System

Abstract

A system is provided.SOLUTION: A system comprising: means for receiving input text; processing means including a generative AI model for analyzing the input text and detecting inappropriate or misleading expressions; means for generating proposed modifications based on problems detected by the processing means; means for presenting the generated proposed modifications; means for confirming and applying the proposed modifications; and means for posting the confirmed and applied text to various communication platforms.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] This invention relates to a system that automatically detects inappropriate or potentially misleading expressions in text and images posted by social networking site users and suggests corrections. This system aims to reduce the possibility of users unintentionally starting a firestorm or causing misunderstandings, thereby realizing safer and smoother communication. Furthermore, these types of problems are increasing with the spread of social networking sites, and appropriate countermeasures are required, so solving this issue is extremely important. [Means for solving the problem]

[0005] To solve the above problems, the present invention provides the following means. First, the system includes a means for receiving a text entered by a user. The received text is analyzed using a generative AI model and checked by a processing means for detecting inappropriate or potentially misleading expressions. Next, the system includes a means for generating suggested revisions based on the detected problems. The suggested revisions are presented to the user, who can confirm and apply them. The system also includes a means for posting the confirmed and applied text to various communication platforms. The system also includes a processing means for similarly analyzing image data to detect inappropriate or misleading content. In addition, the system also includes a means for, if multiple suggested revisions are generated, presenting the multiple suggested revisions to the user and allowing the user to select the most appropriate one. In this way, an environment in which users can use SNS with peace of mind is provided.

[0006] The "means for receiving input text" is a function for importing text data created and input by the user into the system.

[0007] A "generative AI model" is an artificial intelligence model used to analyze text and image data and detect inappropriate or potentially misleading content.

[0008] "Means for analyzing and detecting inappropriate or potentially misleading expressions" refers to a function that uses a generative AI model to analyze input data and identify inappropriate or potentially misleading expressions.

[0009] The "means for generating correction suggestions" is a function that generates appropriate text and image correction suggestions based on problems detected by the analysis.

[0010] The "means for presenting proposed modifications" is a function that displays the generated proposed modifications to the user and prompts the user to confirm and apply them.

[0011] "Means to confirm and apply" refers to the function that allows the user to confirm the proposed revisions, select the revision, and reflect it in the actual post content.

[0012] "Means of posting to various communication platforms" refers to the function of posting the final corrected and confirmed text and images to various SNS and other communication platforms used by the user.

[0013] The "processing means for analyzing image data" is a function that analyzes an image input by a user and determines whether the image contains inappropriate language or misleading content.

[0014] The "means for presenting multiple correction suggestions" is a function that generates multiple correction candidates for a detected problem and presents them to the user as options. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9]1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0016] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0019] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0020] 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.

[0021] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] 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.

[0028] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0033] 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.

[0034] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0035] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0036] The present invention provides a system that detects inappropriate expressions or potentially misleading parts of social media posts entered by users in real time and suggests corrections. Specific embodiments and their operation are described below.

[0037] System configuration

[0038] The system includes the following major components:

[0039] 1. User Device

[0040] It provides an interface for users to input content for posting on social media.

[0041] The input text and image data are sent to the server.

[0042] Receives suggested revisions from the server and presents them to the user.

[0043] The user is then prompted to confirm and apply the proposed changes.

[0044] 2. Server

[0045] It hosts generative AI models and analyzes data received from user devices.

[0046] Detects inappropriate or potentially misleading language and generates suggested corrections.

[0047] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0048] How it works

[0049] User terminal operation

[0050] 1. Input

[0051] The user inputs the content of the SNS post into the interface of the user device. For example, the user might input "Hello, I love dogs!"

[0052] 2. Data Transmission

[0053] Once the input is complete, the user terminal sends this text data to the server. If an image is included, it is also sent.

[0054] 3. Receiving and Displaying Proposed Amendments

[0055] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0056] Server Operation

[0057] 1. Data Analysis

[0058] It analyzes incoming text and image data and uses generative AI models to detect inappropriate or potentially misleading content, suggesting, for example, changing "hello" to "hello" or "dog" to "dog."

[0059] 2. Generate proposed fixes

[0060] Based on the detected problems, a correction suggestion is generated, which is a correction to an appropriate and safe expression while preserving the user's original intent as much as possible.

[0061] 3. Submitting amendments

[0062] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0063] 4. Reconfirmation after application of the proposed amendments

[0064] After the user confirms and applies the proposed edits, they can review the changes and post the finalized text to various communication platforms. For example, the revised text "Hello, I love dogs!" can be safely posted to social media.

[0065] Specific examples

[0066] Example 1: Correcting typos and omissions

[0067] User: Type "Hi, I love dogs!"

[0068] Server: Generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[0069] User: Check the proposed changes and press the "Apply" button.

[0070] Server: Posts the modified text to various communication platforms.

[0071] Example 2: Correcting provocative language

[0072] User: Type "Your opinion is completely meaningless!"

[0073] Server: Generates a correction suggestion saying "I have doubts about your opinion" and sends it to the user's terminal.

[0074] User: Check the proposed changes and press the "Apply" button.

[0075] Server: Posts the modified text to various communication platforms.

[0076] The present invention enables users to prevent misunderstandings and troubles on SNS and to communicate with peace of mind.

[0077] The processing flow will be explained below.

[0078] Step 1:

[0079] The user inputs the content of the SNS post into the input interface of the device. For example, the user inputs "Hello, I love dogs!"

[0080] Step 2:

[0081] The terminal transmits the input text data to the server. If image data is included, the image data is also transmitted.

[0082] Step 3:

[0083] The server passes the received data to the generative AI model to begin analysis, which then analyzes the text to detect inappropriate or potentially misleading expressions.

[0084] Step 4:

[0085] The server generates suggested corrections based on the detected issues. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0086] Step 5:

[0087] The server transmits the generated revision proposal to the user terminal, which displays the received revision proposal on its interface.

[0088] Step 6:

[0089] The user reviews the proposed revisions. If they are appropriate, they press the "Apply" button. If there are multiple revision options, the user selects the most appropriate one.

[0090] Step 7:

[0091] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[0092] Step 8:

[0093] The server posts the verified and applied text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[0094] Step 9:

[0095] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[0096] Example 1

[0097] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0098] It is necessary to prevent problems and misunderstandings that may arise from the use of inappropriate or potentially misleading expressions on communication platforms such as social networking sites, and to provide an environment where users can communicate with peace of mind.

[0099] 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.

[0100] In this invention, the server includes a processing means including a generative AI model that analyzes input text and detects inappropriate or potentially misleading expressions, a means for generating correction suggestions based on the detected problems, and a means for displaying and presenting the generated correction suggestions on a user interface, thereby enabling users to correct the content of their SNS posts and prevent problems and misunderstandings from occurring.

[0101] "Means for receiving input text" refers to the interface that allows users to input social media posts and text data into the system, as well as the technology for receiving such data.

[0102] "Processing means including a generative AI model that analyzes input text" refers to generative AI models and related analytical technologies in general that analyze text data entered by a user and detect inappropriate expressions or parts that may be misleading.

[0103] The "means for generating correction suggestions" refers to a general technology that generates suggestions for correcting inappropriate expressions into appropriate expressions while maintaining the user's original intent as much as possible, based on the detected problems.

[0104] The "means for presenting the generated revision proposal" refers to the interface and technology in general for displaying the revision proposal sent from the server to the user terminal in an easy-to-understand manner for the user.

[0105] "Means for reviewing and applying suggested fixes" refers to the general interface and technology that allows a user to review the suggested fixes and accept operations to actually apply them.

[0106] "Means for posting the confirmed and applied text to various communication platforms" refers to all technologies that allow users to safely post the revised text to various social networking sites and communication platforms.

[0107] "Means for a user terminal to display suggested revisions on a user interface" refers to the general interface and technology that displays suggested revisions sent from a server on a user terminal and allows the user to confirm them.

[0108] "Means for the server to perform final check" refers to the general technology by which the system checks the contents again and performs a final check after the user applies the proposed correction.

[0109] The present invention is a system that analyzes text and images that users are about to post on social networking sites in advance, detects inappropriate expressions or parts that may be misleading, and suggests corrections. The system is composed of a user terminal and a server.

[0110] System configuration and operation

[0111] User terminal

[0112] A user terminal is a device that provides an interface for users to input content for posting on social media. This includes smartphones, tablets, and PCs. For example, if a user types "Hello, I love dogs!", this content is sent to the server via the terminal. The user terminal has a means to receive the input text and immediately send it to the server.

[0113] server

[0114] The server is responsible for analyzing the input data received from the user device. It uses a generative AI model for analysis. Specifically, the server performs the following processes:

[0115] 1. Data Analysis:

[0116] The server uses a generative AI model to analyze text and image data sent from the user's device. For example, it suggests correcting "hello" to "hello" and "dog" to "dog." During this process, the prompt sentence is input into the generative AI model, which then analyzes the data.

[0117] 2. Generate corrections:

[0118] Based on the analysis results, a correction suggestion is generated. The generated correction suggestion is changed to an appropriate expression while maintaining the user's original intent. For example, a correction suggestion might be generated: "Hello, I love dogs!"

[0119] 3. Submitting amendments:

[0120] The generated revision proposal is sent to the user terminal, which displays the revision proposal on a user interface for the user to confirm.

[0121] 4. Reconfirmation after applying the proposed amendment:

[0122] After the user applies the suggested corrections, the server checks them again and posts the finalized text to the social networking platform, for example, "Hello, I love dogs!"

[0123] Specific examples

[0124] Example 1: Correcting typos and omissions

[0125] User: The user types, "Hi, I love dogs!"

[0126] Server: The server generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[0127] User: The user reviews the proposed changes and presses the "Apply" button.

[0128] Server: Post the modified text to the social media platform.

[0129] Prompt Sentence Examples

[0130] Here are some example prompts to input to a generative AI model:

[0131] "Please correct this sentence to correct Japanese: Hello, I love dogs!"

[0132] The present invention allows users to prevent misunderstandings and troubles on SNS and communicate with peace of mind.

[0133] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0134] Step 1:

[0135] The user enters the content to post on SNS.

[0136] The user types "Hello, I love dogs!" into the terminal. The terminal receives this input and stores it as text data.

[0137] Step 2:

[0138] The terminal transmits the text data to the server.

[0139] When the user completes the input and presses the send button, the terminal sends this text data to the server, for example, via an API.

[0140] Step 3:

[0141] The server analyzes the data and detects profanity.

[0142] The server inputs the received text data into a generative AI model to generate a prompt: "Please correct this sentence to correct Japanese: Hello, I love dogs!" The generative AI model analyzes this prompt and detects inappropriate or misleading expressions.

[0143] Step 4:

[0144] The server generates a revision proposal and sends it to the device.

[0145] The server generates a correction suggestion based on the prompt sentence. For example, it generates a correction suggestion such as "Hello, I love dogs!" The generated correction suggestion is sent from the server to the terminal as an HTTP response.

[0146] Step 5:

[0147] The device displays suggested fixes to the user.

[0148] The device displays the suggested corrections received from the server in its user interface, which is presented to the user as a popup or notification. For example, a suggested correction might be "Hi, I love dogs!"

[0149] Step 6:

[0150] The user reviews and applies the proposed fixes.

[0151] The user checks the displayed correction suggestions and presses the "Apply" button. The terminal receives this operation and confirms the corrected text data.

[0152] Step 7:

[0153] The terminal transmits the modified data to the server.

[0154] The device sends the confirmed corrected text "Hello, I love dogs!" again to the server, which receives it for final confirmation.

[0155] Step 8:

[0156] The server performs a final check and posts to the social media platform.

[0157] The server performs a final check of the corrected text data and posts it to various social media platforms via the social media API. As a result, the user's post is published on social media as "Hello, I love dogs!"

[0158] (Application example 1)

[0159] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0160] In traditional brick-and-mortar stores, there is a risk that store staff will use misleading or inappropriate language when dealing with customers, which can lead to lower customer satisfaction and complaints. Furthermore, there was no effective tool for store staff to check and correct appropriate language in real time, making it difficult to communicate promptly on the spot. There is a need to solve the problems that arise from this.

[0161] 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.

[0162] In this invention, the server includes: means for receiving input text; processing means including a generative AI model that analyzes the input text and detects inappropriate or potentially misleading expressions; means for the processing means to analyze voice input from customer interactions in a physical store in real time, determine whether the text contains inappropriate expressions, and suggest appropriate corrections; means for a store clerk to select appropriate expressions based on the suggested corrections; means for suggesting the generated corrections; means for confirming and applying the corrections; and means for posting the confirmed and applied text to various communication platforms. This allows store clerks to confirm and correct appropriate expressions in real time, enabling smooth and secure communication with customers.

[0163] "Means for receiving input text" is a general term for devices and software that provide an interface for transmitting text data input by a user to a server and receiving it.

[0164] "Processing means including a generative AI model that analyzes the input text and detects inappropriate or misleading expressions" is a general term for devices and software that analyze received text data and use natural language processing technology to detect inappropriate or misleading expressions in real time.

[0165] "Means for the processing means to analyze voice inputs made in customer interactions in a physical store in real time, determine whether inappropriate expressions are included, and suggest appropriate corrections" is a general term for devices and software that analyze the voices of store staff in a physical store environment in real time, determine whether inappropriate expressions are included, and suggest corrections to appropriate expressions.

[0166] "Means for allowing the clerk to select an appropriate expression based on the suggested revisions" is a general term for devices and software that present suggested revisions to the clerk visually or audibly and provide an interface that enables the clerk to select an appropriate expression.

[0167] "Means for presenting the generated revision suggestions" is a general term for devices and software for visually or audibly presenting revision suggestions generated by a generative AI model to a user.

[0168] "Means for reviewing and applying the proposed revisions" refers collectively to devices and software that provide an interface through which a user can review the proposed revisions and accept operations to apply them.

[0169] "Means for posting the confirmed and applied text on various communication platforms" is a general term for devices and software that automatically post the corrected text on communication platforms such as SNS.

[0170] This invention relates to a system that enables store clerks to check and correct appropriate expressions in real time when dealing with customers in a physical store. This system allows store clerks to avoid misunderstandings and inappropriate language when communicating with customers, thereby improving service quality.

[0171] System configuration

[0172] The system includes the following major components:

[0173] 1. User Device

[0174] The user terminal is a smartphone or smart glasses used by the store clerk. This terminal provides a voice input interface and receives the store clerk's speech in real time.

[0175] The input voice data is sent to a server, and a correction suggestion is received from the server and presented to the store clerk.

[0176] 2. Server

[0177] The server hosts the generative AI model and analyzes the voice data received from the user device.

[0178] Detects inappropriate or potentially misleading content and generates suggested corrections.

[0179] The generated correction proposal is sent to the user's terminal so that the store clerk can check and apply it.

[0180] Server Operation

[0181] 1. Data Reception

[0182] The server receives the voice data transmitted from the user terminal.

[0183] The audio data is converted into text data using natural language processing techniques.

[0184] 2. Data Analysis

[0185] A generative AI model on the server analyzes the text data and detects any inappropriate or potentially misleading expressions.

[0186] If the user says, "This product is very cheap, so you'd be missing out if you didn't buy it!", the system generates a correction suggestion: "This product is a great deal, so please consider it."

[0187] 3. Generate correction suggestions

[0188] Generate suggested fixes based on detected issues.

[0189] The generated revision suggestions are modified to appropriate and safe expressions while maintaining the user's original intent.

[0190] 4. Submitting amendments

[0191] The generated revision proposal is sent to the user terminal so that the store clerk can check it.

[0192] The store clerk selects the appropriate expression based on the suggested revisions and speaks to the customer.

[0193] Component Details

[0194] Speech recognition library: The speech recognition library used on the user device is, for example, Google's SpeechRecognition API.

[0195] Generative AI model: For the generative AI model implemented on the server side, we use, for example, OpenAI's GPT-3.

[0196] User Interface: The interface of the user terminal is an interface for presenting the proposed revisions to the user visually or audibly and accepting their confirmation and application.

[0197] Communication platform: The system automatically posts the generated revised text to a communication platform (such as a social networking site).

[0198] Specific examples

[0199] Example 1: Voice input and suggested corrections

[0200] User terminal: A salesperson uses smart glasses to tell a customer, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[0201] Server: The server receives the recorded voice data and uses the generation AI to generate a correction suggestion such as, "This product is a great deal, so please consider it."

[0202] User device: The salesperson checks the suggested revisions on the display screen of the smart glasses and communicates them to the customer using appropriate expressions.

[0203] Prompt Sentence Examples

[0204] Example of input prompt: "This product is very cheap, so you'd be missing out if you didn't buy it!" Please change this to an appropriate expression.

[0205] In this way, the present invention enables store staff to communicate appropriately in real time when dealing with customers in a physical store.

[0206] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0207] Step 1:

[0208] The user provides voice input

[0209] Subject: User

[0210] Specific operation: The user (store clerk) wears a smartphone or smart glasses and inputs voice data while talking to a customer. For example, the user might say, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[0211] Input: Voice of store clerk

[0212] Output: Audio data recorded on a smart device

[0213] Step 2:

[0214] Sending audio data

[0215] Subject: Terminal

[0216] Specific operation: The device (smartphone or smart glasses) sends the recorded audio data to the server, which then compresses the data in real time and sends it to the server via a communication line.

[0217] Input: Recorded audio data

[0218] Output: Audio data sent to the server

[0219] Step 3:

[0220] Converting audio data to text

[0221] Subject: Server

[0222] Specific operation: The server uses a speech recognition library (for example, Google's SpeechRecognition API) to convert the received voice data into text data, thereby obtaining the clerk's speech in text format.

[0223] Input: Audio data sent to the server

[0224] Output: Text-formatted speech data

[0225] Step 4:

[0226] Profanity detection

[0227] Subject: Server

[0228] How it works: A generative AI model (such as OpenAI's GPT-3) on the server analyzes text data and detects inappropriate or potentially misleading expressions. For example, it might detect "It's so cheap, you'd be missing out if you didn't buy it."

[0229] Input: Text-formatted speech data

[0230] Output: Text data with inappropriate expressions pointed out

[0231] Step 5:

[0232] Generate correction suggestions

[0233] Subject: Server

[0234] How it works: Based on the detected issues, the server uses a generative AI model to generate appropriate correction suggestions, such as "This product is a great deal, so please consider it."

[0235] Input: Text data containing inappropriate content

[0236] Output: Text data containing suggested revisions

[0237] Step 6:

[0238] Submitting a proposed revision

[0239] Subject: Server

[0240] Specific operation: The server sends the generated revision proposal to the user's device. This transmission is done in real time, so the revision proposal can be viewed instantly on the user's device.

[0241] Input: Text data containing suggested revisions

[0242] Output: The proposed fix sent to the user's device

[0243] Step 7:

[0244] Review and apply the proposed fixes

[0245] Subject: Terminal

[0246] Specific operation: The generated revision suggestions are displayed on the interface of the user device (smart glasses or smartphone). The store clerk can review them and choose whether to adopt the suggestions.

[0247] Input: Proposed correction sent to user device

[0248] Output: Expression data confirmed and applied by the store clerk

[0249] Step 8:

[0250] Communicating the revised statement to customers

[0251] Subject: User

[0252] Specific actions: The salesperson selects the appropriate expression based on the suggested corrections and responds to the customer. For example, they might say, "This product is a great deal, so please consider it."

