.

[1393] (Application example 1)

[1394] Next, a description will be given of Application Example 1 of Form 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".

[1395] In modern companies, improving the efficiency of internal communication and preventing information leaks are important issues. In particular, when communicating using chat tools, it is necessary to avoid inconsistencies with past statements and documents, to grasp the user's emotional state and detect abnormal behavior early, and to prevent the leaking of confidential information. However, it is difficult to realize these functions in a single system, and currently it is necessary to use multiple tools at the same time. This increases the burden on users and makes efficient operation difficult.

[1396] 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.

[1397] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing the user's emotional state and displaying a warning when an abnormal emotion is detected, and a means for detecting comments containing confidential information and displaying a warning. This makes it possible to avoid inconsistencies in internal communication, early detection of abnormal emotional states, and prevention of confidential information leaks.

[1398] "User" refers to any individual or entity that uses the System.

[1399] "Past comments" refers to the content of chats or messages a user has previously made.

[1400] "Created materials" refers to text data such as documents and reports created by users.

[1401] "Input" refers to data or information entered into a system.

[1402] "Contradiction" refers to a discrepancy or inconsistency that arises between past statements or prepared materials and new statements.

[1403] "Suggestions" refers to suggestions or advice that the system presents to the user.

[1404] "Emotional state" refers to a user's psychological state or mood.

[1405] "Analysis" refers to the process of analyzing data or information to extract meaning and patterns.

[1406] "Abnormal emotions" refers to negative emotions or abnormal psychological states that deviate from normal emotional states.

[1407] "Warning" refers to a message or notification that alerts the user.

[1408] "Confidential information" refers to important information that could cause harm if leaked to the outside.

[1409] "Discovery" refers to the process of finding specific conditions or patterns.

[1410] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the user's emotional state and displaying a warning if abnormal emotions are detected, and means for detecting comments that include confidential information and displaying a warning.

[1411] System Programming

[1412] This system is realized using the following hardware and software.

[1413] Hardware

[1414] Server: Manages the database and performs natural language processing and sentiment analysis.

[1415] User device: A smartphone or computer with a chat tool installed.

[1416] software

[1417] Natural Language Processing libraries: TfidfVectorizer, cosine_similarity (using Python's scikit-learn library)

[1418] Sentiment analysis library: TextBlob (Python library)

[1419] Database management system: A database for storing and managing past statements and created materials (e.g., MySQL, PostgreSQL)

[1420] Processing flow

[1421] 1. Receiving input:

[1422] The server retrieves the user's past comments and created materials from a database and receives new comments as input.

[1423] 2. Suggested contradictions:

[1424] The server uses TfidfVectorizer to vectorize new and past comments, and calculates cosine_similarity. If the similarity is low, it is determined that there is a contradiction and presents suggestions to the user.

[1425] 3. Emotional state analysis:

[1426] The server uses TextBlob to calculate the sentiment score of new utterances, and if the sentiment score is abnormal, it displays a warning to the user.

[1427] 4. Detecting Sensitive Information:

[1428] The server uses regular expressions to check whether new posts contain sensitive information, and if so, displays a warning to the user.

[1429] Examples

[1430] Suggested contradictions:

[1431] Previously stated: "Project X is progressing well."

[1432] New statement: "Progress on Project X is slow."

[1433] Result: The inconsistency is detected and the user is prompted with the suggestion, "This statement contradicts something you said previously. Would you like to fix it?"

[1434] Emotional state analysis:

[1435] Say: "I really hate this project."

[1436] Result: Negative sentiment is detected and the user is alerted with "Negative sentiment detected. Is something wrong?"

[1437] Sensitive information detection:

[1438] Say: "This is classified information."

[1439] Result: Because it contains sensitive information, the user will see a warning saying "This post contains sensitive information. Do you want to send this post?"

[1440] Examples of prompt statements

[1441] "Please create a program that checks for inconsistencies in what the user says against past statements and documents, and displays a warning if any inconsistencies are detected. Please also analyze the user's emotional state, and display a warning if abnormal emotions are detected. Furthermore, please add a function to display a warning if confidential information is included."

[1442] In this way, it is possible to avoid inconsistencies in internal communication, detect abnormal emotional states early, and prevent the leakage of confidential information.

[1443] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1444] Step 1:

[1445] The server retrieves the user's past comments and created materials from a database. This provides the past comments and created materials to the server as input. Specifically, a database query is executed to extract past comments and created materials related to the user ID.

[1446] Step 2:

[1447] The server receives the new comment entered by the user. This provides the new comment to the server as input. Specifically, the server receives an API request from the chat tool and obtains the new comment content.

[1448] Step 3:

[1449] The server vectorizes the new and past comments using TfidfVectorizer, which converts the text data into numerical vectors. Specifically, the new and past comments are input into TfidfVectorizer as a single corpus, and a TF-IDF vector is generated for each piece of text.

[1450] Step 4:

[1451] The server calculates cosine_similarity to evaluate the similarity between the new and past comments. This determines whether there is a contradiction. Specifically, it calculates the cosine similarity between the vector of the new comment and the vector of the past comment, and determines that there is a contradiction if the similarity is low.

[1452] Step 5:

[1453] If a contradiction is detected, the server presents a suggestion to the user, giving the user an opportunity to correct the contradiction. Specifically, the server displays a message through the chat tool interface saying, "This is inconsistent with your previous statement. Do you want to correct it?"

[1454] Step 6:

[1455] The server uses TextBlob to calculate the sentiment score of a new comment, which analyzes the sentiment state of the comment. Specifically, a new comment is input to TextBlob and a positive / negative sentiment score is obtained.

[1456] Step 7:

[1457] If the emotion score is abnormal, the server displays a warning to the user. This allows the user to become aware of the abnormal emotional state. Specifically, if the emotion score falls below a certain threshold, the server displays a message saying, "Negative emotion has been detected. Is there a problem?"

[1458] Step 8:

[1459] The server uses regular expressions to check whether new comments contain confidential information. This is how confidential information is detected. Specifically, the server searches new comments for keywords that indicate confidential information (e.g., "confidential" or "secret"), and if any are found, it determines that the new comment contains confidential information.

