System

Abstract

A system is provided.SOLUTION: A system comprising: means for monitoring an online community to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them into a polite style; means for selecting a generative AI model having expertise; means for organizing opinion conflicts by the selected generative AI model and generating consensus comments; and means for generating neutral and constructive advice.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] Disagreements and arguments between users frequently occur on online platforms such as social networking sites and chat rooms. Effectively resolving these issues between the parties can be difficult and can cause emotional distress and stress. Furthermore, if these conflicts escalate, they can pose a risk of litigation or police involvement. The present invention aims to quickly and effectively resolve disputes and arguments within online communities, thereby providing a healthier communication environment. [Means for solving the problem]

[0005] The present invention includes a means for monitoring online communities and automatically detecting offensive comments and arguments. It also provides a means for analyzing detected comments and rewriting them in a polite style. It also provides a means for selecting a generative AI model with specific expertise and using that generative AI model to sort out conflicting opinions and generate empathetic comments. It also includes a means for providing neutral and constructive advice to both parties in a dispute. In this way, online disputes can be quickly resolved and a healthy communication environment can be established.

[0006] An "online community" refers to a virtual space where users can interact with each other via the Internet.

[0007] "Monitoring" means that the system monitors and collects data in real time to detect specific conditions or events.

[0008] "Offensive comments" are statements made with the intent to humiliate or insult others.

[0009] A "fight" refers to a disagreement or argument between users.

[0010] "Automatically detects" means that the system determines and identifies specific events or conditions without human intervention.

[0011] "Analysis" refers to the act of a system analyzing data in detail to clarify its content and intent.

[0012] "Polite writing style" means written expression that is polite, respectful, and inoffensive.

[0013] "Rewriting" refers to the act of changing an existing document into a different style or expression.

[0014] "Generative AI model" means an algorithm or system that uses artificial intelligence to generate text or data.

[0015] "Expertise" refers to having a high level of understanding or skill in a particular field or topic.

[0016] "Conflict of opinion" refers to a situation in which multiple users have different opinions or positions.

[0017] "Empathetic comments" refer to comments that show understanding and empathy for the other person's opinions and feelings.

[0018] "Constructive advice" refers to the act of making positive suggestions or advice toward solving problems or making improvements.

[0019] A "system" is a collection of elements and processes that are configured to achieve a specific purpose. [Brief explanation of the drawings]

[0020] [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

[0021] 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.

[0022] First, the terms used in the following description will be explained.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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."

[0028] [First embodiment]

[0029] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0030] 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.

[0031] 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).

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0037] 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.

[0038] 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.

[0039] 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.

[0040] 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."

[0041] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, which is implemented through the following process.

[0042] System configuration

[0043] 1. Monitoring and Detection

[0044] The server monitors all conversation threads in online communities in real time and automatically detects offensive comments and signs of conflict using natural language processing (NLP) technology, specifically by detecting negative language patterns (e.g., "stupid" or "useless").

[0045] 2. Certification and Selection

[0046] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict. Based on this, it selects a generative AI model with specific expertise and selects it as the appropriate arbitrator. For example, in the case of a technical discussion, a technically specialized GPT (generative AI model) is selected.

[0047] 3. Automatic Rewrite

[0048] The arbitration GPT rewrites offensive comments into polite language. For example, it converts a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?" The rewritten comment is sent to the terminal via the server and displayed on the user's screen.

[0049] 4. Facilitate

[0050] The arbitration GPT organizes the content of comments and identifies the root cause of the conflict. For example, in the case of a technical conflict, it will evenly summarize the advantages and disadvantages of the solution. It also generates empathetic comments to satisfy the user's desire for approval. For example, it generates a comment such as, "Your opinion is very interesting. I was particularly impressed with the part about XX." This is also sent to the device via the server and displayed on the user's screen.

[0051] 5. Reconciliation Advice

[0052] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0053] Specific examples

[0054] Case

[0055] If you encounter technical conflicts such as:

[0056] User A: This code is completely unusable. It's full of bugs.