[0253] Input: Expression data confirmed and applied by the store clerk

[0254] Output: Appropriate response to the customer

[0255] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0256] The present invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, more appropriate revision suggestions can be made. Specific embodiments and their operation are described below.

[0257] System configuration

[0258] The system includes the following major components:

[0259] 1. User Device

[0260] It provides an interface for users to input content to post on social media.

[0261] The input text and image data are sent to the server.

[0262] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[0263] The user is then prompted to confirm and apply the proposed changes.

[0264] 2. Server

[0265] It hosts generative AI models and emotion engines and analyzes data received from users.

[0266] Detect inappropriate or misleading language and generate suggested corrections.

[0267] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[0268] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0269] How it works

[0270] User terminal operation

[0271] 1. Input

[0272] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[0273] 2. Data Transmission

[0274] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[0275] 3. Receiving and Displaying Proposed Amendments

[0276] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0277] Server Operation

[0278] 1. Data Analysis

[0279] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[0280] 2. Generate proposed fixes

[0281] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0282] 3. Emotional Engine Activation

[0283] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[0284] 4. Adjustment of amendments

[0285] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[0286] 5. Submitting amendments

[0287] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0288] 6. Reconfirmation after application of the proposed amendments

[0289] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[0290] Specific examples

[0291] Example 1: Correcting typos and omissions based on emotions

[0292] User: Type "Hi, I love dogs!"

[0293] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[0294] Device: Presents optimal fix suggestions to the user.

[0295] User: Check the proposed changes and press the "Apply" button.

[0296] Server: Posts the finalized text to various communication platforms.

[0297] Example 2: Modifying provocative expressions and reflecting emotions

[0298] User: Type "Your opinion is completely meaningless!"

[0299] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's anger and generates a milder correction suggestion accordingly.

[0300] Device: Presents optimal fix suggestions to the user.

[0301] User: Check the proposed changes and press the "Apply" button.

[0302] Server: Posts the finalized text to various communication platforms.

[0303] This invention allows users to prevent misunderstandings and troubles on social networking sites, allowing them to communicate with peace of mind. In addition, by combining it with an emotion engine, it is possible to provide more appropriate revision suggestions that take into account the user's emotions.

[0304] The processing flow will be explained below.

[0305] Specific processing steps of the operation

[0306] 1. User Input

[0307] Step 1:

[0308] A user inputs text for posting to an SNS into an input interface of a device. For example, the user inputs "Hello, I love dogs!"

[0309] Step 2:

[0310] When the user has completed the input, he / she presses the "Send" button, causing the terminal to send the text data to the server.

[0311] 2.Data analysis by the server

[0312] Step 3:

[0313] The server passes the received text data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate expressions or parts that may be misleading.

[0314] Step 4:

[0315] The server generates a suggested fix based on the detected problem, for example, "Hi, I love dogs!"

[0316] Step 5:

[0317] After generating the revision suggestions, the server uses an emotion engine to recognize the user's emotions and determine whether the input is based on positive, negative, or neutral emotions.

[0318] Step 6:

[0319] The server further examines the proposed revisions based on the emotion engine's judgment and makes adjustments that take into account the user's emotions, for example, softening strong language.

[0320] 3. Present and confirm proposed revisions

[0321] Step 7:

[0322] The server transmits the generated and adjusted correction proposal to the user terminal, for example, "Hello, I love dogs!"

[0323] Step 8:

[0324] The terminal displays the received correction proposals on the user interface in real time. The user checks the proposed correction proposals and selects whether to apply the proposals by pressing the "Apply" button.

[0325] 4. User confirmation and application

[0326] Step 9:

[0327] The user checks the proposed corrections and presses the "Apply" button. If multiple corrections are presented, the user selects the most appropriate one.

[0328] 5. Final submission

[0329] Step 10:

[0330] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[0331] Step 11:

[0332] The server posts the verified and finalized text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[0333] Step 12:

[0334] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[0335] Specific examples

[0336] Example 1: Correcting typos and omissions

[0337] Step 1:

[0338] The user types, "Hi, I love dogs!"

[0339] Step 2:

[0340] The terminal transmits this text data to the server.

[0341] Step 3:

[0342] The server passes the text data to a generative AI model for analysis, which generates suggestions to correct "hello" to "hello" and "wanchan" to "inu."

[0343] Step 4:

[0344] The server applies an emotion engine to the text to recognize positive emotions.

[0345] Step 5:

[0346] The server adjusts the proposed correction "Hello, I love dogs!" based on the emotion and sends it to the user terminal.

[0347] Step 6:

[0348] The terminal presents suggested revisions to the user.

[0349] Step 7:

[0350] The user checks the proposed modifications and presses the "Apply" button.

[0351] Step 8:

[0352] The terminal resends the corrected text to the server.

[0353] Step 9:

[0354] The server posts the confirmed text to the SNS.

[0355] Step 10:

[0356] The terminal and server notify the user that the posting is complete.

[0357] Example 2

[0358] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0359] Posts on conventional social networking services (SNS) often contain inappropriate or potentially misleading language, which can lead to conflicts between users. Furthermore, users often post offensive content influenced by their own emotions, which further contributes to the conflict. To solve these problems, it is necessary to not only properly analyze the content of users' posts and suggest corrections, but also to generate appropriate corrections that take the user's emotions into account.

[0360] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating revision suggestions based on problems detected by the processing means, and means including an emotion engine that recognizes the user's emotions regarding the revision suggestions and generates revision suggestions that take the emotions into consideration. This makes it possible to present not only appropriate revision suggestions for content posted by a user, but also revision suggestions that take the user's emotions into consideration.

[0361] The "means for receiving input text" refers to a device or software having the function of acquiring text data input by the user and incorporating it into the system.

[0362] "Processing means including generative AI models" refers to algorithms or devices that use artificial intelligence to analyze user-entered text and detect inappropriate or potentially misleading language.

[0363] A "means for generating suggested corrections based on problems" is a device or software that has the function of suggesting appropriate correction methods or expressions for inappropriate expressions detected by the generative AI model.

[0364] "Means including an emotion engine" refers to an algorithm or device for recognizing emotions based on user input and generating emotion-sensitive revision suggestions.

[0365] The "means for presenting revision suggestions" is a device or software that has the function of visually displaying the generated revision suggestions to the user.

[0366] "Means for reviewing and applying proposed amendments" means a device or software that has the functionality to allow a user to review proposed amendments and formally adopt the proposed amendments.

[0367] "Means for posting to various communication platforms" refers to devices or software that have the function of posting the text that the user has finally confirmed and applied to a communication platform such as a social networking site.

[0368] The "means for receiving input image data" refers to a device or software having a function for acquiring image data input by a user and incorporating it into the system.

[0369] The "means for presenting multiple revision suggestions and allowing the user to select one" refers to a device or software that has the function of presenting multiple generated revision suggestions to the user and allowing the user to select the most appropriate revision suggestion.

[0370] A "generative AI model" is a model or algorithm that uses artificial intelligence to analyze text data and generate appropriate sentences.

[0371] A "prompt sentence" is an input sentence used to give instructions to a generative AI model, and is specific input data that serves as a reference when analyzing text and generating revision suggestions.

[0372] This invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[0373] System configuration

[0374] The system includes the following major components:

[0375] 1. User Device

[0376] It provides an interface for users to input content to post on social media.

[0377] The input text and image data are sent to the server.

[0378] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[0379] The user is then prompted to confirm and apply the proposed changes.

[0380] 2. Server

[0381] It hosts generative AI models and emotion engines and analyzes data received from users.

[0382] Detect inappropriate or misleading language and generate suggested corrections.

[0383] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[0384] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0385] User terminal processing

[0386] 1. Input:

[0387] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[0388] 2. Data transmission:

[0389] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[0390] 3. Receiving and Viewing Amendments:

[0391] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0392] Server Processing

[0393] 1. Data Analysis:

[0394] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[0395] 2. Generate corrections:

[0396] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0397] 3. Emotional Engine Activation:

[0398] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[0399] 4. Adjustment of amendments:

[0400] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[0401] 5. Submitting amendments:

[0402] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0403] 6. Reconfirmation after applying the proposed amendment:

[0404] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[0405] Specific examples

[0406] Example 1: Correcting typos and omissions based on emotions

[0407] User: Type "Hi, I love dogs!"

[0408] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[0409] Device: Presents optimal fix suggestions to the user.

[0410] User: Check the proposed changes and press the "Apply" button.

[0411] Server: Posts the finalized text to various communication platforms.

[0412] Example 2: Modifying provocative expressions and reflecting emotions

[0413] User: Type "Your opinion is completely meaningless!"

[0414] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's angry emotion and generates a milder correction suggestion accordingly.

[0415] Device: Presents optimal fix suggestions to the user.

[0416] User: Check the proposed changes and press the "Apply" button.

[0417] Server: Posts the finalized text to various communication platforms.

[0418] Prompt Sentence Examples

[0419] "Please fix the typo and inappropriate language in the following text: 'Hi, I love dogs!'"

[0420] "Please change the inflammatory language in this text to a milder one: 'Your opinion is completely meaningless!'"

[0421] This system allows users to post with peace of mind while avoiding misunderstandings and troubles on social media. Furthermore, by combining it with an emotion engine, it can provide appropriate revision suggestions that take into consideration the user's emotions.

[0422] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0423] Step 1: Input

[0424] User: Enter the content they want to post on the SNS into the user device interface. For example, they enter "Hello, I love dogs!" into the text box.

[0425] Input: Social media post content (text data).

[0426] Output: Input data stored on the user's device.

[0427] Step 2: Send data

[0428] Terminal: When the user presses the send button, the terminal sends the entered text and image data (if any) to the server.

[0429] Input: The post content typed by the user.

[0430] Output: Text and image data sent to the server.

[0431] What it does: Data is divided into packets and sent over the Internet to a server using a communication protocol (e.g. HTTP).

[0432] Step 3: Data analysis

[0433] Server: The server passes the received text and image data to a generative AI model, which analyzes the input and detects inappropriate or potentially misleading language.

[0434] Input: User submissions (text and image data).

[0435] Output: A list of detected issues.

[0436] How it works: The generative AI model runs text analysis algorithms and uses natural language processing techniques to identify inappropriate language. Similarly, for images, it runs image analysis algorithms if they contain inappropriate content.

[0437] Step 4: Generate correction suggestions

[0438] Server: The server generates fixes based on the detected issues. The generative AI model considers the user's intent and proposes optimal fixes.

[0439] Input: A list of detected issues.

[0440] Output: A list of suggested fixes.

[0441] How it works: The generative AI model generates and suggests appropriate expressions based on the training dataset. For example, it generates a correction to "Hello, I love dogs!" in response to "Hello, I love dogs!"

[0442] Step 5: Emotion Recognition

[0443] Server: Runs the emotion engine to recognize emotions from the user's text and image data, for example, determining the emotional tone (positive, negative, neutral) of a sentence or image.

[0444] Input: User submissions (text and image data).

[0445] Output: Sentiment tag (e.g. positive, negative, neutral).

[0446] How it works: The emotion engine uses natural language processing and image analysis to analyze the sentiment of text and images and assign emotional tags.

[0447] Step 6: Adjust the proposed amendment

[0448] Server: Based on the recognition results of the emotion engine, the generated correction proposal is adjusted. If a strong emotion is included, the proposal is changed to a more tolerant expression, for example.

[0449] Input: revision suggestion list and sentiment tags.

[0450] Output: A refined list of proposed fixes.

[0451] What it does: A tuning algorithm is run to fine-tune the generated suggestions to take sentiment into account.

[0452] Step 7: Submit your proposed revisions

[0453] Server: Sends the adjusted correction proposal to the user device.

[0454] Input: The adjusted list of amendments.

[0455] Output: The proposed fix sent to the user's device.

[0456] Specific operation: The proposed revision data is packetized and transmitted to the user terminal via the Internet.

[0457] Step 8: Receive and view proposed revisions

[0458] Terminal: The proposed modifications sent from the server are displayed on the user interface. The user can review the proposed modifications and decide whether to apply them.

[0459] Input: The proposed fix sent by the server.

[0460] Output: The suggested fixes displayed in the user interface.

[0461] Specific operation: Retrieves data from the receiving buffer and reflects the received correction suggestions in the UI component.

[0462] Step 9: Review and apply proposed fixes

[0463] User: Review the proposed fix and click the "Apply" button to officially apply the fix.

[0464] Input: The suggested fix.

[0465] Output: Confirmation and applied text.

[0466] What happens: Once the proposed fix is ​​applied, a confirmation dialog will appear, allowing the user to make a final confirmation.

[0467] Step 10: Post

[0468] Server: After the user applies the suggested corrections, it checks them again and posts the finalized text to the social media platform.

[0469] Input: Confirmed and applied text.

[0470] Output: The text posted to the social media platform.

[0471] Specific operation: The confirmed text data is posted via the SNS API.

[0472] (Application example 2)

[0473] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0474] On social media and other online platforms, users often unconsciously use inappropriate or potentially misleading language. Furthermore, emotional comments made by users often lead to trouble and misunderstandings. A system to prevent such problems is needed.

[0475] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating correction suggestions based on the detected problems, and means including an emotion recognition engine that recognizes the user's emotions regarding the generated correction suggestions and presents correction suggestions that take the emotions into consideration. This makes it possible to correct inappropriate expressions or misleading expressions while taking the user's emotions into consideration.

[0476] The "means for receiving input text" refers to a device or software that has the function of acquiring text data input by the user and transmitting it to the system.

[0477] "Processing means including a generative AI model that analyzes and detects inappropriate or potentially misleading language" refers to a system function that includes an AI algorithm that uses natural language processing techniques to analyze input text and identify inappropriate or potentially misleading language.

[0478] The "means for generating proposed fixes based on detected problems" is a processing function for creating alternatives to resolve problems found by the analysis. The means utilizes a generative AI model to automatically generate proposed fixes.

[0479] The "means including an emotion recognition engine" is a processing system for analyzing emotions from user input and providing optimal revision suggestions based on the emotion data.

[0480] "Means for confirming and applying proposed revisions" refers to a function that provides an operational interface for users to confirm revisions proposed by the system and officially select and apply those revisions.

[0481] "Means for posting the confirmed and applied text to various communication platforms" refers to a processing means for sending and posting the amendments selected and applied by the user to the relevant social networking site or other online communication tool.

[0482] This invention is a system that analyzes content posted by users on social networking sites, detects inappropriate or potentially misleading expressions, and suggests corrections. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate corrections. Specific embodiments and their operation are described below.

[0483] System configuration

[0484] The system includes the following major components:

[0485] User terminal

[0486] It provides an interface for users to input content to post on social media.

[0487] The input text and image data are sent to the server.

[0488] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[0489] The user is then prompted to confirm and apply the proposed changes.

[0490] server

[0491] It hosts generative AI models and emotion engines and analyzes data received from users.

[0492] Detect inappropriate or misleading language and generate suggested corrections.

[0493] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[0494] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0495] How it works

[0496] 1. Data Entry

[0497] The user enters the content of the SNS post into the user device interface. For example, they might enter, "This app is really useless!"

[0498] 2. Data Transmission

[0499] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[0500] 3. Data Analysis

[0501] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[0502] 4. Generate correction suggestions

[0503] Based on the detected issues, the generative AI model generates the most appropriate correction suggestions while taking into account the user's intent. For example, it generates a correction suggestion such as, "This app still has room for improvement."

[0504] 5. Emotion recognition

[0505] The server activates an emotion recognition engine to recognize emotions from the user's input and image data. For example, if the input text expresses anger, the server generates a revision suggestion that takes this into consideration.

[0506] 6. Adjustment of amendments

[0507] Based on the emotions recognized by the emotion recognition engine, the proposed corrections are further refined and adjusted to take the user's emotions into consideration. For example, if anger is recognized, the expression will be changed to a calmer one.

[0508] 7. Proposal of amendments

[0509] The generated proposed corrections are sent to the user's terminal for the user to check. The user can confirm the proposed corrections and apply them by pressing the "Apply" button.

[0510] 8. Confirm and post the applied text

[0511] After the user confirms and applies the proposed changes, the finalized text is posted to various communication platforms. For example, the revised version of "This app still has room for improvement, but it has helped me in some ways" can be safely posted on social media.

[0512] Hardware and software used

[0513] Smartphone

[0514] Application execution environment

[0515] Mobile data transmission and reception

[0516] server

[0517] AWS (cloud server)

[0518] Google Cloud (cloud server)

[0519] Hosting generative AI models

[0520] Emotion Engine Hosting

[0521] Generative AI Models

[0522] GPT-4 and other modern natural language processing (NLP) models

[0523] Custom NLP model for profanity detection

[0524] Emotion Recognition Engine

[0525] Emotion recognition model created using TensorFlow and Keras

[0526] Specific examples

[0527] User: Type "This app is completely useless!"

[0528] Generative AI model: Generates suggested fixes such as, "This app still has room for improvement."

[0529] Emotion recognition engine: Recognizes the user's angry emotion and generates adjustment suggestions such as, "This app still has room for improvement, but it has helped in some ways."

[0530] User device: Check the suggested fixes and press the "Apply" button.

[0531] Server: Posts the confirmed text to the social networking site.

[0532] Example prompts to input to a generative AI model:

[0533] Original: "This app is completely useless!"

[0534] Generative AI: "This app still has room for improvement."

[0535] Emotion Engine: Recognizes the user's angry emotion and generates adjustment suggestions such as "This app still has room for improvement, but it has helped in some ways."

[0536] This makes it possible to take into consideration the user's feelings and prevent misunderstandings and problems in online communication.

[0537] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0538] Step 1:

[0539] The user enters the content of the SNS post into the interface of the user device. The entered text is, "This app is really useless!" This input data is sent to the server in the next step.

[0540] Step 2:

[0541] The user terminal sends the input text data to the cloud server. Here, the mobile data is sent using a smartphone, and the cloud server receives the data.

[0542] Step 3:

[0543] The server passes the received text data to a generative AI model (an NLP model such as GPT-4) to begin analysis. The generative AI model analyzes the input text, "This app is really useless!", and detects inappropriate or potentially misleading expressions. In this step, text data analysis and inappropriate expression detection are performed.

[0544] Step 4:

[0545] The generative AI model generates a suggested fix based on the detected issues, for example, "This app still has room for improvement." The server then processes this suggested fix in the next step.

[0546] Step 5:

[0547] The server starts an emotion recognition engine (a model using TensorFlow or Keras) and analyzes the emotion from the user's input text. For example, it recognizes that the input text contains the emotion of anger. In this step, the emotion data is analyzed.

[0548] Step 6:

[0549] The server further refines the generated revision suggestions based on the emotional data analyzed by the emotion recognition engine. For example, it generates an optimal revision suggestion such as, "This app still has room for improvement, but it was helpful in some ways." In this step, data processing is performed taking emotions into consideration.

[0550] Step 7:

[0551] The server sends the final proposed fixes to the user's device, which then displays them for review. For example, the user might see a suggestion like, "This app still has room for improvement, but it helped in some ways."

[0552] Step 8:

[0553] The user checks the proposed revisions and applies them by pressing the "Apply" button. Here, the user selects a revision and the selection is sent from the terminal to the server.

[0554] Step 9:

[0555] The server receives the verified and applied text and posts it to various communication platforms. For example, the revised text "This app still has room for improvement, but it has helped me in some ways" is posted to social media. This step is the final data transmission.

[0556] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0557] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0558] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0559] [Second embodiment]

[0560] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0561] 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.