[1460] Step 9:

[1461] If confidential information is included, the server displays a warning to the user. This allows the user to prevent confidential information from being leaked. Specifically, the server displays a message saying, "Confidential information is included. Do you want to send this message?"

[1462] Example 2

[1463] Next, a second embodiment of the second form will be described. 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".

[1464] In conventional dialogue and chat tools, there is a risk that users may post content that contradicts their past statements or documents they have created. In addition, since suggestions are made without taking into account the user's emotional state, advice that does not match the user's intentions or emotions may be provided. For this reason, there is a demand for a system that provides appropriate suggestions that take into account the user's past statements and emotional state.

[1465] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1466] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether the content to be posted from now on is consistent with inconsistencies, a means for presenting the suggestions to the user, a means for analyzing the user's comments and estimating the user's emotional state, and a means for suggesting whether the content to be posted from now on is consistent with the emotional state based on the estimated emotional state. This makes it possible to make appropriate suggestions taking into account the user's past comments and emotional state.

[1467] "User" refers to an individual or organization that uses the System to create and post comments and materials.

[1468] "Comments" refers to text data that users input and post through the system.

[1469] "Created Materials" refers to documents and data that a User has previously created and stored through the System.

[1470] "Input" refers to the data or information a system receives for processing.

[1471] "Suggestion" refers to suggestions or advice that the system provides to the user.

[1472] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language.

[1473] "Emotional state" refers to the psychological state or emotion inferred from a user's statements.

[1474] "Contradiction" refers to a situation where a user's new statements are inconsistent with or contradict previous statements or materials created.

[1475] "Managed online" means that data and information are stored and managed in a manner that makes them accessible via the Internet.

[1476] "Analysis" refers to the process by which a system examines and evaluates data or information in detail.

[1477] "Inference" refers to the process by which a system predicts a state or outcome based on data and information.

[1478] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1479] The present invention is a system that receives a user's past comments and documents as input, and suggests whether the content to be posted in the future is consistent with the user's past comments and documents. Furthermore, the system has a function of analyzing the user's comments and estimating the user's emotional state, and then suggesting whether the content to be posted in the future is consistent with the user's emotional state based on the estimated emotional state.

[1480] Hardware and Software Used

[1481] The present system is implemented using the following hardware and software.

[1482] Server: Equipped with a database management system (DBMS), it stores and manages users' past comments and documents created by users. The server also provides the computing resources to execute natural language processing and emotion recognition technologies.

[1483] Terminal: The device (PC, smartphone, tablet, etc.) that a user uses to enter utterances.

[1484] Natural Language Processing libraries: Use Python's NLTK library and spaCy to tokenize and semantically analyze text.

[1485] Emotion Recognition Library: Analyzes and estimates the user’s emotional state using Python’s TextBlob and VADER Sentiment Analysis.

[1486] System Operation

[1487] 1. Check before user comments are entered

[1488] When a user enters a new comment, the server retrieves the user's past comments and created materials from the database.

[1489] The server uses natural language processing technology to analyze the content of past statements and documents, and check for any contradictions with new statements.

[1490] Example of how it works: When a user tries to type "The project is on track," the server retrieves the previous statement "The project is delayed" and detects the inconsistency.

[1491] 2. Detecting inconsistencies and displaying warning messages

[1492] If the server detects a discrepancy, it will display a warning message to the user.

[1493] Specific example of behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[1494] 3. Analyzing the user's emotional state

[1495] When a user enters a comment, the server uses natural language processing technology to analyze the text and extract expressions and keywords that indicate emotions.

[1496] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[1497] Example of specific operation: If a user enters "I'm sad that things haven't been going well at work recently," the server will extract the keyword "sad" and estimate the emotional state as "sad."

[1498] 4. Using Emotional States to Suggest

[1499] Based on the estimated emotional state, the system suggests whether the content the user will post next is consistent with that emotional state.

[1500] An example of how this works: When a user tries to type "This project is awesome," the server suggests "Do you really feel that way?"

[1501] Examples of prompt statements

[1502] Conflict check prompt

[1503] Compare the user's previous statements with their new statements and check for inconsistencies. Previous statement: "The project is delayed," New statement: "The project is on track."

[1504] Sentiment Analysis Prompt

[1505] Analyze user utterances and infer their emotional state. utterance: "I'm sad that things haven't been going well at work lately."

[1506] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1507] Step 1:

[1508] The user inputs a new comment.

[1509] Input: The user types text into an input field on the terminal.

[1510] What happens: The user types, "The project is going well."

[1511] Output: The entered text data is sent to the server.

[1512] Step 2:

[1513] The server retrieves the user's past comments and created materials from the database.

[1514] Input: A user's ID and a request for past comments.

[1515] What happens: The server runs a SQL query to retrieve the user's previous statement "The project is delayed."

[1516] Output: Data of past comments and created materials is returned to the server.

[1517] Step 3:

[1518] The server uses natural language processing technology to analyze the content of past statements and documents.

[1519] Input: Text data of past statements and created documents.

[1520] How it works: The server uses Python's NLTK library to tokenize past utterances and perform semantic analysis.

[1521] Output: Structured information of the parsed text data.

[1522] Step 4:

[1523] The server compares the new message with previous messages to check for inconsistencies.

[1524] Input: Text data of new utterances and structured information of previous utterances parsed.

[1525] What happens: The server compares the new statement "The project is on track" with the previous statement "The project is delayed" and detects the inconsistency.

[1526] Output: Information about whether there are any conflicts.

[1527] Step 5:

[1528] If the server detects a discrepancy, it displays a warning message to the user.

[1529] Input: Information about whether there is a conflict.

[1530] Specific behavior: The server displays a warning message on the user's device saying "Please check the project progress again."

[1531] Output: A warning message that is displayed on the user's terminal.

[1532] Step 6:

[1533] The server uses natural language processing technology to analyze the user's comments and extract expressions and keywords that indicate emotions.

[1534] Input: The text of the user's new utterance.