[0057] User B: You're the one who doesn't understand anything.

[0058] Example

[0059] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0060] 2. The server selects a technically-specialized arbitrator, the GPT.

[0061] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0062] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0063] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0064] 4. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0065] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments.

[0066] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0067] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0068] 7. Arbitration GPT generates constructive advice towards a resolution.

[0069] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0070] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[0071] In this way, the system of the present invention can quickly and effectively resolve disputes and conflicts in online communities, providing a healthy communication environment.

[0072] The processing flow will be explained below.

[0073] Step 1:

[0074] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[0075] Step 2:

[0076] The server analyzes the detected offensive comments and uses a comment scoring algorithm to quantify their offensiveness. For example, if a score above a certain threshold is assigned, the comment is deemed definitively offensive.

[0077] Step 3:

[0078] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[0079] Step 4:

[0080] The arbitration GPT takes offensive comments and applies an algorithm to rewrite them in a more polite style, for example rewriting a comment like "Are you stupid?" to "Could you please elaborate a bit more on your point of view?"

[0081] Step 5:

[0082] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[0083] Step 6:

[0084] Arbitration GPT analyzes the background and arguments of the conflicting opinions and identifies the fundamental conflicts. For example, in the case of technical conflicts, it evenly summarizes the advantages and disadvantages of the conflicting solutions.

[0085] Step 7:

[0086] Arbitration GPT generates comments to foster empathy, such as "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?"

[0087] Step 8:

[0088] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[0089] Step 9:

[0090] Arbitration GPT generates neutral and constructive advice, such as "How about creating an opportunity to discuss each other's opinions in more detail?"

[0091] Step 10:

[0092] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[0093] Example 1

[0094] 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."

[0095] As online communities grow, the number of offensive comments and conflicts between users increases, leading to a deterioration in the communication environment. Conventional methods rely on manual moderation, making it difficult to respond quickly and effectively, so an automated system is needed to maintain the health of online communities.

[0096] 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.

[0097] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and conflicts, means for analyzing the detected comments to identify the subject of the conflict and select an appropriate generative AI model, means for automatically rewriting the offensive comments to a polite style using the selected generative AI model, means for the generative AI model to organize the conflicting opinions and generate empathetic comments, and means for generating neutral and constructive advice. This makes it possible to quickly and effectively resolve offensive comments and conflicts in online communities and provide a healthy communication environment.

[0098] An "online community" is a virtual space where people can exchange information and interact via the Internet.

[0099] "Monitoring" means to observe a specific object periodically or continuously to check its trends and status.

[0100] "Offensive comments" are statements that contain negative, insulting, or provocative content toward others.

[0101] "Conflict" refers to a clash or dispute that arises due to different opinions or values.

[0102] "Automatic detection" refers to the ability of a system to detect specific events based on certain conditions or patterns without human intervention.

[0103] A "generative AI model" is an artificial intelligence system that can perform generative tasks (such as text generation or translation).

[0104] "Polite writing style" is a characteristic of writing that uses gentle expressions that convey respect and consideration.

[0105] "Rewriting" means changing the form or wording of an original text while preserving its content.

[0106] An "empathetic comment" is a statement that shows understanding or agreement with the feelings or opinions of others.

[0107] "Constructive advice" refers to making specific and positive suggestions for solving problems or making improvements.

[0108] "Natural language processing technology" refers to technology that allows computers to understand, analyze, and generate human language.

[0109] This invention is a system for quickly and effectively resolving offensive comments and conflicts in online communities, which mainly involves the cooperation of three entities: a server, a terminal, and a user.

[0110] System configuration

[0111] Hardware and software used

[0112] Server: A high-performance server system for monitoring online communities in real time. Data is collected using a communication API.

[0113] Natural Language Processing (NLP): Comment data is analyzed using Google® Cloud Natural Language and SpaCy.