[0562] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0563] 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.

[0564] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0565] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0566] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0567] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0568] 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.

[0569] 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.

[0570] 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.

[0571] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0572] The present invention provides a system that detects inappropriate expressions or potentially misleading parts of social media posts entered by users in real time and suggests corrections. Specific embodiments and their operation are described below.

[0573] System configuration

[0574] The system includes the following major components:

[0575] 1. User Device

[0576] It provides an interface for users to input content for posting on social media.

[0577] The input text and image data are sent to the server.

[0578] Receives suggested revisions from the server and presents them to the user.

[0579] The user is then prompted to confirm and apply the proposed changes.

[0580] 2. Server

[0581] It hosts generative AI models and analyzes data received from user devices.

[0582] Detects inappropriate or potentially misleading language and generates suggested corrections.

[0583] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0584] How it works

[0585] User terminal operation

[0586] 1. Input

[0587] The user inputs the content of the SNS post into the interface of the user device. For example, the user might input "Hello, I love dogs!"

[0588] 2. Data Transmission

[0589] Once the input is complete, the user terminal sends this text data to the server. If an image is included, it is also sent.

[0590] 3. Receiving and Displaying Proposed Amendments

[0591] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0592] Server Operation

[0593] 1. Data Analysis

[0594] It analyzes incoming text and image data and uses generative AI models to detect inappropriate or potentially misleading content, suggesting, for example, changing "hello" to "hello" or "dog" to "dog."

[0595] 2. Generate proposed fixes

[0596] Based on the detected problems, a correction suggestion is generated, which is a correction to an appropriate and safe expression while preserving the user's original intent as much as possible.

[0597] 3. Submitting amendments

[0598] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0599] 4. Reconfirmation after application of the proposed amendments

[0600] After the user confirms and applies the proposed edits, they can review the changes and post the finalized text to various communication platforms. For example, the revised text "Hello, I love dogs!" can be safely posted to social media.

[0601] Specific examples

[0602] Example 1: Correcting typos and omissions

[0603] User: Type "Hi, I love dogs!"

[0604] Server: Generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[0605] User: Check the proposed changes and press the "Apply" button.

[0606] Server: Posts the modified text to various communication platforms.

[0607] Example 2: Correcting provocative language

[0608] User: Type "Your opinion is completely meaningless!"

[0609] Server: Generates a correction suggestion saying "I have doubts about your opinion" and sends it to the user's terminal.

[0610] User: Check the proposed changes and press the "Apply" button.

[0611] Server: Posts the modified text to various communication platforms.

[0612] The present invention enables users to prevent misunderstandings and troubles on SNS and to communicate with peace of mind.

[0613] The processing flow will be explained below.

[0614] Step 1:

[0615] The user inputs the content of the SNS post into the input interface of the device. For example, the user inputs "Hello, I love dogs!"

[0616] Step 2:

[0617] The terminal transmits the input text data to the server. If image data is included, the image data is also transmitted.

[0618] Step 3:

[0619] The server passes the received data to the generative AI model to begin analysis, which then analyzes the text to detect inappropriate or potentially misleading expressions.

[0620] Step 4:

[0621] The server generates suggested corrections based on the detected issues. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0622] Step 5:

[0623] The server transmits the generated revision proposal to the user terminal, which displays the received revision proposal on its interface.

[0624] Step 6:

[0625] The user reviews the proposed revisions. If they are appropriate, they press the "Apply" button. If there are multiple revision options, the user selects the most appropriate one.

[0626] Step 7:

[0627] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[0628] Step 8:

[0629] The server posts the verified and applied text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[0630] Step 9:

[0631] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[0632] Example 1

[0633] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0634] It is necessary to prevent problems and misunderstandings that may arise from the use of inappropriate or potentially misleading expressions on communication platforms such as social networking sites, and to provide an environment where users can communicate with peace of mind.

[0635] 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.

[0636] In this invention, the server includes a processing means including a generative AI model that analyzes input text and detects inappropriate or potentially misleading expressions, a means for generating correction suggestions based on the detected problems, and a means for displaying and presenting the generated correction suggestions on a user interface, thereby enabling users to correct the content of their SNS posts and prevent problems and misunderstandings from occurring.

[0637] "Means for receiving input text" refers to the interface that allows users to input social media posts and text data into the system, as well as the technology for receiving such data.

[0638] "Processing means including a generative AI model that analyzes input text" refers to generative AI models and related analytical technologies in general that analyze text data entered by a user and detect inappropriate expressions or parts that may be misleading.

[0639] The "means for generating correction suggestions" refers to a general technology that generates suggestions for correcting inappropriate expressions into appropriate expressions while maintaining the user's original intent as much as possible, based on the detected problems.

[0640] The "means for presenting the generated revision proposal" refers to the interface and technology in general for displaying the revision proposal sent from the server to the user terminal in an easy-to-understand manner for the user.

[0641] "Means for reviewing and applying suggested fixes" refers to the general interface and technology that allows a user to review the suggested fixes and accept operations to actually apply them.

[0642] "Means for posting the confirmed and applied text to various communication platforms" refers to all technologies that allow users to safely post the revised text to various social networking sites and communication platforms.

[0643] "Means for a user terminal to display suggested revisions on a user interface" refers to the general interface and technology that displays suggested revisions sent from a server on a user terminal and allows the user to confirm them.

[0644] "Means for the server to perform final check" refers to the general technology by which the system checks the contents again and performs a final check after the user applies the proposed correction.

[0645] The present invention is a system that analyzes text and images that users are about to post on social networking sites in advance, detects inappropriate expressions or parts that may be misleading, and suggests corrections. The system is composed of a user terminal and a server.

[0646] System configuration and operation

[0647] User terminal

[0648] A user terminal is a device that provides an interface for users to input content for posting on social media. This includes smartphones, tablets, and PCs. For example, if a user types "Hello, I love dogs!", this content is sent to the server via the terminal. The user terminal has a means to receive the input text and immediately send it to the server.

[0649] server

[0650] The server is responsible for analyzing the input data received from the user device. It uses a generative AI model for analysis. Specifically, the server performs the following processes:

[0651] 1. Data Analysis:

[0652] The server uses a generative AI model to analyze text and image data sent from the user's device. For example, it suggests correcting "hello" to "hello" and "dog" to "dog." During this process, the prompt sentence is input into the generative AI model, which then analyzes the data.

[0653] 2. Generate corrections:

[0654] Based on the analysis results, a correction suggestion is generated. The generated correction suggestion is changed to an appropriate expression while maintaining the user's original intent. For example, a correction suggestion might be generated: "Hello, I love dogs!"

[0655] 3. Submitting amendments:

[0656] The generated revision proposal is sent to the user terminal, which displays the revision proposal on a user interface for the user to confirm.

[0657] 4. Reconfirmation after applying the proposed amendment:

[0658] After the user applies the suggested corrections, the server checks them again and posts the finalized text to the social networking platform, for example, "Hello, I love dogs!"

[0659] Specific examples

[0660] Example 1: Correcting typos and omissions

[0661] User: The user types, "Hi, I love dogs!"

[0662] Server: The server generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[0663] User: The user reviews the proposed changes and presses the "Apply" button.

[0664] Server: Post the modified text to the social media platform.

[0665] Prompt Sentence Examples

[0666] Here are some example prompts to input to a generative AI model:

[0667] "Please correct this sentence to correct Japanese: Hello, I love dogs!"

[0668] The present invention allows users to prevent misunderstandings and troubles on SNS and communicate with peace of mind.

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

[0670] Step 1:

[0671] The user enters the content to post on SNS.

[0672] The user types "Hello, I love dogs!" into the terminal. The terminal receives this input and stores it as text data.

[0673] Step 2:

[0674] The terminal transmits the text data to the server.

[0675] When the user completes the input and presses the send button, the terminal sends this text data to the server. For example, the text data is sent to the server through an API.

[0676] Step 3:

[0677] The server analyzes the data and detects profanity.

[0678] The server inputs the received text data into a generative AI model to generate a prompt: "Please correct this sentence to correct Japanese: Hello, I love dogs!" The generative AI model analyzes this prompt and detects inappropriate or misleading expressions.

[0679] Step 4:

[0680] The server generates a revision proposal and sends it to the device.

[0681] The server generates a correction suggestion based on the prompt sentence. For example, it generates a correction suggestion such as "Hello, I love dogs!" The generated correction suggestion is sent from the server to the terminal as an HTTP response.

[0682] Step 5:

[0683] The device displays suggested fixes to the user.

[0684] The device displays the suggested corrections received from the server in its user interface, which is presented to the user as a popup or notification. For example, a suggested correction might be "Hi, I love dogs!"

[0685] Step 6:

[0686] The user reviews and applies the proposed fixes.

[0687] The user checks the displayed correction suggestions and presses the "Apply" button. The terminal receives this operation and confirms the corrected text data.

[0688] Step 7:

[0689] The terminal transmits the modified data to the server.

[0690] The device sends the confirmed corrected text "Hello, I love dogs!" again to the server, which receives it for final confirmation.

[0691] Step 8:

[0692] The server performs a final check and posts to the social media platform.

[0693] The server performs a final check of the corrected text data and posts it to various social media platforms via the social media API. As a result, the user's post is published on social media as "Hello, I love dogs!"

[0694] (Application example 1)

[0695] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0696] In traditional brick-and-mortar stores, there is a risk that store staff will use misleading or inappropriate language when dealing with customers, which can lead to lower customer satisfaction and complaints. Furthermore, there was no effective tool for store staff to check and correct appropriate language in real time, making it difficult to communicate promptly on the spot. There is a need to solve the problems that arise from this.

[0697] 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.

[0698] In this invention, the server includes: means for receiving input text; processing means including a generative AI model that analyzes the input text and detects inappropriate or potentially misleading expressions; means for the processing means to analyze voice input from customer interactions in a physical store in real time, determine whether the text contains inappropriate expressions, and suggest appropriate corrections; means for a store clerk to select appropriate expressions based on the suggested corrections; means for suggesting the generated corrections; means for confirming and applying the corrections; and means for posting the confirmed and applied text to various communication platforms. This allows store clerks to confirm and correct appropriate expressions in real time, enabling smooth and secure communication with customers.

[0699] "Means for receiving input text" is a general term for devices and software that provide an interface for transmitting text data input by a user to a server and receiving it.

[0700] "Processing means including a generative AI model that analyzes the input text and detects inappropriate or misleading expressions" is a general term for devices and software that analyze received text data and use natural language processing technology to detect inappropriate or misleading expressions in real time.

[0701] "Means for the processing means to analyze voice inputs made in customer interactions in a physical store in real time, determine whether inappropriate expressions are included, and suggest appropriate corrections" is a general term for devices and software that analyze the voices of store staff in a physical store environment in real time, determine whether inappropriate expressions are included, and suggest corrections to appropriate expressions.

[0702] "Means for allowing the clerk to select an appropriate expression based on the suggested revisions" is a general term for devices and software that present suggested revisions to the clerk visually or audibly and provide an interface that enables the clerk to select an appropriate expression.

[0703] "Means for presenting the generated revision suggestions" is a general term for devices and software for visually or audibly presenting revision suggestions generated by a generative AI model to a user.

[0704] "Means for reviewing and applying the proposed revisions" refers collectively to devices and software that provide an interface through which a user can review the proposed revisions and accept operations to apply them.

[0705] "Means for posting the confirmed and applied text on various communication platforms" is a general term for devices and software that automatically post the corrected text on communication platforms such as SNS.

[0706] This invention relates to a system that enables store clerks to check and correct appropriate expressions in real time when dealing with customers in a physical store. This system allows store clerks to avoid misunderstandings and inappropriate language when communicating with customers, thereby improving service quality.

[0707] System configuration

[0708] The system includes the following major components:

[0709] 1. User Device

[0710] The user terminal is a smartphone or smart glasses used by the store clerk. This terminal provides a voice input interface and receives the store clerk's speech in real time.

[0711] The input voice data is sent to a server, and a correction suggestion is received from the server and presented to the store clerk.

[0712] 2. Server

[0713] The server hosts the generative AI model and analyzes the voice data received from the user device.

[0714] Detects inappropriate or potentially misleading content and generates suggested corrections.

[0715] The generated correction proposal is sent to the user's terminal so that the store clerk can check and apply it.

[0716] Server Operation

[0717] 1. Data Reception

[0718] The server receives the voice data transmitted from the user terminal.

[0719] The audio data is converted into text data using natural language processing techniques.

[0720] 2. Data Analysis

[0721] A generative AI model on the server analyzes the text data and detects any inappropriate or potentially misleading expressions.

[0722] If the user says, "This product is very cheap, so you'd be missing out if you didn't buy it!", the system generates a correction suggestion: "This product is a great deal, so please consider it."

[0723] 3. Generate correction suggestions

[0724] Generate suggested fixes based on detected issues.

[0725] The generated revision suggestions are modified to appropriate and safe expressions while maintaining the user's original intent.

[0726] 4. Submitting amendments

[0727] The generated revision proposal is sent to the user terminal so that the store clerk can check it.

[0728] The store clerk selects the appropriate expression based on the suggested revisions and speaks to the customer.

[0729] Component Details

[0730] Speech recognition library: The speech recognition library used on the user device is, for example, Google's SpeechRecognition API.

[0731] Generative AI model: For the generative AI model implemented on the server side, we use, for example, OpenAI's GPT-3.

[0732] User Interface: The interface of the user terminal is an interface for presenting the proposed revisions to the user visually or audibly and accepting their confirmation and application.

[0733] Communication platform: The system automatically posts the generated revised text to a communication platform (such as a social networking site).

[0734] Specific examples

[0735] Example 1: Voice input and suggested corrections

[0736] User terminal: A salesperson uses smart glasses to tell a customer, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[0737] Server: The server receives the recorded voice data and uses the generation AI to generate a correction suggestion such as, "This product is a great deal, so please consider it."

[0738] User device: The salesperson checks the suggested revisions on the display screen of the smart glasses and communicates them to the customer using appropriate expressions.

[0739] Prompt Sentence Examples

[0740] Example of input prompt: "This product is very cheap, so you'd be missing out if you didn't buy it!" Please change this to an appropriate expression.

[0741] In this way, the present invention enables store staff to communicate appropriately in real time when dealing with customers in a physical store.

[0742] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0743] Step 1:

[0744] The user provides voice input

[0745] Subject: User

[0746] Specific operation: The user (store clerk) wears a smartphone or smart glasses and inputs voice data while talking to a customer. For example, the user might say, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[0747] Input: Voice of store clerk

[0748] Output: Audio data recorded on a smart device

[0749] Step 2:

[0750] Sending audio data

[0751] Subject: Terminal

[0752] Specific operation: The device (smartphone or smart glasses) sends the recorded audio data to the server, which then compresses the data in real time and sends it to the server via a communication line.

[0753] Input: Recorded audio data

[0754] Output: Audio data sent to the server

[0755] Step 3:

[0756] Converting audio data to text

[0757] Subject: Server

[0758] Specific operation: The server uses a speech recognition library (for example, Google's SpeechRecognition API) to convert the received voice data into text data, thereby obtaining the clerk's speech in text format.

[0759] Input: Audio data sent to the server

[0760] Output: Text-formatted speech data

[0761] Step 4:

[0762] Profanity detection

[0763] Subject: Server

[0764] How it works: A generative AI model (such as OpenAI's GPT-3) on the server analyzes text data and detects inappropriate or potentially misleading expressions. For example, it might detect "It's so cheap, you'd be missing out if you didn't buy it."

[0765] Input: Text-formatted speech data

[0766] Output: Text data with inappropriate expressions pointed out

[0767] Step 5:

[0768] Generate correction suggestions

[0769] Subject: Server

[0770] How it works: Based on the detected issues, the server uses a generative AI model to generate appropriate correction suggestions, such as "This product is a great deal, so please consider it."

[0771] Input: Text data containing inappropriate content

[0772] Output: Text data containing suggested revisions

[0773] Step 6:

[0774] Submitting a proposed revision

[0775] Subject: Server

[0776] Specific operation: The server sends the generated revision proposal to the user's device. This transmission is done in real time, so the revision proposal can be viewed instantly on the user's device.

[0777] Input: Text data containing suggested revisions

[0778] Output: The proposed fix sent to the user's device

[0779] Step 7:

[0780] Review and apply the proposed fixes

[0781] Subject: Terminal

[0782] Specific operation: The generated revision suggestions are displayed on the interface of the user device (smart glasses or smartphone). The store clerk can review them and choose whether to adopt the suggestions.

[0783] Input: Proposed correction sent to user device

[0784] Output: Expression data confirmed and applied by the store clerk

[0785] Step 8:

[0786] Communicating the revised statement to customers

[0787] Subject: User

[0788] Specific actions: The salesperson selects the appropriate expression based on the suggested corrections and speaks to the customer. For example, they might say, "This product is a great deal, so please consider it."

[0789] Input: Expression data confirmed and applied by the store clerk

[0790] Output: Appropriate response to the customer

[0791] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0792] The present invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining this system with an emotion engine that recognizes the user's emotions, more appropriate revision suggestions can be made. Specific embodiments and their operation are described below.

[0793] System configuration

[0794] The system includes the following major components:

[0795] 1. User Device

[0796] It provides an interface for users to input content to post on social media.

[0797] The input text and image data are sent to the server.

[0798] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[0799] The user is then prompted to confirm and apply the proposed changes.

[0800] 2. Server

[0801] It hosts generative AI models and emotion engines and analyzes data received from users.

[0802] Detect inappropriate or misleading language and generate suggested corrections.

[0803] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[0804] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0805] How it works

[0806] User terminal operation

[0807] 1. Input

[0808] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[0809] 2. Data Transmission

[0810] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[0811] 3. Receiving and Displaying Proposed Amendments

[0812] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0813] Server Operation

[0814] 1. Data Analysis

[0815] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[0816] 2. Generate proposed fixes

[0817] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0818] 3. Emotional Engine Activation

[0819] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[0820] 4. Adjustment of amendments

[0821] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[0822] 5. Submitting amendments

[0823] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0824] 6. Reconfirmation after application of the proposed amendments

[0825] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[0826] Specific examples

[0827] Example 1: Correcting typos and omissions based on emotions

[0828] User: Type "Hi, I love dogs!"

[0829] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[0830] Device: Presents optimal fix suggestions to the user.

[0831] User: Check the proposed changes and press the "Apply" button.

[0832] Server: Posts the finalized text to various communication platforms.

[0833] Example 2: Modifying provocative expressions and reflecting emotions

[0834] User: Type "Your opinion is completely meaningless!"

[0835] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's anger and generates a milder correction suggestion accordingly.

[0836] Device: Presents optimal fix suggestions to the user.

[0837] User: Check the proposed changes and press the "Apply" button.

[0838] Server: Posts the finalized text to various communication platforms.

[0839] This invention allows users to prevent misunderstandings and troubles on social networking sites, allowing them to communicate with peace of mind. In addition, by combining it with an emotion engine, it is possible to provide more appropriate revision suggestions that take into account the user's emotions.

[0840] The processing flow will be explained below.

[0841] Specific processing steps of the operation

[0842] 1. User Input

[0843] Step 1:

[0844] A user inputs text for posting to an SNS into an input interface of a device. For example, the user inputs "Hi, I love dogs!"

[0845] Step 2:

[0846] When the user has completed the input, he / she presses the "Send" button, causing the terminal to send the text data to the server.