[1535] Specific operation: The server uses Python's TextBlob library to extract the keyword "sad" from the comment.

[1536] Output: Extracted sentiment expressions and keywords.

[1537] Step 7:

[1538] The server uses emotion recognition technology to analyze the extracted expressions and keywords and estimate the user's emotional state.

[1539] Input: Extracted sentiment expressions and keywords.

[1540] What happens: The server uses VADER Sentiment Analysis to estimate the user's emotional state as "sad."

[1541] Output: The estimated emotional state.

[1542] Step 8:

[1543] Based on the estimated emotional state, the server suggests whether the content the user is about to post is consistent with that emotional state.

[1544] Input: Estimated emotional state and textual data of new utterances.

[1545] Specific behavior: The server responds to the user's new statement, "This project is awesome," by suggesting, "Do you really feel that way?"

[1546] Output: The suggestion message that is displayed on the user's device.

[1547] (Application example 2)

[1548] Next, a description will be given of Application Example 2 of Form 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".

[1549] Traditional chat tools and conversational tools lack the ability to automatically check whether the content a user posts is consistent with past statements or documents. They also lack the ability to analyze the user's emotional state and provide appropriate suggestions based on that emotion. This creates a risk that users will post inconsistent information or make posts that are not in line with their emotions.

[1550] 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.

[1551] In this invention, the server includes a means for receiving the user's past comments and documents as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state. This enables the user to transmit consistent information and post content appropriate to their emotions.

[1552] "User" means any individual or entity that uses the System to post statements or create materials.

[1553] "Past statements and created materials" refers to statements made by users previously on the system, as well as documents and data they created.

[1554] "Input" refers to the data or information a system receives for processing.

[1555] A "contradiction" is when new statements or materials are inconsistent with or in conflict with previous statements or materials.

[1556] A "suggestion" is a suggestion or advice that the system provides to the user.

[1557] "Presentation" refers to the act of the system displaying information to the user.

[1558] An "emotional state" is a state that indicates a user's current feelings or mood.

[1559] "Analysis" is the act of examining data or information in detail to understand its meaning and structure.

[1560] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[1561] "Managed online" means that data and information are stored in a manner that makes them accessible via the internet.

[1562] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for analyzing the user's emotional state and suggesting the content to be posted based on the emotional state.

[1563] System Configuration

[1564] 1. Hardware:

[1565] Server: A server for storing and processing data.

[1566] Device: The smartphone or computer used by the user.

[1567] 2. Software:

[1568] Natural language processing library: Uses natural language processing technologies such as spaCy.

[1569] Sentiment Analysis Library: Uses sentiment analysis techniques such as TextBlob.

[1570] Program processing explanation

[1571] The server receives input from users' online records of past posts and written materials. It then uses natural language processing technology to compare new posts with past posts and written materials to check for inconsistencies. If a contradiction is detected, the server displays a warning message to the user and suggests correcting the inconsistent content.

[1572] In addition, the server uses emotion analysis technology to analyze the user's emotional state. Based on the analysis results, the server suggests posting content that is consistent with the user's emotional state. This enables users to post information that is consistent with their emotions and appropriate to their feelings.

[1573] Examples

[1574] For example, if a user has previously posted "Today was a very fun day," and then tries to post "Today is very sad," the server detects the inconsistency. In addition, because negative emotions are detected through emotion analysis, the server displays a warning message to the user and suggests that they review the content of their post.

[1575] Examples of prompt statements

[1576] Here is an example of a prompt to be input to the generative AI model:

[1577] A new post from a user reads: "I'm so sad today."

[1578] Previous post: "Today was so much fun."

[1579] Check whether this new post is inconsistent with previous posts and perform a sentiment analysis.

[1580] Using this prompt sentence, the generative AI model can make appropriate suggestions.

[1581] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1582] Step 1:

[1583] The server manages users' past comments and written materials online. These data are contents posted by users in the past and documents created by users, and are stored in a database. The input is the user's past comments and written materials, and the output is a list of these data.

[1584] Step 2:

[1585] The user inputs new post content into the terminal. The terminal sends this input to the server. The input is the new post content entered by the user, and the output is the new post content sent to the server.

[1586] Step 3:

[1587] The server uses natural language processing technology to analyze new posts and past comments and written materials. Specifically, it uses libraries such as spaCy to tokenize the text and perform semantic analysis. The input is the new post and past comments and written materials, and the output is the analysis results.

[1588] Step 4:

[1589] The server uses the analysis results to check whether there are any inconsistencies between the new post and past comments or documents. If a contradiction is detected, the server identifies the content and generates a warning message for the user. The input is the analysis results, and the output is the presence or absence of a contradiction and a warning message.

[1590] Step 5:

[1591] The server analyzes the emotional state of the new post using sentiment analysis techniques, specifically by calculating a sentiment score for the text using a library such as TextBlob. The input is the new post, and the output is the sentiment score.

[1592] Step 6:

[1593] The server estimates the user's emotional state based on the emotion score, and suggests posting content that is consistent with the emotional state. The input is the emotion score, and the output is the suggested content.

[1594] Step 7:

[1595] The server generates warning messages and suggestions and sends them to the user's terminal, which displays them to the user. The input is the warning messages and suggestions, and the output is the information displayed to the user.

[1596] Step 8:

[1597] The user checks the warning messages and suggestions from the server and modifies the new post content as necessary. The input is the information displayed to the user, and the output is the modified post content.

[1598] 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 voice indicating a user input for 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.

[1599] 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> ) and other generation AIs. The data generation model 58 is obtained by making a neural network perform deep learning. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating a voice, text data indicating a text, and image data indicating an image is input. The data generation model 58 performs inference on the input inference data according to 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.

[1600] Another example of generative AI is Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) are mentioned.

[1601] 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.

[1602] [Fourth embodiment]

[1603] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1604] 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.

[1605] 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 wide area network (WAN) and / or a local area network (LAN).

[1606] 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. In addition, the microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1607] 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 the voice according to instructions from the processor 46.