[0114] Generative AI model: An AI system used for generative tasks. The model uses the technically specialized GPT (Generative Pre-trained Transformer).

[0115] Implementation method

[0116] 1. Monitoring

[0117] The server collects comment data from online communities via specific APIs (e.g., Twitter API or Reddit API). The collected data is periodically sent to the server and analyzed in real time using NLP technology.

[0118] 2. Detecting offensive comments

[0119] The server analyzes the collected comment data, searching for negative language patterns and specific keywords (e.g., "stupid," "useless," etc.), using a pre-trained machine learning model (e.g., the BERT model).

[0120] 3. Selecting a Generative AI Model

[0121] The server further analyzes the content of the detected comments to identify the topic of conflict, and then selects the most appropriate generative AI model (e.g., technically specialized GPT) based on the identified topic.

[0122] 4. Rewrite offensive comments

[0123] The server generates a prompt sentence that inputs the detected offensive comment into the arbitration GPT model. For example, a prompt sentence of the form "Please rewrite the following offensive comment in a polite style: 'Are you stupid?'" is created. The arbitration GPT generates the rewritten comment and sends it to the terminal (the user's display device) via the server.

[0124] 5. Generating empathetic comments and advice

[0125] The arbitration GPT analyzes the cause of the conflict and generates empathetic comments and neutral advice. For example, it might create a comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?" These comments are also sent to the terminal via the server and displayed to the user.

[0126] Specific examples

[0127] Case

[0128] If you encounter technical conflicts such as:

[0129] User A: This code is completely unusable. It's full of bugs.

[0130] User B: You're the one who doesn't understand anything.

[0131] Example

[0132] 1. The server monitors comments and detects offensive comments ("useless" or "you don't understand").

[0133] 2. The server selects a technically-specialized arbitrator GPT for the detected comment.

[0134] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0135] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0136] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0137] 4. The server sends the rewritten comments to the terminals and displays them on the screens of User A and User B.

[0138] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments:

[0139] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0140] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0141] 7. Arbitration GPT generates constructive advice towards resolution:

[0142] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0143] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[0144] As a result, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[0145] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0146] Step 1:

[0147] The server collects comment data from online communities. Specifically, it retrieves comment data in real time using a specific API (e.g., Twitter API or Reddit API). The input is the comment data retrieved from the API, and the output is the raw comment data to be analyzed.

[0148] Step 2:

[0149] The server applies natural language processing (NLP) techniques to the collected comment data to detect offensive or conflicting comments. Specifically, it performs text analysis using NLP tools (such as Google Cloud Natural Language or SpaCy). The input is raw comment data, and the output is comment data flagged as offensive or conflicting.

[0150] Step 3:

[0151] The server performs a detailed analysis of the content of the detected offensive comments and identifies the subject of the conflict. Specifically, it uses topic modeling techniques (such as LDA and CLIP) to identify what the conflict is about. The input is flagged comment data, and the output is data identifying the subject of the conflict.

[0152] Step 4:

[0153] The server selects an appropriate generative AI model (arbitration GPT) based on the identified subject of the conflict. The input is data identifying the subject of the conflict, and the output is information about the selected generative AI model.

[0154] Step 5:

[0155] The server generates a prompt for the selected generative AI model to rewrite the offensive comment. An example of a specific prompt is "Please rewrite the following offensive comment in a polite manner: 'Are you stupid?'" The input is the offensive comment and information about the selected generative AI model, and the output is the prompt.

[0156] Step 6:

[0157] The arbitration GPT receives the generated prompt and rewrites the offensive comments to a more polite style. The input is the prompt, and the output is the rewritten comment.

[0158] Step 7:

[0159] The server sends the rewritten comment to the terminal and displays it on the user's screen. The input is the rewritten comment, and the output is the rewritten comment displayed on the user's screen.