[0847] 2.Data analysis by the server

[0848] Step 3:

[0849] The server passes the received text data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate expressions or parts that may be misleading.

[0850] Step 4:

[0851] The server generates a suggested fix based on the detected problem, for example, "Hi, I love dogs!"

[0852] Step 5:

[0853] After generating the revision suggestions, the server uses an emotion engine to recognize the user's emotions and determine whether the input is based on positive, negative, or neutral emotions.

[0854] Step 6:

[0855] The server further examines the proposed revisions based on the emotion engine's judgment and makes adjustments that take into account the user's emotions, for example, softening strong language.

[0856] 3. Present and confirm proposed revisions

[0857] Step 7:

[0858] The server transmits the generated and adjusted correction proposal to the user terminal, for example, "Hello, I love dogs!"

[0859] Step 8:

[0860] The terminal displays the received correction proposals on the user interface in real time. The user checks the proposed correction proposals and selects whether to apply the proposals by pressing the "Apply" button.

[0861] 4. User confirmation and application

[0862] Step 9:

[0863] The user checks the proposed corrections and presses the "Apply" button. If multiple corrections are presented, the user selects the most appropriate one.

[0864] 5. Final submission

[0865] Step 10:

[0866] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[0867] Step 11:

[0868] The server posts the verified and finalized text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[0869] Step 12:

[0870] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[0871] Specific examples

[0872] Example 1: Correcting typos and omissions

[0873] Step 1:

[0874] The user types, "Hi, I love dogs!"

[0875] Step 2:

[0876] The terminal transmits this text data to the server.

[0877] Step 3:

[0878] The server passes the text data to a generative AI model for analysis, which generates suggestions to correct "hello" to "hello" and "wanchan" to "inu."

[0879] Step 4:

[0880] The server applies an emotion engine to the text to recognize positive emotions.

[0881] Step 5:

[0882] The server adjusts the proposed correction "Hello, I love dogs!" based on the emotion and sends it to the user terminal.

[0883] Step 6:

[0884] The terminal presents suggested revisions to the user.

[0885] Step 7:

[0886] The user checks the proposed modifications and presses the "Apply" button.

[0887] Step 8:

[0888] The terminal resends the corrected text to the server.

[0889] Step 9:

[0890] The server posts the confirmed text to the SNS.

[0891] Step 10:

[0892] The terminal and server notify the user that the posting is complete.

[0893] Example 2

[0894] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0895] Posts on conventional social networking services (SNS) often contain inappropriate or potentially misleading language, which can lead to conflicts between users. Furthermore, users often post offensive content influenced by their own emotions, which further contributes to the conflict. To solve these problems, it is necessary to not only properly analyze the content of users' posts and suggest corrections, but also to generate appropriate corrections that take the user's emotions into account.

[0896] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating revision suggestions based on problems detected by the processing means, and means including an emotion engine that recognizes the user's emotions regarding the revision suggestions and generates revision suggestions that take the emotions into consideration. This makes it possible to present not only appropriate revision suggestions for content posted by a user, but also revision suggestions that take the user's emotions into consideration.

[0897] The "means for receiving input text" refers to a device or software having the function of acquiring text data input by the user and incorporating it into the system.

[0898] "Processing means including generative AI models" refers to algorithms or devices that use artificial intelligence to analyze user-entered text and detect inappropriate or potentially misleading language.

[0899] A "means for generating suggested corrections based on problems" is a device or software that has the function of suggesting appropriate correction methods or expressions for inappropriate expressions detected by the generative AI model.

[0900] "Means including an emotion engine" refers to an algorithm or device for recognizing emotions based on user input and generating emotion-sensitive revision suggestions.

[0901] The "means for presenting revision suggestions" is a device or software that has the function of visually displaying the generated revision suggestions to the user.

[0902] "Means for reviewing and applying proposed amendments" means a device or software that has the functionality to allow a user to review proposed amendments and formally adopt the proposed amendments.

[0903] "Means for posting to various communication platforms" refers to devices or software that have the function of posting the text that the user has finally confirmed and applied to a communication platform such as a social networking site.

[0904] The "means for receiving input image data" refers to a device or software having a function for acquiring image data input by a user and incorporating it into the system.

[0905] The "means for presenting multiple revision suggestions and allowing the user to select one" refers to a device or software that has the function of presenting multiple generated revision suggestions to the user and allowing the user to select the most appropriate revision suggestion.

[0906] A "generative AI model" is a model or algorithm that uses artificial intelligence to analyze text data and generate appropriate sentences.

[0907] A "prompt sentence" is an input sentence used to give instructions to a generative AI model, and is specific input data that serves as a reference when analyzing text and generating revision suggestions.

[0908] This invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[0909] System configuration

[0910] The system includes the following major components:

[0911] 1. User Device

[0912] It provides an interface for users to input content to post on social media.

[0913] The input text and image data are sent to the server.

[0914] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[0915] The user is then prompted to confirm and apply the proposed changes.

[0916] 2. Server

[0917] It hosts generative AI models and emotion engines and analyzes data received from users.

[0918] Detect inappropriate or misleading language and generate suggested corrections.

[0919] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[0920] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[0921] User terminal processing

[0922] 1. Input:

[0923] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[0924] 2. Data transmission:

[0925] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[0926] 3. Receiving and Viewing Amendments:

[0927] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[0928] Server Processing

[0929] 1. Data Analysis:

[0930] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[0931] 2. Generate corrections:

[0932] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[0933] 3. Emotional Engine Activation:

[0934] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[0935] 4. Adjustment of amendments:

[0936] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[0937] 5. Submitting amendments:

[0938] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[0939] 6. Reconfirmation after applying the proposed amendment:

[0940] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[0941] Specific examples

[0942] Example 1: Correcting typos and omissions based on emotions

[0943] User: Type "Hi, I love dogs!"

[0944] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[0945] Device: Presents optimal fix suggestions to the user.

[0946] User: Check the proposed changes and press the "Apply" button.

[0947] Server: Posts the finalized text to various communication platforms.

[0948] Example 2: Modifying provocative expressions and reflecting emotions

[0949] User: Type "Your opinion is completely meaningless!"

[0950] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's anger and generates a milder correction suggestion accordingly.

[0951] Device: Presents optimal fix suggestions to the user.

[0952] User: Check the proposed changes and press the "Apply" button.

[0953] Server: Posts the finalized text to various communication platforms.

[0954] Prompt Sentence Examples

[0955] "Please fix the typo and inappropriate language in the following text: 'Hi, I love dogs!'"

[0956] "Please change the inflammatory language in this text to a milder one: 'Your opinion is completely meaningless!'"

[0957] This system allows users to post with peace of mind while avoiding misunderstandings and troubles on social media. Furthermore, by combining it with an emotion engine, it can provide appropriate revision suggestions that take into consideration the user's emotions.

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

[0959] Step 1: Input

[0960] User: Enter the content they want to post on the SNS into the user device interface. For example, they might enter "Hello, I love dogs!" into a text box.

[0961] Input: Social media post content (text data).

[0962] Output: Input data stored on the user's device.

[0963] Step 2: Send data

[0964] Terminal: When the user presses the send button, the terminal sends the entered text and image data (if any) to the server.

[0965] Input: The post content typed by the user.

[0966] Output: Text and image data sent to the server.

[0967] What it does: Data is divided into packets and sent over the Internet to a server using a communication protocol (e.g. HTTP).

[0968] Step 3: Data analysis

[0969] Server: The server passes the received text and image data to a generative AI model, which analyzes the input and detects inappropriate or potentially misleading language.

[0970] Input: User submissions (text and image data).

[0971] Output: A list of detected issues.

[0972] How it works: The generative AI model runs text analysis algorithms and uses natural language processing techniques to identify inappropriate language. Similarly, for images, it runs image analysis algorithms if they contain inappropriate content.

[0973] Step 4: Generate correction suggestions

[0974] Server: The server generates fixes based on the detected issues. The generative AI model considers the user's intent and proposes optimal fixes.

[0975] Input: A list of detected issues.

[0976] Output: A list of suggested fixes.

[0977] How it works: The generative AI model generates and suggests appropriate expressions based on the training dataset. For example, it generates a correction to "Hello, I love dogs!" in response to "Hello, I love dogs!"

[0978] Step 5: Emotion Recognition

[0979] Server: Runs the emotion engine to recognize emotions from the user's text and image data, for example, determining the emotional tone (positive, negative, neutral) of a sentence or image.

[0980] Input: User submissions (text and image data).

[0981] Output: Sentiment tag (e.g. positive, negative, neutral).

[0982] How it works: The emotion engine uses natural language processing and image analysis to analyze the sentiment of text and images and assign emotional tags.

[0983] Step 6: Adjust the proposed amendment

[0984] Server: Based on the recognition results of the emotion engine, the generated correction proposal is adjusted. If a strong emotion is included, the proposal is changed to a more tolerant expression, for example.

[0985] Input: revision suggestion list and sentiment tags.

[0986] Output: A refined list of proposed fixes.

[0987] What it does: A tuning algorithm is run to fine-tune the generated suggestions to take sentiment into account.

[0988] Step 7: Submit your proposed revisions

[0989] Server: Sends the adjusted correction proposal to the user device.

[0990] Input: The adjusted list of amendments.

[0991] Output: The proposed fix sent to the user's device.

[0992] Specific operation: The proposed revision data is packetized and transmitted to the user terminal via the Internet.

[0993] Step 8: Receive and view proposed revisions

[0994] Terminal: The proposed modifications sent from the server are displayed on the user interface. The user can review the proposed modifications and decide whether to apply them.

[0995] Input: The proposed fix sent by the server.

[0996] Output: The suggested fixes displayed in the user interface.

[0997] Specific operation: Retrieves data from the receiving buffer and reflects the received correction suggestions in the UI component.

[0998] Step 9: Review and apply proposed fixes

[0999] User: Review the proposed fix and click the "Apply" button to officially apply the fix.

[1000] Input: The suggested fix.

[1001] Output: Confirmation and applied text.

[1002] What happens: Once the proposed fix is ​​applied, a confirmation dialog will appear, allowing the user to make a final confirmation.

[1003] Step 10: Post

[1004] Server: After the user applies the suggested corrections, it checks them again and posts the finalized text to the social media platform.

[1005] Input: Confirmed and applied text.

[1006] Output: The text posted to the social media platform.

[1007] Specific operation: The confirmed text data is posted via the SNS API.

[1008] (Application example 2)

[1009] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1010] On social media and other online platforms, users often unconsciously use inappropriate or potentially misleading language. Furthermore, emotional comments made by users often lead to trouble and misunderstandings. A system to prevent such problems is needed.

[1011] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating correction suggestions based on the detected problems, and means including an emotion recognition engine that recognizes the user's emotions regarding the generated correction suggestions and presents correction suggestions that take the emotions into consideration. This makes it possible to correct inappropriate expressions or misleading expressions while taking the user's emotions into consideration.

[1012] The "means for receiving input text" refers to a device or software that has the function of acquiring text data input by the user and transmitting it to the system.

[1013] "Processing means including a generative AI model that analyzes and detects inappropriate or potentially misleading language" refers to a system function that includes an AI algorithm that uses natural language processing techniques to analyze input text and identify inappropriate or potentially misleading language.

[1014] The "means for generating proposed fixes based on detected problems" is a processing function for creating alternatives to resolve problems found by the analysis. The means utilizes a generative AI model to automatically generate proposed fixes.

[1015] The "means including an emotion recognition engine" is a processing system for analyzing emotions from user input and providing optimal revision suggestions based on the emotion data.

[1016] "Means for confirming and applying proposed revisions" refers to a function that provides an operational interface for users to confirm revisions proposed by the system and officially select and apply those revisions.

[1017] "Means for posting the confirmed and applied text to various communication platforms" refers to a processing means for sending and posting the amendments selected and applied by the user to the relevant social networking site or other online communication tool.

[1018] This invention is a system that analyzes content posted by users on social networking sites, detects inappropriate or potentially misleading expressions, and suggests corrections. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate corrections. Specific embodiments and their operation are described below.

[1019] System configuration

[1020] The system includes the following major components:

[1021] User terminal

[1022] It provides an interface for users to input content to post on social media.

[1023] The input text and image data are sent to the server.

[1024] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1025] The user is then prompted to confirm and apply the proposed changes.

[1026] server

[1027] It hosts generative AI models and emotion engines and analyzes data received from users.

[1028] Detect inappropriate or misleading language and generate suggested corrections.

[1029] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1030] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1031] How it works

[1032] 1. Data Entry

[1033] The user enters the content of the SNS post into the user device interface. For example, they might enter, "This app is really useless!"

[1034] 2. Data Transmission

[1035] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1036] 3. Data Analysis

[1037] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[1038] 4. Generate correction suggestions

[1039] Based on the detected issues, the generative AI model generates the most appropriate correction suggestions while taking into account the user's intent. For example, it generates a correction suggestion such as, "This app still has room for improvement."

[1040] 5. Emotion recognition

[1041] The server activates an emotion recognition engine to recognize emotions from the user's input and image data. For example, if the input text expresses anger, the server generates a revision suggestion that takes this into consideration.

[1042] 6. Adjustment of amendments

[1043] Based on the emotions recognized by the emotion recognition engine, the proposed corrections are further refined and adjusted to take the user's emotions into consideration. For example, if anger is recognized, the expression will be changed to a calmer one.

[1044] 7. Proposal of amendments

[1045] The generated proposed corrections are sent to the user's terminal for the user to review. The user can review the proposed corrections and apply them by pressing the "Apply" button.

[1046] 8. Confirm and post the applied text

[1047] After the user confirms and applies the proposed changes, the finalized text is posted to various communication platforms. For example, the revised version of "This app still has room for improvement, but it has helped me in some ways" can be safely posted on social media.

[1048] Hardware and software used

[1049] Smartphone

[1050] Application execution environment

[1051] Mobile data transmission and reception

[1052] server

[1053] AWS (cloud server)

[1054] Google Cloud (cloud server)

[1055] Hosting generative AI models

[1056] Emotion Engine Hosting

[1057] Generative AI Models

[1058] GPT-4 and other modern natural language processing (NLP) models

[1059] Custom NLP model for profanity detection

[1060] Emotion Recognition Engine

[1061] Emotion recognition model created using TensorFlow and Keras

[1062] Specific examples

[1063] User: Type "This app is completely useless!"

[1064] Generative AI model: Generates suggested fixes such as, "This app still has room for improvement."

[1065] Emotion recognition engine: Recognizes the user's angry emotion and generates adjustment suggestions such as, "This app still has room for improvement, but it has helped in some ways."

[1066] User device: Check the suggested fixes and press the "Apply" button.

[1067] Server: Posts the confirmed text to the social networking site.

[1068] Example prompts to input to a generative AI model:

[1069] Original: "This app is completely useless!"

[1070] Generative AI: "This app still has room for improvement."

[1071] Emotion Engine: Recognizes the user's angry emotion and generates adjustment suggestions such as "This app still has room for improvement, but it has helped in some ways."

[1072] This makes it possible to take into consideration the user's feelings and prevent misunderstandings and problems in online communication.

[1073] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1074] Step 1:

[1075] The user enters the content of a social media post into the interface of the user device. The entered text is, "This app is really useless!" This input data is sent to the server in the next step.

[1076] Step 2:

[1077] The user terminal sends the input text data to the cloud server. Here, the mobile data is sent using a smartphone, and the cloud server receives the data.

[1078] Step 3:

[1079] The server passes the received text data to a generative AI model (an NLP model such as GPT-4) to begin analysis. The generative AI model analyzes the input text, "This app is really useless!", and detects inappropriate or potentially misleading expressions. In this step, text data analysis and inappropriate expression detection are performed.

[1080] Step 4:

[1081] The generative AI model generates a suggested fix based on the detected issues, for example, "This app still has room for improvement." The server then processes this suggested fix in the next step.

[1082] Step 5:

[1083] The server starts an emotion recognition engine (a model using TensorFlow or Keras) and analyzes the emotion from the user's input text. For example, it recognizes that the input text contains the emotion of anger. In this step, the emotion data is analyzed.

[1084] Step 6:

[1085] The server further refines the generated revision suggestions based on the emotional data analyzed by the emotion recognition engine. For example, it generates an optimal revision suggestion such as, "This app still has room for improvement, but it was helpful in some ways." In this step, data processing is performed taking emotions into consideration.

[1086] Step 7:

[1087] The server sends the final proposed fixes to the user's device, which then displays them for review. For example, the user might see a suggestion like, "This app still has room for improvement, but it helped in some ways."

[1088] Step 8:

[1089] The user checks the proposed revisions and applies them by pressing the "Apply" button. Here, the user selects a revision and the selection is sent from the terminal to the server.

[1090] Step 9:

[1091] The server receives the verified and applied text and posts it to various communication platforms. For example, the revised text "This app still has room for improvement, but it has helped me in some ways" is posted to social media. This step is the final data transmission.

[1092] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1093] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1094] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1095] [Third embodiment]

[1096] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1097] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1098] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1099] 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.

[1100] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1101] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1102] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1103] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1104] 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.

[1105] 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.

[1106] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1107] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1108] The present invention provides a system that detects inappropriate expressions or potentially misleading parts of social media posts entered by users in real time and suggests corrections. Specific embodiments and their operation are described below.

[1109] System configuration

[1110] The system includes the following major components:

[1111] 1. User Device

[1112] It provides an interface for users to input content for posting on social media.

[1113] The input text and image data are sent to the server.

[1114] Receives suggested revisions from the server and presents them to the user.

[1115] The user is then prompted to confirm and apply the proposed changes.

[1116] 2. Server

[1117] It hosts generative AI models and analyzes data received from user devices.

[1118] Detects inappropriate or potentially misleading language and generates suggested corrections.

[1119] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1120] How it works

[1121] User terminal operation

[1122] 1. Input

[1123] The user inputs the content of the SNS post into the interface of the user device. For example, the user might input "Hello, I love dogs!"

[1124] 2. Data Transmission

[1125] Once the input is complete, the user terminal sends this text data to the server. If an image is included, it is also sent.

[1126] 3. Receiving and Displaying Proposed Amendments

[1127] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[1128] Server Operation

[1129] 1. Data Analysis

[1130] It analyzes incoming text and image data and uses generative AI models to detect inappropriate or potentially misleading content, suggesting, for example, changing "hello" to "hello" or "dog" to "dog."

[1131] 2. Generate proposed fixes

[1132] Based on the detected problems, a correction suggestion is generated, which is a correction to an appropriate and safe expression while preserving the user's original intent as much as possible.

[1133] 3. Submitting amendments

[1134] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[1135] 4. Reconfirmation after application of the proposed amendments

[1136] After the user confirms and applies the proposed edits, they can review the changes and post the finalized text to various communication platforms. For example, the revised text "Hello, I love dogs!" can be safely posted to social media.

[1137] Specific examples

[1138] Example 1: Correcting typos and omissions

[1139] User: Type "Hi, I love dogs!"

[1140] Server: Generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[1141] User: Check the proposed changes and press the "Apply" button.

[1142] Server: Posts the modified text to various communication platforms.

[1143] Example 2: Correcting provocative language

[1144] User: Type "Your opinion is completely meaningless!"

[1145] Server: Generates a correction suggestion saying "I have doubts about your opinion" and sends it to the user's terminal.

[1146] User: Check the proposed changes and press the "Apply" button.

[1147] Server: Posts the modified text to various communication platforms.