[1608] Camera 42 is a small digital camera equipped with an optical system including a lens, an aperture, and a 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 (e.g., an imaging range defined by an angle of view equivalent to the width of the field of vision of an average healthy person).

[1609] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for transmitting and receiving various types of information between the processor 46 and the processor 28 via the network 54. The transmission and reception of various types of information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is performed in a secure state.

[1610] The control target 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, legs, etc. The posture and behavior of the robot 414 are controlled by controlling the motors of the arms, hands, legs, etc. 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.

[1611] Fig. 8 shows an example of 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.

[1612] 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.

[1613] 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.

[1614] In the robot 414, 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.

[1615] Next, the specific processing by the specific processing unit 290 of the data processing device 12 will be described.

[1616] "Example 1"

[1617] As one embodiment of the present invention, the present invention is implemented as a plug-in for a chat tool. When a user makes a comment on the chat tool, the plug-in compares the comment content with past comments and created materials, and automatically checks for inconsistencies. Specifically, when a user inputs a new comment, the plug-in analyzes the comment content and compares it with past comments and created materials. If an inconsistency is detected, the plug-in displays a warning message to the user and suggests that the inconsistent content be corrected.

[1618] "Example 2"

[1619] Another embodiment of the present invention is implemented as an interactive tool that automatically checks for inconsistencies before a user inputs a new statement by comparing it with past statements and written materials. Specifically, before a user inputs a new statement, the tool analyzes past statements and written materials and checks for inconsistencies with the contents of the new statement. If an inconsistency is detected, the tool displays a warning message to the user and suggests that the user correct the inconsistent contents.

[1620] The process flow of each embodiment will be described below.

[1621] "Example 1"

[1622] Step 1: A user types a new message into the chat tool.

[1623] Step 2: The plugin analyzes the comment and compares it with previous comments and created materials.

[1624] Step 3: If any conflicts are detected, the plugin will display a warning message and suggest you fix the conflicts.

[1625] "Example 2"

[1626] Step 1: Before the user enters a new comment, the interactive tool analyzes past comments and written materials.

[1627] Step 2: Check whether the interactive tool is consistent with the content.

[1628] Step 3: If any conflicts are detected, the interactive tool will display a warning message and suggest ways to fix the conflicts.

[1629] Example 1

[1630] Next, a description will be given of Example 1 of the embodiment 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".

[1631] Conventional chat tools have the problem that users may post content that contradicts past statements or documents created, resulting in inconsistent information. This problem can cause confusion in project progress and communication. Therefore, a system that supports users to avoid contradicting past statements or documents created is required.

[1632] 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.

[1633] In this invention, the server includes means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, means for analyzing the content of a new comment when the user inputs the comment, means for comparing the analyzed content of the comment with the past comments and created materials, and means for displaying a warning message to the user when a contradiction is detected and suggesting that the contradictory content be corrected. This makes it possible to maintain consistency in the content of user comments and prevent confusion in the progress of projects and communication.

[1634] "User" refers to an individual or organization that uses chat tools to make statements or create materials.

[1635] "Past comments" refers to records of comments a user previously made on the chat tool.

[1636] "Created Materials" refers to documents and data created by Users on chat tools or related platforms.

[1637] "Input" refers to the data or information a system receives for processing.

[1638] "Suggestion" refers to suggestions or advice that the system provides to the user.

[1639] "Means" refers to a method or apparatus for achieving a specific function or purpose.

[1640] "Chat Tool" refers to software or platform that enables users to exchange messages in real time.

[1641] A "plugin" refers to an extension or module that is added to a chat tool.

[1642] An "interactive tool" refers to software or a platform that allows users and systems to communicate with each other.

[1643] "Managed online" means that data or information is stored in a manner that makes it accessible via the internet.

[1644] "Natural language processing technology" refers to technology that enables computers to understand and process human language.

[1645] "Analysis" refers to the process by which a system understands and extracts meaning from input data or information.

[1646] "Comparison" refers to the process in which the system compares new statements with past statements and created materials to detect differences.

[1647] "Warning message" refers to a message that the system displays to alert the user.

[1648] "Correction" refers to the act of a user changing statements or materials to resolve a contradiction.

[1649] This invention is a system that automatically checks for inconsistencies when a user makes a comment on a chat tool by comparing the comment with past comments and created documents. This system is implemented as a plug-in for the chat tool.

[1650] Generating a Program

[1651] The server generates a program implemented as a plug-in for the chat tool. This program has a function of automatically checking for inconsistencies by comparing the content of a user's comment with past comments and created documents when the user comments on the chat tool.

[1652] Program Processing

[1653] When a user types a new comment into the chat tool, the device sends the comment to the plugin, which then uses natural language processing (NLP) techniques to analyze the comment. This analysis can be done using a generative AI model, such as OpenAI's GPT-3.

[1654] The analyzed content of the statement is compared with past statements and documents stored on the server. This comparison process uses MySQL or PostgreSQL as a database management system (DBMS). The server compares the new statement with past statements and documents to check for any inconsistencies.

[1655] If a conflict is detected, the server sends that information back to the device, which then displays a warning message to the user with suggestions for correcting the conflict. The warning message includes the specific conflict and the suggested correction.

[1656] Examples

[1657] For example, if a user says, "The project is due next Friday," the plugin checks past statements and documentation. If a user has previously said, "The project is due this Friday," a contradiction is detected.

[1658] In this case, the user will see a warning message like this:

[1659] "There is a contradiction with your previous statement. Previously you said, 'The project deadline is this Friday.' Would you like to correct that?"

[1660] Examples of prompt statements

[1661] Examples of prompts to input to a generative AI model include:

[1662] "A user has recently stated, 'The project deadline is next Friday.' Please check for inconsistencies against previous statements and documentation."

[1663] In this way, the system helps users maintain consistency in what they say.

[1664] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1665] Step 1:

[1666] A user types a comment into a chat tool.

[1667] The user types a new statement into the chat tool's input field, for example, "The project is due next Friday," and the statement is sent to the next processing step.

[1668] Step 2:

[1669] The device sends the speech content to the plugin.

[1670] The terminal sends the user's input to the plugin. This transmission is done through the chat tool's internal API. The input is the user's input, and the output is the data sent to the plugin.

[1671] Step 3:

[1672] The plugin analyzes the content of the comment.

[1673] The plugin analyzes the received utterances using natural language processing (NLP) techniques. Specifically, it uses a generative AI model (e.g., OpenAI's GPT-3) to understand the meaning of the utterances. The input is the user's utterance, and the output is the analyzed utterance data.

[1674] Step 4:

[1675] The server compares the content of the statement with past statements and created materials.

[1676] The server compares the analyzed comment content with past comments and created documents stored in a database. MySQL or PostgreSQL is used as a database management system (DBMS) for this comparison process. The input is the analyzed comment content and past data, and the output is the comparison result.

[1677] Step 5:

[1678] The server detects the discrepancy.

[1679] The server compares new statements with past statements and documents to check for inconsistencies. For example, if someone previously said, "The project deadline is this Friday," a contradiction is detected. The input is the comparison result, and the output is information about whether there is a contradiction.

[1680] Step 6:

[1681] The server sends the conflicting information back to the terminal.

[1682] The server returns the detected conflict information to the terminal, including the specific conflicts and suggested corrections. The input is the conflict information, and the output is the data transmission to the terminal.

[1683] Step 7:

[1684] The device will display a warning message to the user.

[1685] The terminal displays a warning message to the user, for example, "There is a contradiction with your previous statement. You previously said, 'The project deadline is this Friday.' Would you like to correct it?" The input is the contradiction information from the server, and the output is the warning message to the user.

[1686] Step 8:

[1687] The user edits what was said.

[1688] The user checks the warning message and modifies the comment if necessary. The modified comment is sent to the plugin again, where the same check is performed. The input is the warning message, and the output is the modified comment.

[1689] (Application example 1)

[1690] Next, a description will be given of Application Example 1 of Form 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".

[1691] Work instructions and reports within the factory may contain information that contradicts previous statements or documents, which can lead to reduced work efficiency and safety. In addition, there is no way to immediately respond to contradictions, which can lead to problems becoming more serious.

[1692] 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. In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, a means for analyzing work instructions and report contents in the factory in real time and checking for inconsistencies by comparing them with past instructions and report contents, and a means for generating a warning message and presenting it to the user when an inconsistency is detected. This maintains the consistency of work instructions and report contents in the factory, enabling efficient and safe work.

[1693] "User" means an individual or organization that uses the System.

[1694] "Past comments" are records of comments a user has made in the past.

[1695] "Created Materials" means documents and data records created by a User.

[1696] "Input" refers to data or information that is entered into a system.

[1697] "Content to be posted" refers to the new information or instructions that the user wishes to state.

[1698] "Contradiction" refers to a situation in which new statements are inconsistent with or in conflict with previous statements or documents.

[1699] "Suggest" refers to the system making suggestions or advice to the user.

[1700] "Present" refers to the system displaying information to the user.

[1701] "Factory" refers to the facility or location where manufacturing operations take place.

[1702] "Work instructions" are information that instructs the specific work content and procedures to be performed in a factory.

[1703] "Report content" refers to information about the progress and results of work.

[1704] "Real-time" refers to processing and display occurring immediately, without delay.

[1705] "Analysis" refers to the detailed examination of data or information to find meaning and patterns.

[1706] A "warning message" is a notification or alert intended to alert the user.

[1707] To implement this invention, a specific program must be installed on the robots and terminals used in the factory. This program has the function of receiving the user's past comments and documents as input, and suggesting whether there are any inconsistencies in the content to be posted in the future. It also includes a function of analyzing the work instructions and report contents in the factory in real time, and checking for inconsistencies by comparing them with past instructions and reports.

[1708] The server uses the following hardware and software:

[1709] Hardware: Factory robots

[1710] Software: Python, JSON files, natural language processing technology

[1711] The server first reads past comments and created documents from a JSON file. Then, when a new comment is entered, it checks for inconsistencies by comparing it with past comments and created documents. If a contradiction is detected, a warning message is generated and displayed to the user. If there is no contradiction, the new comment is added to the list of past comments and saved in a JSON file.

[1712] As a concrete example, consider instructions and reports for machine maintenance in a factory. If a past statement such as "Maintenance of machine A has started" is recorded, and a new statement such as "Maintenance of machine A has been completed" is entered, the system will compare these statements to check for inconsistencies.

[1713] Here are some example prompts using a generative AI model:

[1714] Previously stated: "Maintenance has begun on Machine A."

[1715] New statement: "Maintenance on Machine A has been completed."

[1716] Check for inconsistencies.

[1717] In this way, consistency in work instructions and reports within the factory can be maintained, enabling efficient and safe work.

[1718] The flow of the specific process in the application example 1 will be described with reference to FIG.

[1719] Step 1:

[1720] The server reads past comments and created documents from a JSON file. Specifically, it opens a JSON file at a specified path from the file system and converts its contents into a data structure (list or dictionary). The input is the path to the JSON file, and the output is the data structure of past comments and created documents.

[1721] Step 2:

[1722] The user inputs a new utterance. The terminal receives the input from the user and sends the contents to the server. The input is the user's new utterance, and the output is the utterance data to be sent to the server.

[1723] Step 3:

[1724] The server compares the new statement with past statements and documents to check for inconsistencies. Specifically, it loops through the data structures of past statements and documents and compares them with the new statement. The input is the new statement and the data structures of past statements and documents, and the output is the presence or absence of inconsistencies and a list of contradictory statements.

[1725] Step 4:

[1726] The server generates a warning message when a contradiction is detected. Specifically, the server creates a warning message based on the list of contradictory statements and presents it to the user. The input is the list of contradictory statements, and the output is the warning message.

[1727] Step 5:

[1728] If no contradictions are detected, the server adds the new comment to the list of past comments and saves it in a JSON file. Specifically, the server adds the new comment to a data structure, converts the data structure to JSON format, and writes it to a file. The input is the new comment and the data structure of past comments and created materials, and the output is the updated JSON file.