[1148] The present invention enables users to prevent misunderstandings and troubles on SNS and to communicate with peace of mind.

[1149] The processing flow will be explained below.

[1150] Step 1:

[1151] The user inputs the content of the SNS post into the input interface of the device. For example, the user inputs "Hello, I love dogs!"

[1152] Step 2:

[1153] The terminal transmits the input text data to the server. If image data is included, the image data is also transmitted.

[1154] Step 3:

[1155] The server passes the received data to the generative AI model to begin analysis, which then analyzes the text to detect inappropriate or potentially misleading expressions.

[1156] Step 4:

[1157] The server generates suggested corrections based on the detected issues. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[1158] Step 5:

[1159] The server transmits the generated revision proposal to the user terminal, which displays the received revision proposal on its interface.

[1160] Step 6:

[1161] The user reviews the proposed revisions. If they are appropriate, they press the "Apply" button. If there are multiple revision options, the user selects the most appropriate one.

[1162] Step 7:

[1163] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[1164] Step 8:

[1165] The server posts the verified and applied text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[1166] Step 9:

[1167] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[1168] Example 1

[1169] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1170] It is necessary to prevent problems and misunderstandings that may arise from the use of inappropriate or potentially misleading expressions on communication platforms such as social networking sites, and to provide an environment where users can communicate with peace of mind.

[1171] 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.

[1172] In this invention, the server includes a processing means including a generative AI model that analyzes input text and detects inappropriate or potentially misleading expressions, a means for generating suggested revisions based on the detected problems, and a means for displaying and presenting the generated suggested revisions on a user interface, thereby enabling users to correct their SNS posts and prevent problems or misunderstandings from occurring.

[1173] "Means for receiving input text" refers to the interface that allows users to input social media posts and text data into the system, as well as the technology for receiving such data.

[1174] "Processing means including a generative AI model that analyzes input text" refers to generative AI models and related analytical technologies in general that analyze text data entered by a user and detect inappropriate expressions or parts that may be misleading.

[1175] The "means for generating correction suggestions" refers to a general technology that generates suggestions for correcting inappropriate expressions into appropriate expressions while maintaining the user's original intent as much as possible, based on the detected problems.

[1176] The "means for presenting the generated revision proposal" refers to the interface and technology in general for displaying the revision proposal sent from the server to the user terminal in an easy-to-understand manner for the user.

[1177] "Means for reviewing and applying suggested fixes" refers to the general interface and technology that allows a user to review the suggested fixes and accept operations to actually apply them.

[1178] "Means for posting the confirmed and applied text to various communication platforms" refers to all technologies that allow users to safely post the revised text to various social networking sites and communication platforms.

[1179] "Means for a user terminal to display suggested revisions on a user interface" refers to the general interface and technology that displays suggested revisions sent from a server on a user terminal and allows the user to confirm them.

[1180] "Means for the server to perform final check" refers to the general technology by which the system checks the contents again and performs a final check after the user applies the proposed correction.

[1181] The present invention is a system that analyzes text and images that users are about to post on social networking sites in advance, detects inappropriate expressions or parts that may be misleading, and suggests corrections. The system is composed of a user terminal and a server.

[1182] System configuration and operation

[1183] User terminal

[1184] A user terminal is a device that provides an interface for users to input content for posting on social media. This includes smartphones, tablets, and PCs. For example, if a user types "Hello, I love dogs!", this content is sent to the server via the terminal. The user terminal has a means to receive the input text and immediately send it to the server.

[1185] server

[1186] The server is responsible for analyzing the input data received from the user device. It uses a generative AI model for analysis. Specifically, the server performs the following processes:

[1187] 1. Data Analysis:

[1188] The server uses a generative AI model to analyze text and image data sent from the user's device. For example, it suggests correcting "hello" to "hello" and "dog" to "dog." During this process, the prompt sentence is input into the generative AI model, which then analyzes the data.

[1189] 2. Generate corrections:

[1190] Based on the analysis results, a correction suggestion is generated. The generated correction suggestion is changed to an appropriate expression while maintaining the user's original intent. For example, a correction suggestion might be generated: "Hello, I love dogs!"

[1191] 3. Submitting amendments:

[1192] The generated revision proposal is sent to the user terminal, which displays the revision proposal on a user interface for the user to confirm.

[1193] 4. Reconfirmation after applying the proposed amendment:

[1194] After the user applies the suggested corrections, the server checks them again and posts the finalized text to the social networking platform, for example, "Hello, I love dogs!"

[1195] Specific examples

[1196] Example 1: Correcting typos and omissions

[1197] User: The user types, "Hi, I love dogs!"

[1198] Server: The server generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[1199] User: The user reviews the proposed changes and presses the "Apply" button.

[1200] Server: Post the modified text to the social media platform.

[1201] Prompt Sentence Examples

[1202] Here are some example prompts to input to a generative AI model:

[1203] "Please correct this sentence to correct Japanese: Hello, I love dogs!"

[1204] The present invention allows users to prevent misunderstandings and troubles on SNS and communicate with peace of mind.

[1205] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1206] Step 1:

[1207] The user enters the content to post on SNS.

[1208] The user types "Hello, I love dogs!" into the terminal. The terminal receives this input and stores it as text data.

[1209] Step 2:

[1210] The terminal transmits the text data to the server.

[1211] When the user completes the input and presses the send button, the terminal sends this text data to the server. For example, the text data is sent to the server through an API.

[1212] Step 3:

[1213] The server analyzes the data and detects profanity.

[1214] The server inputs the received text data into a generative AI model to generate a prompt: "Please correct this sentence to correct Japanese: Hello, I love dogs!" The generative AI model analyzes this prompt and detects inappropriate or misleading expressions.

[1215] Step 4:

[1216] The server generates a revision proposal and sends it to the device.

[1217] The server generates a correction suggestion based on the prompt sentence. For example, it generates a correction suggestion such as "Hello, I love dogs!" The generated correction suggestion is sent from the server to the terminal as an HTTP response.

[1218] Step 5:

[1219] The device displays suggested fixes to the user.

[1220] The device displays the suggested corrections received from the server in its user interface, which is presented to the user as a popup or notification. For example, a suggested correction might be "Hello, I love dogs!"

[1221] Step 6:

[1222] The user reviews and applies the proposed fixes.

[1223] The user checks the displayed correction suggestions and presses the "Apply" button. The terminal receives this operation and confirms the corrected text data.

[1224] Step 7:

[1225] The terminal transmits the modified data to the server.

[1226] The device sends the confirmed corrected text "Hello, I love dogs!" again to the server, which receives it for final confirmation.

[1227] Step 8:

[1228] The server performs a final check and posts to the social media platform.

[1229] The server performs a final check of the corrected text data and posts it to various social media platforms via the social media API. As a result, the user's post is published on social media as "Hello, I love dogs!"

[1230] (Application example 1)

[1231] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1232] In traditional brick-and-mortar stores, there is a risk that store staff will use misleading or inappropriate language when dealing with customers, which can lead to lower customer satisfaction and complaints. Furthermore, there was no effective tool for store staff to check and correct appropriate language in real time, making it difficult to communicate promptly on the spot. There is a need to solve the problems that arise from this.

[1233] 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.

[1234] In this invention, the server includes: means for receiving input text; processing means including a generative AI model that analyzes the input text and detects inappropriate or potentially misleading expressions; means for the processing means to analyze voice input from customer interactions in a physical store in real time, determine whether the text contains inappropriate expressions, and suggest appropriate corrections; means for a store clerk to select appropriate expressions based on the suggested corrections; means for suggesting the generated corrections; means for confirming and applying the corrections; and means for posting the confirmed and applied text to various communication platforms. This allows store clerks to confirm and correct appropriate expressions in real time, enabling smooth and secure communication with customers.

[1235] "Means for receiving input text" is a general term for devices and software that provide an interface for transmitting text data input by a user to a server and receiving it.

[1236] "Processing means including a generative AI model that analyzes the input text and detects inappropriate or misleading expressions" is a general term for devices and software that analyze received text data and use natural language processing technology to detect inappropriate or misleading expressions in real time.

[1237] "Means for the processing means to analyze voice inputs made in customer interactions in a physical store in real time, determine whether inappropriate expressions are included, and suggest appropriate corrections" is a general term for devices and software that analyze the voices of store staff in a physical store environment in real time, determine whether inappropriate expressions are included, and suggest corrections to appropriate expressions.

[1238] "Means for allowing the clerk to select an appropriate expression based on the suggested revisions" is a general term for devices and software that present suggested revisions to the clerk visually or audibly and provide an interface that enables the clerk to select an appropriate expression.

[1239] "Means for presenting the generated revision suggestions" is a general term for devices and software for visually or audibly presenting revision suggestions generated by a generative AI model to a user.

[1240] "Means for reviewing and applying the proposed revisions" refers collectively to devices and software that provide an interface through which a user can review the proposed revisions and accept operations to apply them.

[1241] "Means for posting the confirmed and applied text to various communication platforms" is a general term for devices and software that automatically post the corrected text to communication platforms such as SNS.

[1242] This invention relates to a system that enables store clerks to check and correct appropriate expressions in real time when dealing with customers in a physical store. This system allows store clerks to avoid misunderstandings and inappropriate language when communicating with customers, thereby improving service quality.

[1243] System configuration

[1244] The system includes the following major components:

[1245] 1. User Device

[1246] The user terminal is a smartphone or smart glasses used by the store clerk. This terminal provides a voice input interface and receives the clerk's speech in real time.

[1247] The input voice data is sent to a server, and a correction suggestion is received from the server and presented to the store clerk.

[1248] 2. Server

[1249] The server hosts the generative AI model and analyzes the voice data received from the user device.

[1250] Detects inappropriate or potentially misleading content and generates suggested corrections.

[1251] The generated correction proposal is sent to the user's terminal so that the store clerk can check and apply it.

[1252] Server Operation

[1253] 1. Data Reception

[1254] The server receives the voice data transmitted from the user terminal.

[1255] The audio data is converted into text data using natural language processing techniques.

[1256] 2. Data Analysis

[1257] A generative AI model on the server analyzes the text data and detects any inappropriate or potentially misleading expressions.

[1258] If the user says, "This product is very cheap, so you'd be missing out if you didn't buy it!", the system generates a correction suggestion: "This product is a great deal, so please consider buying it."

[1259] 3. Generate correction suggestions

[1260] Generate suggested fixes based on detected issues.

[1261] The generated revision suggestions are modified to appropriate and safe expressions while maintaining the user's original intent.

[1262] 4. Submitting amendments

[1263] The generated revision proposal is sent to the user terminal so that the store clerk can check it.

[1264] The store clerk selects the appropriate expression based on the suggested revisions and speaks to the customer.

[1265] Component Details

[1266] Speech recognition library: The speech recognition library used on the user device is, for example, Google's SpeechRecognition API.

[1267] Generative AI model: For the generative AI model implemented on the server side, we use, for example, OpenAI's GPT-3.

[1268] User Interface: The interface of the user terminal is an interface for presenting the proposed revisions to the user visually or audibly and accepting confirmation and application.

[1269] Communication platform: The system automatically posts the generated revised text to a communication platform (such as a social networking site).

[1270] Specific examples

[1271] Example 1: Voice input and suggested corrections

[1272] User terminal: A salesperson uses smart glasses to tell a customer, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[1273] Server: The server receives the recorded voice data and uses the generation AI to generate a correction suggestion such as, "This product is a great deal, so please consider it."

[1274] User device: The salesperson checks the suggested revisions on the display screen of the smart glasses and communicates them to the customer using appropriate expressions.

[1275] Prompt Sentence Examples

[1276] Example of input prompt: "This product is very cheap, so you'd be missing out if you didn't buy it!" Please change this to an appropriate expression.

[1277] In this way, the present invention enables store staff to communicate appropriately in real time when dealing with customers in a physical store.

[1278] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1279] Step 1:

[1280] The user provides voice input

[1281] Subject: User

[1282] Specific operation: The user (store clerk) wears a smartphone or smart glasses and inputs voice data while talking to a customer. For example, the user might say, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[1283] Input: Voice of store clerk

[1284] Output: Audio data recorded on a smart device

[1285] Step 2:

[1286] Sending audio data

[1287] Subject: Terminal

[1288] Specific operation: The device (smartphone or smart glasses) sends the recorded audio data to the server, which then compresses the data in real time and sends it to the server via a communication line.

[1289] Input: Recorded audio data

[1290] Output: Audio data sent to the server

[1291] Step 3:

[1292] Converting audio data to text

[1293] Subject: Server

[1294] Specific operation: The server uses a speech recognition library (for example, Google's SpeechRecognition API) to convert the received voice data into text data, thereby obtaining the clerk's speech in text format.

[1295] Input: Audio data sent to the server

[1296] Output: Text-formatted speech data

[1297] Step 4:

[1298] Profanity detection

[1299] Subject: Server

[1300] How it works: A generative AI model (such as OpenAI's GPT-3) on the server analyzes text data and detects inappropriate or potentially misleading expressions. For example, it might detect "It's so cheap, you'd be missing out if you didn't buy it."

[1301] Input: Text-formatted speech data

[1302] Output: Text data with inappropriate expressions pointed out

[1303] Step 5:

[1304] Generate correction suggestions

[1305] Subject: Server

[1306] How it works: Based on the detected issues, the server uses a generative AI model to generate appropriate correction suggestions, such as "This product is a great deal, so please consider it."

[1307] Input: Text data containing inappropriate content

[1308] Output: Text data containing suggested revisions

[1309] Step 6:

[1310] Submitting a proposed revision

[1311] Subject: Server

[1312] Specific operation: The server sends the generated revision proposal to the user's device. This transmission is done in real time, so the revision proposal can be viewed instantly on the user's device.

[1313] Input: Text data containing suggested revisions

[1314] Output: The proposed fix sent to the user's device

[1315] Step 7:

[1316] Review and apply the proposed fixes

[1317] Subject: Terminal

[1318] Specific operation: The generated revision suggestions are displayed on the interface of the user device (smart glasses or smartphone). The store clerk can review them and choose whether to adopt the suggestions.

[1319] Input: Proposed correction sent to user device

[1320] Output: Expression data confirmed and applied by the store clerk

[1321] Step 8:

[1322] Communicating the revised statement to customers

[1323] Subject: User

[1324] Specific actions: The salesperson selects the appropriate expression based on the suggested corrections and responds to the customer. For example, they might say, "This product is a great deal, so please consider it."

[1325] Input: Expression data confirmed and applied by the store clerk

[1326] Output: Appropriate response to the customer

[1327] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1328] The present invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[1329] System configuration

[1330] The system includes the following major components:

[1331] 1. User Device

[1332] It provides an interface for users to input content to post on social media.

[1333] The input text and image data are sent to the server.

[1334] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1335] The user is then prompted to confirm and apply the proposed changes.

[1336] 2. Server

[1337] It hosts generative AI models and emotion engines and analyzes data received from users.

[1338] Detect inappropriate or misleading language and generate suggested corrections.

[1339] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1340] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1341] How it works

[1342] User terminal operation

[1343] 1. Input

[1344] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[1345] 2. Data Transmission

[1346] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1347] 3. Receiving and Displaying Proposed Amendments

[1348] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[1349] Server Operation

[1350] 1. Data Analysis

[1351] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[1352] 2. Generate proposed fixes

[1353] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[1354] 3. Emotional Engine Activation

[1355] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[1356] 4. Adjustment of amendments

[1357] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[1358] 5. Submitting amendments

[1359] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[1360] 6. Reconfirmation after application of the proposed amendments

[1361] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[1362] Specific examples

[1363] Example 1: Correcting typos and omissions based on emotions

[1364] User: Type "Hi, I love dogs!"

[1365] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[1366] Device: Presents optimal fix suggestions to the user.

[1367] User: Check the proposed changes and press the "Apply" button.

[1368] Server: Posts the finalized text to various communication platforms.

[1369] Example 2: Modifying provocative expressions and reflecting emotions

[1370] User: Type "Your opinion is completely meaningless!"

[1371] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's angry emotion and generates a milder correction suggestion accordingly.

[1372] Device: Presents optimal fix suggestions to the user.

[1373] User: Check the proposed changes and press the "Apply" button.

[1374] Server: Posts the finalized text to various communication platforms.

[1375] This invention allows users to prevent misunderstandings and troubles on social networking sites, allowing them to communicate with peace of mind. In addition, by combining it with an emotion engine, it is possible to provide more appropriate revision suggestions that take into account the user's emotions.

[1376] The processing flow will be explained below.

[1377] Specific processing steps of the operation

[1378] 1. User Input

[1379] Step 1:

[1380] A user inputs text for posting to an SNS into an input interface of a device. For example, the user inputs "Hello, I love dogs!"

[1381] Step 2:

[1382] When the user has completed the input, he / she presses the "Send" button, causing the terminal to send the text data to the server.

[1383] 2.Data analysis by the server

[1384] Step 3:

[1385] The server passes the received text data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate expressions or parts that may be misleading.

[1386] Step 4:

[1387] The server generates a suggested fix based on the detected problem, for example, "Hi, I love dogs!"

[1388] Step 5:

[1389] After generating the revision suggestions, the server uses an emotion engine to recognize the user's emotions and determine whether the input is based on positive, negative, or neutral emotions.

[1390] Step 6:

[1391] The server further examines the proposed revisions based on the results of the emotion engine and makes adjustments that take into account the user's emotions, for example, softening strong language.

[1392] 3. Present and confirm proposed revisions

[1393] Step 7:

[1394] The server transmits the generated and adjusted correction proposal to the user terminal, for example, "Hello, I love dogs!"

[1395] Step 8:

[1396] The terminal displays the received correction proposals on the user interface in real time. The user checks the proposed correction proposals and selects whether to apply the proposals by pressing the "Apply" button.

[1397] 4. User confirmation and application

[1398] Step 9:

[1399] The user checks the proposed corrections and presses the "Apply" button. If multiple corrections are presented, the user selects the most appropriate one.

[1400] 5. Final submission

[1401] Step 10:

[1402] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[1403] Step 11:

[1404] The server posts the verified and finalized text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[1405] Step 12:

[1406] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[1407] Specific examples

[1408] Example 1: Correcting typos and omissions

[1409] Step 1:

[1410] The user types, "Hi, I love dogs!"

[1411] Step 2:

[1412] The terminal transmits this text data to the server.

[1413] Step 3:

[1414] The server passes the text data to a generative AI model for analysis, which generates suggestions to correct "hello" to "hello" and "wanchan" to "inu."

[1415] Step 4:

[1416] The server applies an emotion engine to the text to recognize positive emotions.

[1417] Step 5:

[1418] The server adjusts the proposed correction "Hello, I love dogs!" based on the emotion and sends it to the user terminal.

[1419] Step 6:

[1420] The terminal presents suggested revisions to the user.

[1421] Step 7:

[1422] The user checks the proposed corrections and presses the "Apply" button.

[1423] Step 8:

[1424] The terminal resends the corrected text to the server.

[1425] Step 9:

[1426] The server posts the confirmed text to the SNS.

[1427] Step 10:

[1428] The terminal and server notify the user that the posting is complete.

[1429] Example 2

[1430] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1431] Posts on conventional social networking services (SNS) often contain inappropriate or potentially misleading language, which can lead to conflicts between users. Furthermore, users often post offensive content influenced by their own emotions, which further contributes to the conflict. To solve these problems, it is necessary to not only properly analyze the content of users' posts and suggest corrections, but also to generate appropriate corrections that take the user's emotions into account.

[1432] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating revision suggestions based on problems detected by the processing means, and means including an emotion engine that recognizes the user's emotions regarding the revision suggestions and generates revision suggestions that take the emotions into consideration. This makes it possible to present not only appropriate revision suggestions for content posted by a user, but also revision suggestions that take the user's emotions into consideration.

[1433] The "means for receiving input text" refers to a device or software having the function of acquiring text data input by the user and incorporating it into the system.

[1434] "Processing means including generative AI models" refers to algorithms or devices that use artificial intelligence to analyze user-entered text and detect inappropriate or potentially misleading language.

[1435] A "means for generating suggested corrections based on problems" is a device or software that has the function of suggesting appropriate correction methods or expressions for inappropriate expressions detected by the generative AI model.

[1436] "Means including an emotion engine" refers to an algorithm or device for recognizing emotions based on user input and generating emotion-sensitive revision suggestions.

[1437] The "means for presenting revision suggestions" is a device or software that has the function of visually displaying the generated revision suggestions to the user.

[1438] "Means for reviewing and applying proposed amendments" means a device or software that has the functionality to allow a user to review proposed amendments and formally adopt the proposed amendments.

[1439] "Means for posting to various communication platforms" refers to devices or software that have the function of posting the text that the user has finally confirmed and applied to a communication platform such as a social networking site.

[1440] The "means for receiving input image data" refers to a device or software having a function for acquiring image data input by a user and incorporating it into the system.

[1441] The "means for presenting multiple revision suggestions and allowing the user to select one" refers to a device or software that has the function of presenting multiple generated revision suggestions to the user and allowing the user to select the most appropriate revision suggestion.

[1442] A "generative AI model" is a model or algorithm that uses artificial intelligence to analyze text data and generate appropriate sentences.

[1443] A "prompt sentence" is an input sentence used to give instructions to a generative AI model, and is specific input data that serves as a reference when analyzing text and generating revision suggestions.

[1444] This invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[1445] System configuration

[1446] The system includes the following major components:

[1447] 1. User Device

[1448] It provides an interface for users to input content to post on social media.

[1449] The input text and image data are sent to the server.

[1450] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1451] The user is then prompted to confirm and apply the proposed changes.

[1452] 2. Server

[1453] It hosts generative AI models and emotion engines and analyzes data received from users.

[1454] Detect inappropriate or misleading language and generate suggested corrections.

[1455] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1456] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1457] User terminal processing

[1458] 1. Input:

[1459] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[1460] 2. Data transmission:

[1461] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1462] 3. Receiving and Viewing Proposed Amendments:

[1463] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[1464] Server Processing

[1465] 1. Data Analysis:

[1466] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[1467] 2. Generate corrections:

[1468] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[1469] 3. Emotional Engine Activation:

[1470] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[1471] 4. Adjustment of amendments:

[1472] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[1473] 5. Submitting amendments:

[1474] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[1475] 6. Reconfirmation after applying the proposed amendment:

[1476] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[1477] Specific examples

[1478] Example 1: Correcting typos and omissions based on emotions

[1479] User: Type "Hi, I love dogs!"

[1480] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[1481] Device: Presents optimal fix suggestions to the user.

[1482] User: Check the proposed changes and press the "Apply" button.

[1483] Server: Posts the finalized text to various communication platforms.

[1484] Example 2: Modifying provocative expressions and reflecting emotions

[1485] User: Type "Your opinion is completely meaningless!"

[1486] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's angry emotion and generates a milder correction suggestion accordingly.

[1487] Device: Presents optimal fix suggestions to the user.

[1488] User: Check the proposed changes and press the "Apply" button.

[1489] Server: Posts the finalized text to various communication platforms.

[1490] Prompt Sentence Examples

[1491] "Please fix the typo and inappropriate language in the following text: 'Hi, I love dogs!'"

[1492] "Please change the inflammatory language in this text to a milder one: 'Your opinion is completely meaningless!'"

[1493] This system allows users to post with peace of mind while avoiding misunderstandings and troubles on social media. Furthermore, by combining it with an emotion engine, it can provide appropriate revision suggestions that take into consideration the user's emotions.

[1494] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1495] Step 1: Input

[1496] User: Enter the content they want to post on the SNS into the user device interface. For example, they enter "Hello, I love dogs!" into the text box.

[1497] Input: Social media post content (text data).

[1498] Output: Input data stored on the user's device.

[1499] Step 2: Send data

[1500] Terminal: When the user presses the send button, the terminal sends the entered text and image data (if any) to the server.

[1501] Input: The post content typed by the user.

[1502] Output: Text and image data sent to the server.

[1503] What it does: Data is divided into packets and sent over the Internet to a server using a communication protocol (e.g. HTTP).

[1504] Step 3: Data analysis

[1505] Server: The server passes the received text and image data to a generative AI model, which analyzes the input and detects inappropriate or potentially misleading language.

[1506] Input: User submissions (text and image data).

[1507] Output: A list of detected issues.

[1508] How it works: The generative AI model runs text analysis algorithms and uses natural language processing techniques to identify inappropriate language. Similarly, for images, it runs image analysis algorithms if they contain inappropriate content.

[1509] Step 4: Generate correction suggestions

[1510] Server: The server generates fixes based on the detected issues. The generative AI model considers the user's intent and proposes optimal fixes.

[1511] Input: A list of detected issues.

[1512] Output: A list of suggested fixes.

[1513] How it works: The generative AI model generates and suggests appropriate expressions based on the training dataset. For example, it generates a correction to "Hello, I love dogs!" in response to "Hello, I love dogs!"

[1514] Step 5: Emotion Recognition

[1515] Server: Runs the emotion engine to recognize emotions from the user's text and image data, for example, determining the emotional tone (positive, negative, neutral) of a sentence or image.

[1516] Input: User submissions (text and image data).

[1517] Output: Sentiment tag (e.g. positive, negative, neutral).

[1518] How it works: The emotion engine uses natural language processing and image analysis to analyze the sentiment of text and images and assign emotional tags.

[1519] Step 6: Adjust the proposed amendment

[1520] Server: Based on the recognition results of the emotion engine, the generated correction proposal is adjusted. If a strong emotion is included, the proposal is changed to a more tolerant expression, for example.

[1521] Input: revision suggestion list and sentiment tags.

[1522] Output: A refined list of proposed fixes.

[1523] What it does: A tuning algorithm is run to fine-tune the generated suggestions to take sentiment into account.

[1524] Step 7: Submit your proposed revisions

[1525] Server: Sends the adjusted correction proposal to the user device.

[1526] Input: The adjusted list of amendments.

[1527] Output: The proposed fix sent to the user's device.

[1528] Specific operation: The proposed revision data is packetized and transmitted to the user terminal via the Internet.

[1529] Step 8: Receive and view proposed revisions

[1530] Terminal: The proposed modifications sent from the server are displayed on the user interface. The user can review the proposed modifications and decide whether to apply them.

[1531] Input: The proposed fix sent by the server.

[1532] Output: The suggested fixes displayed in the user interface.

[1533] Specific operation: Retrieves data from the receiving buffer and reflects the received correction suggestions in the UI component.

[1534] Step 9: Review and apply proposed fixes

[1535] User: Review the proposed fix and click the "Apply" button to officially apply the fix.

[1536] Input: The suggested fix.

[1537] Output: Confirmation and applied text.

[1538] What happens: Once the proposed fix is ​​applied, a confirmation dialog will appear, allowing the user to make a final confirmation.

[1539] Step 10: Post

[1540] Server: After the user applies the suggested corrections, it checks them again and posts the finalized text to the social media platform.

[1541] Input: Confirmed and applied text.

[1542] Output: The text posted to the social media platform.

[1543] Specific operation: The confirmed text data is posted via the SNS API.

[1544] (Application example 2)

[1545] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1546] On social media and other online platforms, users often unconsciously use inappropriate or potentially misleading language. Furthermore, emotional comments made by users often lead to trouble and misunderstandings. A system to prevent such problems is needed.

[1547] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating correction suggestions based on the detected problems, and means including an emotion recognition engine that recognizes the user's emotions regarding the generated correction suggestions and presents correction suggestions that take the emotions into consideration. This makes it possible to correct inappropriate expressions or misleading expressions while taking the user's emotions into consideration.

[1548] The "means for receiving input text" refers to a device or software that has the function of acquiring text data input by the user and transmitting it to the system.

[1549] "Processing means including a generative AI model that analyzes and detects inappropriate or potentially misleading language" refers to a system function that includes an AI algorithm that uses natural language processing techniques to analyze input text and identify inappropriate or potentially misleading language.

[1550] The "means for generating proposed fixes based on detected problems" is a processing function for creating alternatives to resolve problems found by the analysis. The means utilizes a generative AI model to automatically generate proposed fixes.

[1551] The "means including an emotion recognition engine" is a processing system for analyzing emotions from user input and providing optimal revision suggestions based on the emotion data.

[1552] "Means for confirming and applying proposed revisions" refers to a function that provides an operational interface for users to confirm revisions proposed by the system and officially select and apply those revisions.

[1553] "Means for posting the confirmed and applied text to various communication platforms" refers to a processing means for sending and posting the amendments selected and applied by the user to the relevant social networking site or other online communication tool.

[1554] This invention is a system that analyzes content posted by users on social networking sites, detects inappropriate or potentially misleading expressions, and suggests corrections. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate corrections. Specific embodiments and their operation are described below.

[1555] System configuration

[1556] The system includes the following major components:

[1557] User terminal

[1558] It provides an interface for users to input content to post on social media.

[1559] The input text and image data are sent to the server.

[1560] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1561] The user is then prompted to confirm and apply the proposed changes.

[1562] server

[1563] It hosts generative AI models and emotion engines and analyzes data received from users.

[1564] Detect inappropriate or misleading language and generate suggested corrections.

[1565] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1566] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1567] How it works

[1568] 1. Data Entry

[1569] The user enters the content of the SNS post into the user device interface. For example, they might enter, "This app is really useless!"

[1570] 2. Data Transmission

[1571] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1572] 3. Data Analysis

[1573] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[1574] 4. Generate correction suggestions

[1575] Based on the detected issues, the generative AI model generates the most appropriate correction suggestions while taking into account the user's intent. For example, it generates a correction suggestion such as, "This app still has room for improvement."

[1576] 5. Emotion recognition

[1577] The server activates an emotion recognition engine to recognize emotions from the user's input and image data. For example, if the input text expresses anger, the server generates a revision suggestion that takes this into consideration.

[1578] 6. Adjustment of amendments

[1579] Based on the emotions recognized by the emotion recognition engine, the proposed corrections are further refined and adjusted to take the user's emotions into consideration. For example, if anger is recognized, the expression will be changed to a calmer one.

[1580] 7. Proposal of amendments

[1581] The generated proposed corrections are sent to the user's terminal for the user to review. The user can review the proposed corrections and apply them by pressing the "Apply" button.

[1582] 8. Confirm and post the applied text

[1583] After the user confirms and applies the proposed changes, the finalized text is posted to various communication platforms. For example, the revised version of "This app still has room for improvement, but it has helped me in some ways" can be safely posted on social media.

[1584] Hardware and software used

[1585] Smartphone

[1586] Application execution environment

[1587] Mobile data transmission and reception

[1588] server

[1589] AWS (cloud server)

[1590] Google Cloud (cloud server)

[1591] Hosting generative AI models

[1592] Emotion Engine Hosting

[1593] Generative AI Models

[1594] GPT-4 and other modern natural language processing (NLP) models

[1595] Custom NLP model for profanity detection

[1596] Emotion Recognition Engine

[1597] Emotion recognition model created using TensorFlow and Keras

[1598] Specific examples

[1599] User: Type "This app is completely useless!"

[1600] Generative AI model: Generates suggested fixes such as, "This app still has room for improvement."

[1601] Emotion recognition engine: Recognizes the user's angry emotion and generates adjustment suggestions such as, "This app still has room for improvement, but it has helped in some ways."

[1602] User device: Check the suggested fixes and press the "Apply" button.

[1603] Server: Posts the confirmed text to the social networking site.

[1604] Example prompts to input to a generative AI model:

[1605] Original: "This app is completely useless!"

[1606] Generative AI: "This app still has room for improvement."

[1607] Emotion Engine: Recognizes the user's angry emotion and generates adjustment suggestions such as "This app still has room for improvement, but it has helped in some ways."

[1608] This makes it possible to take into consideration the user's feelings and prevent misunderstandings and problems in online communication.

[1609] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1610] Step 1:

[1611] The user enters the content of a social media post into the interface of the user device. The entered text is, "This app is really useless!" This input data is sent to the server in the next step.

[1612] Step 2:

[1613] The user terminal sends the input text data to the cloud server. Here, the mobile data is sent using a smartphone, and the cloud server receives the data.

[1614] Step 3:

[1615] The server passes the received text data to a generative AI model (an NLP model such as GPT-4) to begin analysis. The generative AI model analyzes the input text, "This app is really useless!", and detects inappropriate or potentially misleading expressions. In this step, text data analysis and inappropriate expression detection are performed.

[1616] Step 4:

[1617] The generative AI model generates a suggested fix based on the detected issues, for example, "This app still has room for improvement." The server then processes this suggested fix in the next step.

[1618] Step 5:

[1619] The server starts an emotion recognition engine (a model using TensorFlow or Keras) and analyzes the emotion from the user's input text. For example, it recognizes that the input text contains the emotion of anger. In this step, the emotion data is analyzed.

[1620] Step 6:

[1621] The server further refines the generated revision suggestions based on the emotional data analyzed by the emotion recognition engine. For example, it generates an optimal revision suggestion such as, "This app still has room for improvement, but it was helpful in some ways." In this step, data processing is performed taking emotions into consideration.

[1622] Step 7:

[1623] The server sends the final proposed fixes to the user's device, which then displays them for review. For example, the user might see a suggestion like, "This app still has room for improvement, but it helped in some ways."

[1624] Step 8:

[1625] The user checks the proposed revisions and applies them by pressing the "Apply" button. Here, the user selects a revision and the selection is sent from the terminal to the server.

[1626] Step 9:

[1627] The server receives the verified and applied text and posts it to various communication platforms. For example, the revised text "This app still has room for improvement, but it has helped me in some ways" is posted to social media. This step is the final data transmission.

[1628] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1629] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1630] 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.

[1631] [Fourth embodiment]

[1632] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1633] 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.

[1634] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1635] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1636] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1637] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1638] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1639] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1640] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1641] 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.

[1642] 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.

[1643] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1644] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1645] The present invention provides a system that detects inappropriate expressions or potentially misleading parts of social media posts entered by users in real time and suggests corrections. Specific embodiments and their operation are described below.

[1646] System configuration

[1647] The system includes the following major components:

[1648] 1. User Device

[1649] It provides an interface for users to input content for posting on social media.

[1650] The input text and image data are sent to the server.

[1651] Receives suggested revisions from the server and presents them to the user.

[1652] The user is then prompted to confirm and apply the proposed changes.

[1653] 2. Server

[1654] It hosts generative AI models and analyzes data received from user devices.

[1655] Detects inappropriate or potentially misleading language and generates suggested corrections.

[1656] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1657] How it works

[1658] User terminal operation

[1659] 1. Input

[1660] The user inputs the content of the SNS post into the interface of the user device. For example, the user might input "Hello, I love dogs!"

[1661] 2. Data Transmission

[1662] Once the input is complete, the user terminal sends this text data to the server. If an image is included, it is also sent.

[1663] 3. Receiving and Displaying Proposed Amendments

[1664] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[1665] Server Operation

[1666] 1. Data Analysis

[1667] It analyzes incoming text and image data and uses generative AI models to detect inappropriate or potentially misleading content, suggesting, for example, changing "hello" to "hello" or "dog" to "dog."

[1668] 2. Generate proposed fixes

[1669] Based on the detected problems, a correction suggestion is generated, which is a correction to an appropriate and safe expression while preserving the user's original intent as much as possible.

[1670] 3. Submitting amendments

[1671] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[1672] 4. Reconfirmation after application of the proposed amendments

[1673] After the user confirms and applies the proposed edits, they can review the changes and post the finalized text to various communication platforms. For example, the revised text "Hello, I love dogs!" can be safely posted to social media.

[1674] Specific examples

[1675] Example 1: Correcting typos and omissions

[1676] User: Type "Hi, I love dogs!"

[1677] Server: Generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[1678] User: Check the proposed changes and press the "Apply" button.

[1679] Server: Posts the modified text to various communication platforms.

[1680] Example 2: Correcting provocative language

[1681] User: Type "Your opinion is completely meaningless!"

[1682] Server: Generates a correction suggestion saying "I have doubts about your opinion" and sends it to the user's terminal.

[1683] User: Check the proposed changes and press the "Apply" button.

[1684] Server: Posts the modified text to various communication platforms.

[1685] The present invention enables users to prevent misunderstandings and troubles on SNS and to communicate with peace of mind.

[1686] The processing flow will be explained below.

[1687] Step 1:

[1688] The user inputs the content of the SNS post into the input interface of the device. For example, the user inputs "Hello, I love dogs!"

[1689] Step 2:

[1690] The terminal transmits the input text data to the server. If image data is included, the image data is also transmitted.

[1691] Step 3:

[1692] The server passes the received data to the generative AI model to begin analysis, which then analyzes the text to detect inappropriate or potentially misleading expressions.

[1693] Step 4:

[1694] The server generates suggested corrections based on the detected issues. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[1695] Step 5:

[1696] The server transmits the generated revision proposal to the user terminal, which displays the received revision proposal on its interface.

[1697] Step 6:

[1698] The user reviews the proposed revisions. If they are appropriate, they press the "Apply" button. If there are multiple revision options, the user selects the most appropriate one.

[1699] Step 7:

[1700] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[1701] Step 8:

[1702] The server posts the verified and applied text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[1703] Step 9:

[1704] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[1705] Example 1

[1706] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1707] It is necessary to prevent problems and misunderstandings that may arise from the use of inappropriate or potentially misleading expressions on communication platforms such as social networking sites, and to provide an environment where users can communicate with peace of mind.

[1708] 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.

[1709] In this invention, the server includes a processing means including a generative AI model that analyzes input text and detects inappropriate or potentially misleading expressions, a means for generating suggested revisions based on the detected problems, and a means for displaying and presenting the generated suggested revisions on a user interface, thereby enabling users to correct their SNS posts and prevent problems or misunderstandings from occurring.

[1710] "Means for receiving input text" refers to the interface that allows users to input social media posts and text data into the system, as well as the technology for receiving such data.

[1711] "Processing means including a generative AI model that analyzes input text" refers to generative AI models and related analytical technologies in general that analyze text data entered by a user and detect inappropriate expressions or parts that may be misleading.

[1712] The "means for generating correction suggestions" refers to a general technology that generates suggestions for correcting inappropriate expressions into appropriate expressions while maintaining the user's original intent as much as possible, based on the detected problems.

[1713] The "means for presenting the generated revision proposal" refers to the interface and technology in general for displaying the revision proposal sent from the server to the user terminal in an easy-to-understand manner for the user.

[1714] "Means for reviewing and applying suggested fixes" refers to the general interface and technology that allows a user to review the suggested fixes and accept operations to actually apply them.

[1715] "Means for posting the confirmed and applied text to various communication platforms" refers to all technologies that allow users to safely post the revised text to various social networking sites and communication platforms.