[1729] Step 6:

[1730] The server notifies the user that the new comment has been added successfully. Specifically, it sends a message to the user's terminal to inform them that the comment has been added. The input is the result of adding the new comment, and the output is a notification message to the user.

[1731] Example 2

[1732] Next, a second embodiment of the second embodiment will be described. 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".

[1733] Conventional chat tools and dialogue tools lack the functionality to automatically check for inconsistencies with past comments or documents when a user makes a new comment. This means that users must manually check past comments and documents, which is inefficient and increases the risk of sending incorrect information. The present invention aims to solve these problems and enable users to send information efficiently and accurately.

[1734] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1735] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the past comments and created materials using natural language processing technology, a means for checking whether there are any inconsistencies between a new comment and the past comments and created materials using a generative AI model, and a means for displaying a warning message to the user and suggesting corrections when a inconsistency is detected. This enables the user to automatically check for inconsistencies with past comments and created materials when making a new comment, and to transmit information efficiently and accurately.

[1736] "User" means an individual or organization that uses the system to make statements or create materials.

[1737] "Past comments" are records of comments or statements a user has previously made through the system.

[1738] "Created Materials" means documents and data records created by Users through the System.

[1739] "Input" refers to data or information entered into a system.

[1740] "Suggestion" means that the system makes suggestions or advice to the user.

[1741] "Natural language processing technology" is a technology that enables computers to understand and analyze human language.

[1742] A "generative AI model" is a model that uses artificial intelligence to generate and analyze data.

[1743] A "contradiction" is a state in which there is a discrepancy or inconsistency between a new statement and previous statements or documents.

[1744] A "warning message" is a message that the system displays to alert the user.

[1745] "Suggesting corrections" means that the system suggests to the user to correct the content of their statements or materials.

[1746] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1747] The present invention is a system that automatically checks for inconsistencies with past statements and created materials when a user makes a new statement, enabling the user to disseminate information efficiently and accurately. The system includes a server, a user terminal, and a generative AI model.

[1748] Hardware and Software Used

[1749] Hardware: Servers, user devices (PCs, smartphones, etc.)

[1750] Software: Generative AI models (e.g. GPT-4), Natural Language Processing libraries (e.g. spaCy, NLTK), Database Management Systems (e.g. MySQL)

[1751] System Operation

[1752] 1. The user enters a new comment

[1753] The user uses the terminal to enter a new utterance into an input field of the interactive tool. For example, the user enters, "The start date for the new project is next Monday."

[1754] 2. The server retrieves past comments and documents

[1755] The server connects to the database management system and queries the user's past comments and documents. For example, a past comment such as "The project start date is the Monday after next" is retrieved from the database.

[1756] 3. The server performs natural language processing

[1757] The server uses natural language processing libraries to analyze the retrieved past utterances and written materials, specifically tokenizing sentences, tagging parts of speech, and analyzing dependencies.

[1758] 4. The server performs a conflict check

[1759] The server uses the generative AI model to check whether the new statement is consistent with previous statements or documents. For example, the generative AI model is given the following prompt: "The user has entered a new statement. Please check whether there are any inconsistencies with previous statements or documents. New statement: 'The start date of the new project is next Monday.' Previous statement: 'The start date of the project is the Monday after next.'"

[1760] 5. The server notifies the user of the result

[1761] If a conflict is detected, the server generates a warning message for the user and displays it on the terminal, for example "Isn't the start date of your new project on Monday after next?", and suggests ways to correct the conflict.

[1762] Examples

[1763] User input: "The start date for my new project is next Monday."

[1764] Previously stated: "The project start date is the Monday after next."

[1765] In this case, the system proceeds as follows.

[1766] 1. A user uses a terminal and types, "The start date for my new project is next Monday."

[1767] 2. The server retrieves the past statement "The project start date is the Monday after next" from the database.

[1768] 3. The server uses a natural language processing library to analyze past utterances.

[1769] 4. The server uses the generative AI model to check for inconsistencies between new and previous utterances.

[1770] 5. Because an inconsistency is detected, the server displays a warning message to the user asking "Isn't the start date of your new project on Monday after next?" and suggests a correction.

[1771] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[1772] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1773] Program processing flow

[1774] Step 1: User enters new utterance

[1775] What happens: The user uses the terminal to enter a new utterance into the input field of the interactive tool.

[1776] Input: The new utterance typed by the user (e.g., "The start date for the new project is next Monday.")

[1777] Output: The user's new utterance is sent to the system.

[1778] Step 2: The server retrieves past statements and documents

[1779] Specific operation: The server connects to a database management system and queries and retrieves the user's past comments and created materials.

[1780] Input: User ID or user identification

[1781] Data processing: Search the database for past statements and created materials and extract relevant data.

[1782] Output: Data from past statements and documents (e.g. "The project start date is the Monday after next week.")

[1783] Step 3: The server performs natural language processing

[1784] Specific operation: The server uses a natural language processing library to analyze the retrieved past statements and written materials.

[1785] Input: Data of past statements and created documents

[1786] Data processing: Sentence tokenization, part-of-speech tagging, dependency analysis, etc.

[1787] Output: Structured data of analyzed past statements and documents

[1788] Step 4: Server performs inconsistency check

[1789] How it works: The server uses a generative AI model to check for inconsistencies between new statements and previous statements or created materials.

[1790] Input: New statements, parsed past statements, and structured data of created documents

[1791] Data calculation: Input a prompt sentence into the generative AI model and perform inconsistency checks.

[1792] Example prompt: "The user has entered a new statement. Please check for inconsistencies with previous statements and documentation. New statement: 'The new project start date is next Monday.' Previous statement: 'The project start date is the Monday after next.'"

[1793] Output: Result on whether there is a conflict (e.g., conflict detected)

[1794] Step 5: The server notifies the user of the result

[1795] Specific Behavior: If a conflict is detected, the server generates a warning message for the user and displays it on the terminal.

[1796] Input: Result regarding whether or not there is a conflict

[1797] Data processing: Generate warning messages (e.g. "Isn't the start date of your new project on Monday after next?")

[1798] Output: A warning message that is displayed on the user's terminal.