[1716] "Means for a user terminal to display suggested revisions on a user interface" refers to the general interface and technology that displays suggested revisions sent from a server on a user terminal and allows the user to confirm them.

[1717] "Means for the server to perform final check" refers to the general technology by which the system checks the contents again and performs a final check after the user applies the proposed correction.

[1718] The present invention is a system that analyzes text and images that users are about to post on social networking sites in advance, detects inappropriate expressions or parts that may be misleading, and suggests corrections. The system is composed of a user terminal and a server.

[1719] System configuration and operation

[1720] User terminal

[1721] A user terminal is a device that provides an interface for users to input content for posting on social media. This includes smartphones, tablets, and PCs. For example, if a user types "Hello, I love dogs!", this content is sent to the server via the terminal. The user terminal has a means to receive the input text and immediately send it to the server.

[1722] server

[1723] The server is responsible for analyzing the input data received from the user device. It uses a generative AI model for analysis. Specifically, the server performs the following processes:

[1724] 1. Data Analysis:

[1725] The server uses a generative AI model to analyze text and image data sent from the user's device. For example, it suggests correcting "hello" to "hello" and "dog" to "dog." During this process, the prompt sentence is input into the generative AI model, which then analyzes the data.

[1726] 2. Generate corrections:

[1727] Based on the analysis results, a correction suggestion is generated. The generated correction suggestion is changed to an appropriate expression while maintaining the user's original intent. For example, a correction suggestion might be generated: "Hello, I love dogs!"

[1728] 3. Submitting amendments:

[1729] The generated revision proposal is sent to the user terminal, which displays the revision proposal on a user interface for the user to confirm.

[1730] 4. Reconfirmation after applying the proposed amendment:

[1731] After the user applies the suggested corrections, the server checks them again and posts the finalized text to the social networking platform, for example, "Hello, I love dogs!"

[1732] Specific examples

[1733] Example 1: Correcting typos and omissions

[1734] User: The user types, "Hi, I love dogs!"

[1735] Server: The server generates a correction suggestion, "Hello, I love dogs!", and sends it to the user device.

[1736] User: The user reviews the proposed changes and presses the "Apply" button.

[1737] Server: Post the modified text to the social media platform.

[1738] Prompt Sentence Examples

[1739] Here are some example prompts to input to a generative AI model:

[1740] "Please correct this sentence to correct Japanese: Hello, I love dogs!"

[1741] The present invention allows users to prevent misunderstandings and troubles on SNS and communicate with peace of mind.

[1742] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1743] Step 1:

[1744] The user enters the content to post on SNS.

[1745] The user types "Hello, I love dogs!" into the terminal. The terminal receives this input and stores it as text data.

[1746] Step 2:

[1747] The terminal transmits the text data to the server.

[1748] When the user completes the input and presses the send button, the terminal sends this text data to the server. For example, the text data is sent to the server through an API.

[1749] Step 3:

[1750] The server analyzes the data and detects profanity.

[1751] The server inputs the received text data into a generative AI model to generate a prompt: "Please correct this sentence to correct Japanese: Hello, I love dogs!" The generative AI model analyzes this prompt and detects inappropriate or misleading expressions.

[1752] Step 4:

[1753] The server generates a revision proposal and sends it to the device.

[1754] The server generates a correction suggestion based on the prompt sentence. For example, it generates a correction suggestion such as "Hello, I love dogs!" The generated correction suggestion is sent from the server to the terminal as an HTTP response.

[1755] Step 5:

[1756] The device displays suggested fixes to the user.

[1757] The device displays the suggested corrections received from the server in its user interface, which is presented to the user as a popup or notification. For example, a suggested correction might be "Hello, I love dogs!"

[1758] Step 6:

[1759] The user reviews and applies the proposed fixes.

[1760] The user checks the displayed correction suggestions and presses the "Apply" button. The terminal receives this operation and confirms the corrected text data.

[1761] Step 7:

[1762] The terminal transmits the modified data to the server.

[1763] The device sends the confirmed corrected text "Hello, I love dogs!" again to the server, which receives it for final confirmation.

[1764] Step 8:

[1765] The server performs a final check and posts to the social media platform.

[1766] The server performs a final check of the corrected text data and posts it to various social media platforms via the social media API. As a result, the user's post is published on social media as "Hello, I love dogs!"

[1767] (Application example 1)

[1768] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1769] In traditional brick-and-mortar stores, there is a risk that store staff will use misleading or inappropriate language when dealing with customers, which can lead to lower customer satisfaction and complaints. Furthermore, there was no effective tool for store staff to check and correct appropriate language in real time, making it difficult to communicate promptly on the spot. There is a need to solve the problems that arise from this.

[1770] 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.

[1771] In this invention, the server includes: means for receiving input text; processing means including a generative AI model that analyzes the input text and detects inappropriate or potentially misleading expressions; means for the processing means to analyze voice input from customer interactions in a physical store in real time, determine whether the text contains inappropriate expressions, and suggest appropriate corrections; means for a store clerk to select appropriate expressions based on the suggested corrections; means for suggesting the generated corrections; means for confirming and applying the corrections; and means for posting the confirmed and applied text to various communication platforms. This allows store clerks to confirm and correct appropriate expressions in real time, enabling smooth and secure communication with customers.

[1772] "Means for receiving input text" is a general term for devices and software that provide an interface for transmitting text data input by a user to a server and receiving it.

[1773] "Processing means including a generative AI model that analyzes the input text and detects inappropriate or misleading expressions" is a general term for devices and software that analyze received text data and use natural language processing technology to detect inappropriate or misleading expressions in real time.

[1774] "Means for the processing means to analyze voice inputs made in customer interactions in a physical store in real time, determine whether inappropriate expressions are included, and suggest appropriate corrections" is a general term for devices and software that analyze the voices of store staff in a physical store environment in real time, determine whether inappropriate expressions are included, and suggest corrections to appropriate expressions.

[1775] "Means for allowing the clerk to select an appropriate expression based on the suggested revisions" is a general term for devices and software that present suggested revisions to the clerk visually or audibly and provide an interface that enables the clerk to select an appropriate expression.

[1776] "Means for presenting the generated revision suggestions" is a general term for devices and software for visually or audibly presenting revision suggestions generated by a generative AI model to a user.

[1777] "Means for reviewing and applying the proposed revisions" refers collectively to devices and software that provide an interface through which a user can review the proposed revisions and accept operations to apply them.

[1778] "Means for posting the confirmed and applied text to various communication platforms" is a general term for devices and software that automatically post the corrected text to communication platforms such as SNS.

[1779] This invention relates to a system that enables store clerks to check and correct appropriate expressions in real time when dealing with customers in a physical store. This system allows store clerks to avoid misunderstandings and inappropriate language when communicating with customers, thereby improving service quality.

[1780] System configuration

[1781] The system includes the following major components:

[1782] 1. User Device

[1783] The user terminal is a smartphone or smart glasses used by the store clerk. This terminal provides a voice input interface and receives the clerk's speech in real time.

[1784] The input voice data is sent to a server, and a correction suggestion is received from the server and presented to the store clerk.

[1785] 2. Server

[1786] The server hosts the generative AI model and analyzes the voice data received from the user device.

[1787] Detects inappropriate or potentially misleading content and generates suggested corrections.

[1788] The generated correction proposal is sent to the user's terminal so that the store clerk can check and apply it.

[1789] Server Operation

[1790] 1. Data Reception

[1791] The server receives the voice data transmitted from the user terminal.

[1792] The audio data is converted into text data using natural language processing techniques.

[1793] 2. Data Analysis

[1794] A generative AI model on the server analyzes the text data and detects any inappropriate or potentially misleading expressions.

[1795] If the user says, "This product is very cheap, so you'd be missing out if you didn't buy it!", the system generates a correction suggestion: "This product is a great deal, so please consider buying it."

[1796] 3. Generate correction suggestions

[1797] Generate suggested fixes based on detected issues.

[1798] The generated revision suggestions are modified to appropriate and safe expressions while maintaining the user's original intent.

[1799] 4. Submitting amendments

[1800] The generated revision proposal is sent to the user terminal so that the store clerk can check it.

[1801] The store clerk selects the appropriate expression based on the suggested revisions and speaks to the customer.

[1802] Component Details

[1803] Speech recognition library: The speech recognition library used on the user device is, for example, Google's SpeechRecognition API.

[1804] Generative AI model: For the generative AI model implemented on the server side, we use, for example, OpenAI's GPT-3.

[1805] User Interface: The interface of the user terminal is an interface for presenting the proposed revisions to the user visually or audibly and accepting confirmation and application.

[1806] Communication platform: The system automatically posts the generated revised text to a communication platform (such as a social networking site).

[1807] Specific examples

[1808] Example 1: Voice input and suggested corrections

[1809] User terminal: A salesperson uses smart glasses to tell a customer, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[1810] Server: The server receives the recorded voice data and uses the generation AI to generate a correction suggestion such as, "This product is a great deal, so please consider it."

[1811] User device: The salesperson checks the suggested revisions on the display screen of the smart glasses and communicates them to the customer using appropriate expressions.

[1812] Prompt Sentence Examples

[1813] Example of input prompt: "This product is very cheap, so you'd be missing out if you didn't buy it!" Please change this to an appropriate expression.

[1814] In this way, the present invention enables store staff to communicate appropriately in real time when dealing with customers in a physical store.

[1815] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1816] Step 1:

[1817] The user provides voice input

[1818] Subject: User

[1819] Specific operation: The user (store clerk) wears a smartphone or smart glasses and inputs voice data while talking to a customer. For example, the user might say, "This product is very cheap, so you'd be missing out if you didn't buy it!"

[1820] Input: Voice of store clerk

[1821] Output: Audio data recorded on a smart device

[1822] Step 2:

[1823] Sending audio data

[1824] Subject: Terminal

[1825] Specific operation: The device (smartphone or smart glasses) sends the recorded audio data to the server, which then compresses the data in real time and sends it to the server via a communication line.

[1826] Input: Recorded audio data

[1827] Output: Audio data sent to the server

[1828] Step 3:

[1829] Converting audio data to text

[1830] Subject: Server

[1831] Specific operation: The server uses a speech recognition library (for example, Google's SpeechRecognition API) to convert the received voice data into text data, thereby obtaining the clerk's speech in text format.

[1832] Input: Audio data sent to the server

[1833] Output: Text-formatted speech data

[1834] Step 4:

[1835] Profanity detection

[1836] Subject: Server

[1837] How it works: A generative AI model (such as OpenAI's GPT-3) on the server analyzes text data and detects inappropriate or potentially misleading expressions. For example, it might detect "It's so cheap, you'd be missing out if you didn't buy it."

[1838] Input: Text-formatted speech data

[1839] Output: Text data with inappropriate expressions pointed out

[1840] Step 5:

[1841] Generate correction suggestions

[1842] Subject: Server

[1843] How it works: Based on the detected issues, the server uses a generative AI model to generate appropriate correction suggestions, such as "This product is a great deal, so please consider it."

[1844] Input: Text data containing inappropriate content

[1845] Output: Text data containing suggested revisions

[1846] Step 6:

[1847] Submitting a proposed revision

[1848] Subject: Server

[1849] Specific operation: The server sends the generated revision proposal to the user's device. This transmission is done in real time, so the revision proposal can be viewed instantly on the user's device.

[1850] Input: Text data containing suggested revisions

[1851] Output: The proposed fix sent to the user's device

[1852] Step 7:

[1853] Review and apply the proposed fixes

[1854] Subject: Terminal

[1855] Specific operation: The generated revision suggestions are displayed on the interface of the user device (smart glasses or smartphone). The store clerk can review them and choose whether to adopt the suggestions.

[1856] Input: Proposed correction sent to user device

[1857] Output: Expression data confirmed and applied by the store clerk

[1858] Step 8:

[1859] Communicating the revised statement to customers

[1860] Subject: User

[1861] Specific actions: The salesperson selects the appropriate expression based on the suggested corrections and responds to the customer. For example, they might say, "This product is a great deal, so please consider it."

[1862] Input: Expression data confirmed and applied by the store clerk

[1863] Output: Appropriate response to the customer

[1864] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1865] The present invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[1866] System configuration

[1867] The system includes the following major components:

[1868] 1. User Device

[1869] It provides an interface for users to input content to post on social media.

[1870] The input text and image data are sent to the server.

[1871] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1872] The user is then prompted to confirm and apply the proposed changes.

[1873] 2. Server

[1874] It hosts generative AI models and emotion engines and analyzes data received from users.

[1875] Detect inappropriate or misleading language and generate suggested corrections.

[1876] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1877] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1878] How it works

[1879] User terminal operation

[1880] 1. Input

[1881] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[1882] 2. Data Transmission

[1883] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1884] 3. Receiving and Displaying Proposed Amendments

[1885] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[1886] Server Operation

[1887] 1. Data Analysis

[1888] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[1889] 2. Generate proposed fixes

[1890] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[1891] 3. Emotional Engine Activation

[1892] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[1893] 4. Adjustment of amendments

[1894] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[1895] 5. Submitting amendments

[1896] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[1897] 6. Reconfirmation after application of the proposed amendments

[1898] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[1899] Specific examples

[1900] Example 1: Correcting typos and omissions based on emotions

[1901] User: Type "Hi, I love dogs!"

[1902] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[1903] Device: Presents optimal fix suggestions to the user.

[1904] User: Check the proposed changes and press the "Apply" button.

[1905] Server: Posts the finalized text to various communication platforms.

[1906] Example 2: Modifying provocative expressions and reflecting emotions

[1907] User: Type "Your opinion is completely meaningless!"

[1908] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's angry emotion and generates a milder correction suggestion accordingly.

[1909] Device: Presents optimal fix suggestions to the user.

[1910] User: Check the proposed changes and press the "Apply" button.

[1911] Server: Posts the finalized text to various communication platforms.

[1912] This invention allows users to prevent misunderstandings and troubles on social networking sites, allowing them to communicate with peace of mind. In addition, by combining it with an emotion engine, it is possible to provide more appropriate revision suggestions that take into account the user's emotions.

[1913] The processing flow will be explained below.

[1914] Specific processing steps of the operation

[1915] 1. User Input

[1916] Step 1:

[1917] A user inputs text for posting to an SNS into an input interface of a device. For example, the user inputs "Hello, I love dogs!"

[1918] Step 2:

[1919] When the user has completed the input, he / she presses the "Send" button, causing the terminal to send the text data to the server.

[1920] 2.Data analysis by the server

[1921] Step 3:

[1922] The server passes the received text data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate expressions or parts that may be misleading.

[1923] Step 4:

[1924] The server generates a suggested fix based on the detected problem, for example, "Hi, I love dogs!"

[1925] Step 5:

[1926] After generating the revision suggestions, the server uses an emotion engine to recognize the user's emotions and determine whether the input is based on positive, negative, or neutral emotions.

[1927] Step 6:

[1928] The server further examines the proposed revisions based on the results of the emotion engine and makes adjustments that take into account the user's emotions, for example, softening strong language.

[1929] 3. Present and confirm proposed revisions

[1930] Step 7:

[1931] The server transmits the generated and adjusted correction proposal to the user terminal, for example, "Hello, I love dogs!"

[1932] Step 8:

[1933] The terminal displays the received correction proposals on the user interface in real time. The user checks the proposed correction proposals and selects whether to apply the proposals by pressing the "Apply" button.

[1934] 4. User confirmation and application

[1935] Step 9:

[1936] The user checks the proposed corrections and presses the "Apply" button. If multiple corrections are presented, the user selects the most appropriate one.

[1937] 5. Final submission

[1938] Step 10:

[1939] The device resends the suggested corrections selected and applied by the user to the server, and the corrected text data is returned to the server.

[1940] Step 11:

[1941] The server posts the verified and finalized text via the API of various communication platforms. For example, "Hello, I love dogs!" is posted to SNS.

[1942] Step 12:

[1943] The device and server notify the user that the post has been successfully completed, allowing the user to confirm that the revised post has been successfully updated on the SNS.

[1944] Specific examples

[1945] Example 1: Correcting typos and omissions

[1946] Step 1:

[1947] The user types, "Hi, I love dogs!"

[1948] Step 2:

[1949] The terminal transmits this text data to the server.

[1950] Step 3:

[1951] The server passes the text data to a generative AI model for analysis, which generates suggestions to correct "hello" to "hello" and "wanchan" to "inu."

[1952] Step 4:

[1953] The server applies an emotion engine to the text to recognize positive emotions.

[1954] Step 5:

[1955] The server adjusts the proposed correction "Hello, I love dogs!" based on the emotion and sends it to the user terminal.

[1956] Step 6:

[1957] The terminal presents suggested revisions to the user.

[1958] Step 7:

[1959] The user checks the proposed corrections and presses the "Apply" button.

[1960] Step 8:

[1961] The terminal resends the corrected text to the server.

[1962] Step 9:

[1963] The server posts the confirmed text to the SNS.

[1964] Step 10:

[1965] The terminal and server notify the user that the posting is complete.

[1966] Example 2

[1967] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1968] Posts on conventional social networking services (SNS) often contain inappropriate or potentially misleading language, which can lead to conflicts between users. Furthermore, users often post offensive content influenced by their own emotions, which further contributes to the conflict. To solve these problems, it is necessary to not only properly analyze the content of users' posts and suggest corrections, but also to generate appropriate corrections that take the user's emotions into account.

[1969] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating revision suggestions based on problems detected by the processing means, and means including an emotion engine that recognizes the user's emotions regarding the revision suggestions and generates revision suggestions that take the emotions into consideration. This makes it possible to present not only appropriate revision suggestions for content posted by a user, but also revision suggestions that take the user's emotions into consideration.

[1970] The "means for receiving input text" refers to a device or software having the function of acquiring text data input by the user and incorporating it into the system.

[1971] "Processing means including generative AI models" refers to algorithms or devices that use artificial intelligence to analyze user-entered text and detect inappropriate or potentially misleading language.

[1972] A "means for generating suggested corrections based on problems" is a device or software that has the function of suggesting appropriate correction methods or expressions for inappropriate expressions detected by the generative AI model.

[1973] "Means including an emotion engine" refers to an algorithm or device for recognizing emotions based on user input and generating emotion-sensitive revision suggestions.

[1974] The "means for presenting revision suggestions" is a device or software that has the function of visually displaying the generated revision suggestions to the user.

[1975] "Means for reviewing and applying proposed amendments" means a device or software that has the functionality to allow a user to review proposed amendments and formally adopt the proposed amendments.

[1976] "Means for posting to various communication platforms" refers to devices or software that have the function of posting the text that the user has finally confirmed and applied to a communication platform such as a social networking site.

[1977] The "means for receiving input image data" refers to a device or software having a function for acquiring image data input by a user and incorporating it into the system.

[1978] The "means for presenting multiple revision suggestions and allowing the user to select one" refers to a device or software that has the function of presenting multiple generated revision suggestions to the user and allowing the user to select the most appropriate revision suggestion.

[1979] A "generative AI model" is a model or algorithm that uses artificial intelligence to analyze text data and generate appropriate sentences.

[1980] A "prompt sentence" is an input sentence used to give instructions to a generative AI model, and is specific input data that serves as a reference when analyzing text and generating revision suggestions.

[1981] This invention relates to a system that analyzes content posted by users on social media, detects inappropriate or potentially misleading expressions using a generative AI model, and suggests revisions. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate revisions. Specific embodiments and their operation are described below.

[1982] System configuration

[1983] The system includes the following major components:

[1984] 1. User Device

[1985] It provides an interface for users to input content to post on social media.

[1986] The input text and image data are sent to the server.

[1987] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[1988] The user is then prompted to confirm and apply the proposed changes.

[1989] 2. Server

[1990] It hosts generative AI models and emotion engines and analyzes data received from users.

[1991] Detect inappropriate or misleading language and generate suggested corrections.

[1992] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[1993] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[1994] User terminal processing

[1995] 1. Input:

[1996] The user enters the content of the SNS post into the interface of the user device. For example, the user enters "Hello, I love dogs!"

[1997] 2. Data transmission:

[1998] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[1999] 3. Receiving and Viewing Proposed Amendments:

[2000] The analysis results and suggested modifications are received from the server and displayed on the user interface. The user can check the suggested modifications and apply them by pressing the "Apply" button.

[2001] Server Processing

[2002] 1. Data Analysis:

[2003] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[2004] 2. Generate corrections:

[2005] Based on the detected issues, a suggested correction is generated. The generative AI model considers the user's intent and creates the most appropriate correction. For example, it generates a suggestion like "Hello, I love dogs!"

[2006] 3. Emotional Engine Activation:

[2007] The server activates an emotion engine to recognize emotions from the user's input and image data. For example, if the input text has a positive emotion, it generates a revision suggestion that takes this into consideration.

[2008] 4. Adjustment of amendments:

[2009] Based on the emotions identified by the emotion engine, the proposed revisions are further refined and adjusted to better reflect the user's emotions, for example, by softening strong language.

[2010] 5. Submitting amendments:

[2011] The generated revision proposal is sent to the user terminal so that the user can confirm it.

[2012] 6. Reconfirmation after applying the proposed amendment:

[2013] After users confirm and apply the proposed edits, they can review them again and post the finalized text to various communication platforms. For example, the revised version of "Hello, I love dogs!" can be safely posted to social media.

[2014] Specific examples

[2015] Example 1: Correcting typos and omissions based on emotions

[2016] User: Type "Hi, I love dogs!"

[2017] Server: The generative AI generates a correction suggestion such as "Hi, I love dogs!", and the emotion engine recognizes the user's positive emotions and adjusts the correction suggestion accordingly.

[2018] Device: Presents optimal fix suggestions to the user.

[2019] User: Check the proposed changes and press the "Apply" button.

[2020] Server: Posts the finalized text to various communication platforms.

[2021] Example 2: Modifying provocative expressions and reflecting emotions

[2022] User: Type "Your opinion is completely meaningless!"

[2023] Server: The generative AI generates a correction suggestion such as "I doubt your opinion," and the emotion engine recognizes the user's angry emotion and generates a milder correction suggestion accordingly.

[2024] Device: Presents optimal fix suggestions to the user.

[2025] User: Check the proposed changes and press the "Apply" button.

[2026] Server: Posts the finalized text to various communication platforms.

[2027] Prompt Sentence Examples

[2028] "Please fix the typo and inappropriate language in the following text: 'Hi, I love dogs!'"

[2029] "Please change the inflammatory language in this text to a milder one: 'Your opinion is completely meaningless!'"

[2030] This system allows users to post with peace of mind while avoiding misunderstandings and troubles on social media. Furthermore, by combining it with an emotion engine, it can provide appropriate revision suggestions that take into consideration the user's emotions.

[2031] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2032] Step 1: Input

[2033] User: Enter the content they want to post on the SNS into the user device interface. For example, they enter "Hello, I love dogs!" into the text box.

[2034] Input: Social media post content (text data).

[2035] Output: Input data stored on the user's device.

[2036] Step 2: Send data

[2037] Terminal: When the user presses the send button, the terminal sends the entered text and image data (if any) to the server.

[2038] Input: The post content typed by the user.

[2039] Output: Text and image data sent to the server.

[2040] What it does: Data is divided into packets and sent over the Internet to a server using a communication protocol (e.g. HTTP).

[2041] Step 3: Data analysis

[2042] Server: The server passes the received text and image data to a generative AI model, which analyzes the input and detects inappropriate or potentially misleading language.

[2043] Input: User submissions (text and image data).

[2044] Output: A list of detected issues.

[2045] How it works: The generative AI model runs text analysis algorithms and uses natural language processing techniques to identify inappropriate language. Similarly, for images, it runs image analysis algorithms if they contain inappropriate content.

[2046] Step 4: Generate correction suggestions

[2047] Server: The server generates fixes based on the detected issues. The generative AI model considers the user's intent and proposes optimal fixes.

[2048] Input: A list of detected issues.

[2049] Output: A list of suggested fixes.

[2050] How it works: The generative AI model generates and suggests appropriate expressions based on the training dataset. For example, it generates a correction to "Hello, I love dogs!" in response to "Hello, I love dogs!"

[2051] Step 5: Emotion Recognition

[2052] Server: Runs the emotion engine to recognize emotions from the user's text and image data, for example, determining the emotional tone (positive, negative, neutral) of a sentence or image.

[2053] Input: User submissions (text and image data).

[2054] Output: Sentiment tag (e.g. positive, negative, neutral).

[2055] How it works: The emotion engine uses natural language processing and image analysis to analyze the sentiment of text and images and assign emotional tags.

[2056] Step 6: Adjust the proposed amendment

[2057] Server: Based on the recognition results of the emotion engine, the generated correction proposal is adjusted. If a strong emotion is included, the proposal is changed to a more tolerant expression, for example.

[2058] Input: revision suggestion list and sentiment tags.

[2059] Output: A refined list of proposed fixes.

[2060] What it does: A tuning algorithm is run to fine-tune the generated suggestions to take sentiment into account.

[2061] Step 7: Submit your proposed revisions

[2062] Server: Sends the adjusted correction proposal to the user device.

[2063] Input: The adjusted list of amendments.

[2064] Output: The proposed fix sent to the user's device.

[2065] Specific operation: The proposed revision data is packetized and transmitted to the user terminal via the Internet.

[2066] Step 8: Receive and view proposed revisions

[2067] Terminal: The proposed modifications sent from the server are displayed on the user interface. The user can review the proposed modifications and decide whether to apply them.

[2068] Input: The proposed fix sent by the server.

[2069] Output: The suggested fixes displayed in the user interface.

[2070] Specific operation: Retrieves data from the receiving buffer and reflects the received correction suggestions in the UI component.

[2071] Step 9: Review and apply proposed fixes

[2072] User: Review the proposed fix and click the "Apply" button to officially apply the fix.

[2073] Input: The suggested fix.

[2074] Output: Confirmation and applied text.

[2075] What happens: Once the proposed fix is ​​applied, a confirmation dialog will appear, allowing the user to make a final confirmation.

[2076] Step 10: Post

[2077] Server: After the user applies the suggested corrections, it checks them again and posts the finalized text to the social media platform.

[2078] Input: Confirmed and applied text.

[2079] Output: The text posted to the social media platform.

[2080] Specific operation: The confirmed text data is posted via the SNS API.

[2081] (Application example 2)

[2082] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2083] On social media and other online platforms, users often unconsciously use inappropriate or potentially misleading language. Furthermore, emotional comments made by users often lead to trouble and misunderstandings. A system to prevent such problems is needed.

[2084] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes processing means including a generative AI model that analyzes input sentences and detects inappropriate expressions or expressions that may be misleading, means for generating correction suggestions based on the detected problems, and means including an emotion recognition engine that recognizes the user's emotions regarding the generated correction suggestions and presents correction suggestions that take the emotions into consideration. This makes it possible to correct inappropriate expressions or misleading expressions while taking the user's emotions into consideration.

[2085] The "means for receiving input text" refers to a device or software that has the function of acquiring text data input by the user and transmitting it to the system.

[2086] "Processing means including a generative AI model that analyzes and detects inappropriate or potentially misleading language" refers to a system function that includes an AI algorithm that uses natural language processing techniques to analyze input text and identify inappropriate or potentially misleading language.

[2087] The "means for generating proposed fixes based on detected problems" is a processing function for creating alternatives to resolve problems found by the analysis. The means utilizes a generative AI model to automatically generate proposed fixes.

[2088] The "means including an emotion recognition engine" is a processing system for analyzing emotions from user input and providing optimal revision suggestions based on the emotion data.

[2089] "Means for confirming and applying proposed revisions" refers to a function that provides an operational interface for users to confirm revisions proposed by the system and officially select and apply those revisions.

[2090] "Means for posting the confirmed and applied text to various communication platforms" refers to a processing means for sending and posting the amendments selected and applied by the user to the relevant social networking site or other online communication tool.

[2091] This invention is a system that analyzes content posted by users on social networking sites, detects inappropriate or potentially misleading expressions, and suggests corrections. Furthermore, by combining it with an emotion engine that recognizes the user's emotions, it is possible to propose more appropriate corrections. Specific embodiments and their operation are described below.

[2092] System configuration

[2093] The system includes the following major components:

[2094] User terminal

[2095] It provides an interface for users to input content to post on social media.

[2096] The input text and image data are sent to the server.

[2097] The server presents the user with suggested revisions and sentiment-sensitive suggestions.

[2098] The user is then prompted to confirm and apply the proposed changes.

[2099] server

[2100] It hosts generative AI models and emotion engines and analyzes data received from users.

[2101] Detect inappropriate or misleading language and generate suggested corrections.

[2102] An emotion engine is used to recognize the user's emotions and reflect them in suggested revisions.

[2103] The proposed revisions are sent to the user's device, and the confirmed and applied text is posted on various communication platforms.

[2104] How it works

[2105] 1. Data Entry

[2106] The user enters the content of the SNS post into the user device interface. For example, they might enter, "This app is really useless!"

[2107] 2. Data Transmission

[2108] Once the input is complete, the user terminal sends this text data to the server. If image data is included, it is also sent.

[2109] 3. Data Analysis

[2110] The server passes the received text and image data to the generative AI model to begin analysis. The generative AI model analyzes the text and detects inappropriate or potentially misleading expressions.

[2111] 4. Generate correction suggestions

[2112] Based on the detected issues, the generative AI model generates the most appropriate correction suggestions while taking into account the user's intent. For example, it generates a correction suggestion such as, "This app still has room for improvement."

[2113] 5. Emotion recognition

[2114] The server activates an emotion recognition engine to recognize emotions from the user's input and image data. For example, if the input text expresses anger, the server generates a revision suggestion that takes this into consideration.

[2115] 6. Adjustment of amendments

[2116] Based on the emotions recognized by the emotion recognition engine, the proposed corrections are further refined and adjusted to take the user's emotions into consideration. For example, if anger is recognized, the expression will be changed to a calmer one.

[2117] 7. Proposal of amendments

[2118] The generated proposed corrections are sent to the user's terminal for the user to review. The user can review the proposed corrections and apply them by pressing the "Apply" button.

[2119] 8. Confirm and post the applied text

[2120] After the user confirms and applies the proposed changes, the finalized text is posted to various communication platforms. For example, the revised version of "This app still has room for improvement, but it has helped me in some ways" can be safely posted on social media.

[2121] Hardware and software used

[2122] Smartphone

[2123] Application execution environment

[2124] Mobile data transmission and reception

[2125] server

[2126] AWS (cloud server)

[2127] Google Cloud (cloud server)

[2128] Hosting generative AI models

[2129] Emotion Engine Hosting

[2130] Generative AI Models

[2131] GPT-4 and other modern natural language processing (NLP) models

[2132] Custom NLP model for profanity detection

[2133] Emotion Recognition Engine

[2134] Emotion recognition model created using TensorFlow and Keras

[2135] Specific examples

[2136] User: Type "This app is completely useless!"

[2137] Generative AI model: Generates suggested fixes such as, "This app still has room for improvement."

[2138] Emotion recognition engine: Recognizes the user's angry emotion and generates adjustment suggestions such as, "This app still has room for improvement, but it has helped in some ways."

[2139] User device: Check the suggested fixes and press the "Apply" button.

[2140] Server: Posts the confirmed text to the social networking site.

[2141] Example prompts to input to a generative AI model:

[2142] Original: "This app is completely useless!"

[2143] Generative AI: "This app still has room for improvement."

[2144] Emotion Engine: Recognizes the user's angry emotion and generates adjustment suggestions such as "This app still has room for improvement, but it has helped in some ways."

[2145] This makes it possible to take into consideration the user's feelings and prevent misunderstandings and problems in online communication.

[2146] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2147] Step 1:

[2148] The user enters the content of a social media post into the interface of the user device. The entered text is, "This app is really useless!" This input data is sent to the server in the next step.

[2149] Step 2:

[2150] The user terminal sends the input text data to the cloud server. Here, the mobile data is sent using a smartphone, and the cloud server receives the data.

[2151] Step 3:

[2152] The server passes the received text data to a generative AI model (an NLP model such as GPT-4) to begin analysis. The generative AI model analyzes the input text, "This app is really useless!", and detects inappropriate or potentially misleading expressions. In this step, text data analysis and inappropriate expression detection are performed.

[2153] Step 4:

[2154] The generative AI model generates a suggested fix based on the detected issues, for example, "This app still has room for improvement." The server then processes this suggested fix in the next step.

[2155] Step 5:

[2156] The server starts an emotion recognition engine (a model using TensorFlow or Keras) and analyzes the emotion from the user's input text. For example, it recognizes that the input text contains the emotion of anger. In this step, the emotion data is analyzed.

[2157] Step 6:

[2158] The server further refines the generated revision suggestions based on the emotional data analyzed by the emotion recognition engine. For example, it generates an optimal revision suggestion such as, "This app still has room for improvement, but it was helpful in some ways." In this step, data processing is performed taking emotions into consideration.

[2159] Step 7:

[2160] The server sends the final proposed fixes to the user's device, which then displays them for review. For example, the user might see a suggestion like, "This app still has room for improvement, but it helped in some ways."

[2161] Step 8:

[2162] The user checks the proposed revisions and applies them by pressing the "Apply" button. Here, the user selects a revision and the selection is sent from the terminal to the server.

[2163] Step 9:

[2164] The server receives the verified and applied text and posts it to various communication platforms. For example, the revised text "This app still has room for improvement, but it has helped me in some ways" is posted to social media. This step is the final data transmission.

[2165] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2166] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2167] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2168] The emotion identification model 59 as an emotion engine may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2169] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2170] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2171] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2172] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2173] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2174] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2175] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2176] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2177] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2178] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[2179] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2180] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2181] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[2182] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2183] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2184] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2185] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2186] The following is further disclosed regarding the above embodiment.

[2187] (Claim 1)

[2188] means for receiving input text;

[2189] A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions;

[2190] means for generating suggested modifications based on the problems detected by said processing means;

[2191] means for presenting the generated revision proposal;

[2192] means for reviewing and applying said amendments; and

[2193] A means for posting the confirmed and applied text on various communication platforms;

[2194] A system including:

[2195] (Claim 2)

[2196] 10. The system of claim 1, further comprising a processing means including a generative AI model that receives input image data and analyzes the image data to detect inappropriate or potentially misleading content.

[2197] (Claim 3)

[2198] 2. The system according to claim 1, further comprising means for presenting a plurality of revision suggestions to a user when a plurality of revision suggestions are generated, and allowing the user to select an optimum revision suggestion.

[2199] "Example 1"

[2200] (Claim 1)

[2201] means for receiving input text;

[2202] A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions;

[2203] means for generating suggested modifications based on the problems detected by said processing means;

[2204] means for presenting the generated revision proposal;

[2205] means for reviewing and applying said amendments; and

[2206] A means for posting the confirmed and applied text on various communication platforms;

[2207] means for the user terminal to display the proposed revision on a user interface;

[2208] a means for the server to perform a final check after applying said proposed amendment;

[2209] A system including:

[2210] (Claim 2)

[2211] 10. The system of claim 1, further comprising a processing means including a generative AI model that receives input image data and analyzes the image data to detect inappropriate or potentially misleading content.

[2212] (Claim 3)

[2213] 2. The system according to claim 1, further comprising means for presenting a plurality of revision suggestions to a user when a plurality of revision suggestions are generated, and allowing the user to select an optimum revision suggestion.

[2214] "Application Example 1"

[2215] (Claim 1)

[2216] means for receiving input text;

[2217] A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions;

[2218] means for generating suggested modifications based on the problems detected by said processing means;

[2219] means for presenting the generated revision proposal;

[2220] means for reviewing and applying said amendments; and

[2221] A means for posting the confirmed and applied text on various communication platforms;

[2222] The processing means analyzes the voice input of customer service in the physical store in real time, determines whether or not it contains inappropriate expressions, and proposes appropriate corrections;

[2223] A means for a store clerk to select an appropriate expression based on the proposed revision;

[2224] A system including:

[2225] (Claim 2)

[2226] 10. The system of claim 1, further comprising: processing means for receiving input image data and including a generative AI model for analyzing the image data to detect inappropriate or potentially misleading content.

[2227] (Claim 3)

[2228] 10. The system of claim 1, further comprising means for presenting the generated revision suggestions to the user in the event that there are multiple revision suggestions, and allowing the user to select the most appropriate revision suggestion.

[2229] "Example 2: Combining Emotion Engines"

[2230] (Claim 1)

[2231] means for receiving input text;

[2232] A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions;

[2233] means for generating suggested modifications based on the problems detected by said processing means;

[2234] means including an emotion engine for recognizing a user's emotion with respect to the revision proposal and generating a revision proposal that takes the emotion into consideration;

[2235] means for presenting the generated revision proposal;

[2236] means for reviewing and applying said amendments; and

[2237] A means for posting the confirmed and applied text on various communication platforms;

[2238] A system including:

[2239] (Claim 2)

[2240] 10. The system of claim 1, further comprising a processing means including a generative AI model that receives input image data and analyzes the image data to detect inappropriate or potentially misleading content.

[2241] (Claim 3)

[2242] 2. The system according to claim 1, further comprising means for presenting a plurality of revision suggestions to a user when a plurality of revision suggestions are generated, and allowing the user to select an optimum revision suggestion.

[2243] "Application example 2 when combining emotion engines"

[2244] (Claim 1)

[2245] means for receiving input text;

[2246] A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions;

[2247] means for generating suggested modifications based on the problems detected by said processing means;

[2248] means including an emotion recognition engine that recognizes a user's emotion regarding the generated revision proposal and presents a revision proposal that takes the emotion into consideration;

[2249] A means for verifying and applying the proposed amendments; and

[2250] A means for posting the confirmed and applied text on various communication platforms;

[2251] A system including:

[2252] (Claim 2)

[2253] 10. The system of claim 1, further comprising a processing means including a generative AI model that receives input image data and analyzes the image data to detect inappropriate or potentially misleading content.

[2254] (Claim 3)

[2255] 2. The system according to claim 1, further comprising means for presenting a plurality of revision suggestions to a user when a plurality of revision suggestions are generated, and allowing the user to select an optimum revision suggestion. [Explanation of symbols]

[2256] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving input text; A processing means including a generative AI model that analyzes the input sentence and detects inappropriate or potentially misleading expressions; means for generating suggested modifications based on the problems detected by said processing means; means for presenting the generated revision proposal; means for verifying and applying said amendments; A means for posting the confirmed and applied text on various communication platforms; A system including:

2. 10. The system of claim 1, further comprising a processing means including a generative AI model that receives input image data and analyzes the image data to detect inappropriate or potentially misleading content.

3. The system according to claim 1 , further comprising means for presenting the generated revision suggestions to the user when there are multiple revision suggestions, and allowing the user to select the most suitable revision suggestion.

Images