[1799] In this way, the system automatically checks for inconsistencies between a user's new comments and past comments or written materials, and displays warnings to the user if necessary.

[1800] (Application example 2)

[1801] Next, a description will be given of Application Example 2 of Form 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".

[1802] Conventional dialogue tools and chat tools lack the functionality to automatically check for inconsistencies with past statements and prepared documents when a user makes a new statement. This makes it difficult to maintain consistency of important information, particularly in security services, and there is a risk of incorrect information being transmitted. The present invention aims to solve these problems and provide a system that automatically checks for inconsistencies with past statements and prepared documents when a user makes a new statement, and suggests corrections as necessary.

[1803] 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.

[1804] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there are any inconsistencies in the content to be posted in the future, a means for presenting the suggestions to the user, and a means for using a generative AI model to suggest corrections when a contradiction is detected. This makes it possible to automatically check for inconsistencies with past comments and created materials when the user makes a new comment, and to suggest corrections as necessary.

[1805] "User" means an individual or entity that uses the System to input statements or create materials.

[1806] "Past comments" are records of comments or statements a user has previously made within the system.

[1807] "Created Materials" refers to documents and data created or uploaded by a User within the System.

[1808] "Input" refers to the data or information that a system receives for processing.

[1809] "Content to be posted" refers to the new information or comments that a user intends to make within the system.

[1810] "Inconsistency" refers to a situation in which new statements or posts are inconsistent with or in conflict with previous statements or materials.

[1811] "Suggestions" refers to suggestions or advice that the system provides to the user.

[1812] A "generative AI model" refers to an algorithm or program that uses artificial intelligence techniques to analyze data and generate appropriate suggestions or revisions.

[1813] "Present" refers to the act of the system displaying information or suggestions to the user.

[1814] A system for implementing this invention includes means for receiving a user's past comments and created materials as input and suggesting whether there are any inconsistencies in the content to be posted in the future, means for presenting the suggestions to the user, and means for using a generative AI model to suggest corrections when an inconsistency is detected.

[1815] When a user inputs a new comment, the server retrieves past comments and written materials from the database and analyzes them. Natural language processing technology is used for the analysis to check for inconsistencies between the past comments and written materials and the new comment. If a contradiction is detected, a generative AI model is used to generate a correction proposal and present it to the user.

[1816] Specifically, the server performs the following process. First, when a user inputs a new statement, the statement is sent to the server. The server retrieves past statements and created documents from a database and analyzes this data using natural language processing technology. For example, OpenAI's API can be used for the analysis. If the analysis detects a contradiction between the new statement and past statements or created documents, the server uses a generative AI model to generate a correction proposal. This correction proposal is presented to the user as a warning message.

[1817] The hardware used includes a server and a user's terminal. The server is a computer equipped with a high-performance processor and a large amount of memory, and the user's terminal is a general device such as a smartphone or a PC. The software used includes libraries for implementing natural language processing technology (e.g., Python's difflib library) and APIs for executing generative AI models (e.g., OpenAI's API).

[1818] As a concrete example, consider the case of checking statements regarding security policies. Assume that a past statement recorded is "All passwords must be at least eight characters long." If the user attempts to newly state "All passwords must be at least six characters long," the server analyzes this new statement and detects any inconsistencies with the past statement. If a contradiction is detected, the generative AI model is used to generate a prompt sentence like the one below, and a correction suggestion is presented.

[1819] Example prompt:

[1820] New statement: All passwords must be at least six characters long.

[1821] Contradictory previous statement: All passwords must be at least eight characters long.

[1822] Generate suggestions for fixing the inconsistencies.

[1823] In this way, when a user makes a new statement, the system automatically checks for inconsistencies with previous statements and documents, and suggests corrections if necessary.

[1824] The flow of the specific process in the application example 2 will be described with reference to FIG.

[1825] Step 1:

[1826] The user inputs a new comment. The input comment is sent from the user's terminal to the server. The input data is the text of the new comment.

[1827] Step 2:

[1828] The server retrieves the user's past comments and created documents from the database. The retrieved data is the text of the past comments and created documents.

[1829] Step 3:

[1830] The server uses natural language processing technology to analyze new comments and past comments and created documents. Specifically, it uses Python's difflib library to calculate the similarity between comments and documents. The input data is the text of the new comment and past comments and created documents, and the output data is the similarity score.

[1831] Step 4:

[1832] The server checks whether there are any inconsistencies between the new comment and past comments or documents based on the analysis results. If a inconsistency is detected, the information is recorded. The input data is the similarity score, and the output data is a flag indicating whether or not there is a inconsistency.

[1833] Step 5:

[1834] If a contradiction is detected, the server uses a generative AI model to generate a correction proposal. Specifically, it uses OpenAI's API to generate a prompt sentence and obtain a correction proposal. The input data is the text of past statements or created documents that contradict the new statement, and the output data is the text of the correction proposal.

[1835] Step 6:

[1836] The server presents the generated revision proposal to the user, which is displayed as a warning message on the user's terminal. The input data is the text of the revision proposal, and the output data is the warning message presented to the user.

[1837] Step 7:

[1838] The user checks the proposed revisions and revises the new utterance if necessary. The revised utterance is sent to the server again, and the same process is repeated. The input data is the text of the revised utterance, and the output data is the text of the final utterance.

[1839] In addition, an emotion engine that estimates the emotion of the user may be further combined. That is, the identification processing unit 290 may estimate the emotion of the user using the emotion identification model 59, and perform identification processing using the emotion of the user.

[1840] "Example 1"

[1841] As one embodiment of the present invention, a chat tool plug-in or interactive tool combined with an emotion engine is provided. The emotion engine analyzes the user's emotional state from the user's past comments and documents created. Specifically, it analyzes the text of messages posted by the user in the past and documents created by the user, and extracts emotional expressions and keywords suggesting emotions contained therein. Then, the user's emotional state is estimated based on this information. For example, if the user has used many positive expressions such as "fun" and "happy" in the past, the user's emotional state is estimated to be "joy" or "happiness." The emotional state estimated in this way is used as reference information when suggesting whether there are any inconsistencies in the content the user will post in the future.

[1842] "Example 2"

[1843] In another embodiment of the present invention, a system is provided in which an emotion engine analyzes a user's emotional state using natural language processing technology and emotion recognition technology. Specifically, natural language processing technology is used to analyze text, and expressions and keywords indicating emotions are extracted from the text. Next, emotion recognition technology is used to analyze the extracted expressions and keywords to estimate the user's emotional state. For example, if a user uses a lot of negative expressions such as "sad" and "disappointed," the user's emotional state is estimated to be "sadness" or "disappointment." The emotional state estimated in this way is used as reference information when suggesting whether the content the user is about to post is consistent with the user's emotional state.

[1844] The process flow of each embodiment will be described below.

[1845] "Example 1"

[1846] Step 1: User types a message into the chat tool.

[1847] Step 2: The emotion engine analyzes the user's emotional state from their past statements and documents. Specifically, it analyzes the text of past messages and documents and extracts emotional expressions and keywords that suggest emotions.

[1848] Step 3: The emotion engine estimates the user's emotional state based on the extracted information.

[1849] Step 4: Based on the estimated emotional state, we suggest inconsistencies in the content the user will post next.

[1850] "Example 2"

[1851] Step 1: User types a message into the chat tool.

[1852] Step 2: The emotion engine uses natural language processing techniques to analyze the text and extract expressions and keywords that indicate emotions.

[1853] Step 3: The emotion engine uses emotion recognition technology to analyze the extracted expressions and keywords and infer the user's emotional state.

[1854] Step 4: Based on the estimated emotional state, we suggest whether the content of the user's future post is consistent with that emotional state.

[1855] Example 1

[1856] Next, a description will be given of Example 1 of the embodiment 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".

[1857] Conventional chat tools and conversational tools lack the ability to automatically detect and warn users when their statements contradict previous statements or written materials. In addition, suggestions do not take into account the user's emotional state, making it difficult to provide appropriate advice that is in line with the user's intentions. This can lead to a lack of consistency in communication, leading to misunderstandings and confusion.

[1858] 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.

[1859] In this invention, the server includes a means for receiving the user's past comments and created materials as input, suggesting whether there is any inconsistency in the content to be posted, a means for presenting the suggestions to the user, a means for analyzing the user's comments using natural language processing technology, a means for comparing the analyzed comments with the past comments and created materials stored in the database, a means for displaying a warning message to the user when a contradiction is detected, a means for analyzing the user's emotional state from the user's past comments and created materials, and a means for making suggestions based on the analyzed emotional state. This makes it possible to automatically check whether the user's comments are inconsistent with the past comments and created materials, and to provide appropriate warnings and suggestions. In addition, by making suggestions taking the user's emotional state into consideration, more consistent communication can be achieved.

[1860] "User" refers to an individual or organization that makes statements or creates materials using chat tools or interactive tools.

[1861] A "utterance" is a text message entered by a user through a chat or interactive tool.

[1862] "Created Materials" refers to documents and data files previously created by the User, including reports, memos, presentation materials, etc.

[1863] "Input" refers to the data or information a system receives for analysis or comparison.

[1864] "Suggestion" refers to suggestions or advice that the system provides to the user.

[1865] "Natural language processing technology" refers to technology that enables computers to understand, analyze, and generate human language, and includes tokenization, grammatical analysis, and semantic analysis.

[1866] A "database" refers to a collection of information that stores past statements and created materials and can be used for searching and comparison.

[1867] "Contradiction" refers to an inconsistency that occurs when a user's new statements do not match previous statements or materials they have created.

[1868] "Warning message" refers to a notification sent by the system to the user to inform them of a discrepancy and encourage them to correct it.

[1869] "Emotional state" refers to a user's psychological state inferred from their past statements and written materials, and may include joy, sadness, anger, etc.

[1870] "Analysis" refers to the process by which the system processes the data it receives and understands its meaning and structure.

[1871] "Comparison" refers to the process of comparing the analyzed statements with past statements and created materials to check for matches and inconsistencies.

[1872] The present invention is a system that automatically checks whether there are any inconsistencies with past statements or documents created when a user uses a chat tool or an interactive tool, and provides appropriate suggestions. A specific embodiment of this system will be described below.

[1873] System Configuration

[1874] This system is primarily composed of three elements: a server, a terminal, and a user. The server is responsible for analyzing what users say and comparing it with past comments and created materials. Terminals are devices that users use to operate chat tools and interactive tools, such as PCs and smartphones. Users are individuals or groups who use chat tools and interactive tools to make comments and create materials.

[1875] Hardware and Software Used

[1876] The server is a computer system equipped with a high-performance processor and a large amount of memory. The database uses a relational database management system (RDBMS) or a NoSQL database to store past statements and created materials. Specifically, MySQL is one such example.

[1877] The analysis uses natural language processing technology. Specifically, OpenAI's GPT-3 and Google's BERT are used as generative AI models. These models are used to tokenize the user's speech and analyze its grammar and meaning.

[1878] Data processing and calculation

[1879] When a user enters a new comment, the server receives the comment and analyzes it using natural language processing technology. The analyzed comment is compared with past comments and created materials stored in the database. This comparison is made using Elasticsearch, a full-text search engine. Based on the comparison results, it is determined whether there are any inconsistencies, and if any are detected, a warning message is displayed to the user.

[1880] In addition, an emotion engine is used to analyze the user's emotional state from their past comments and written materials, using tools such as IBM Watson's Natural Language Understanding. The...

Claims

1. A means for analyzing the content of a statement input by a user; A means for comparing the analyzed comment content with the user's past comment content and created materials; a means for suggesting, when a content that contradicts the past comment content and the created materials is detected in the input comment content, to correct the contradictory content; A system including:

2. The suggesting means suggests the revised utterance content of the user based on the emotional state of the user. The system of claim 1 .

3. The suggesting means suggests the content of the user's utterance using a generative AI model based on a prompt sentence corresponding to the emotional state of the user. The system of claim 2.

Images