[0160] Step 8:

[0161] The arbitration GPT generates empathetic comments based on the rewritten comments and conflict analysis data. Specifically, it generates phrases that ease the conflict and deepen understanding between users. The inputs are the rewritten comments and conflict analysis data, and the output is empathetic comments.

[0162] Step 9:

[0163] The server sends the generated empathy comments to the terminal and displays them to the user. The input is the empathy comments, and the output is the empathy comments displayed on the user's screen.

[0164] Step 10:

[0165] The arbitration GPT ultimately generates constructive advice for resolving the conflict. Specifically, it suggests something like, "How about setting up an opportunity to discuss each other's opinions in more detail?" The inputs are conflict analysis data and the context of the dialogue, and the output is constructive advice.

[0166] Step 11:

[0167] The server sends the generated advice comment to the terminal and displays it on the user's screen. The advice comment is the input, and the advice comment displayed on the user's screen is obtained as the output.

[0168] (Application example 1)

[0169] 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."

[0170] Currently, offensive comments and conflicts are common in online community comment sections, hindering healthy communication. Applications used on smart devices, in particular, lack the ability to detect and appropriately address these issues in real time, resulting in a poor user experience. This invention aims to provide a system that quickly and effectively detects offensive comments and promotes constructive communication.

[0171] 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.

[0172] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for monitoring the comment section of an application used on a smart device to detect offensive comments in real time, and means for rewriting offensive comments to polite expressions and generating empathetic comments and constructive advice. This makes it possible to effectively suppress offensive behavior in the comment section and promote healthy and constructive communication.

[0173] An "online community" is a virtual space on the Internet where users can exchange opinions and share information.

[0174] "Monitoring" is the process by which a system continuously monitors all conversation threads in an online community to detect problematic comments in real time.

[0175] "Offensive comments" are comments that contain expressions or words intended to harm other users or that may cause conflict.

[0176] A "fight" refers to an emotional or aggressive conflict between users within an online community.

[0177] "Polite writing style" refers to the transformation of aggressive comments into expressions that respect others and allow for polite communication.

[0178] A "generative AI model" is an algorithm or system that has the ability to generate natural language using artificial intelligence technology.

[0179] An "empathetic comment" is a comment that includes expressions that empathize with the opinions and feelings of others and show positive and understanding.

[0180] "Neutral and constructive advice" is practical advice given from an impartial standpoint to resolve conflicts.

[0181] A "smart device" is a computerized portable device that can connect to the Internet and perform multiple functions. Examples include smartphones, smart glasses, and head-mounted displays.

[0182] "Real-time" means that data and information are processed almost immediately and results are obtained without any time lag.

[0183] "Polite expression" is language that is polite and carefully chosen so as not to offend others.

[0184] MODE FOR CARRYING OUT THE INVENTION

[0185] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. The system includes an application and a server that are run on smart devices.

[0186] Overall system configuration

[0187] The server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them to a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, and means for generating neutral and constructive advice. The server also includes means for monitoring comment sections in an application used on a smart device and detecting offensive comments in real time.

[0188] Hardware and Software Used

[0189] Hardware:

[0190] Smartphone

[0191] Smart Glasses

[0192] head-mounted display

[0193] software:

[0194] Natural language processing (NLP) models for monitoring (e.g., parsers such as Hugging Face's BERT)

[0195] Generative AI models (e.g., GPT models) for detecting and rewriting offensive comments

[0196] Data processing and calculation

[0197] The server monitors online community comments in real time and uses NLP technology, such as Hugging Face's "sentiment-analysis" pipeline, to detect offensive comments. Detected offensive comments generate appropriate prompts, which are then fed into a generative AI model (e.g., a GPT model). This generative AI model rewrites offensive comments into politer expressions and also generates empathetic comments and constructive advice.

[0198] Specific examples

[0199] Below is a concrete example of the process for detecting offensive comments, rewriting them, and generating empathetic comments and constructive advice.

[0200] Comment example

[0201] User A: "This video is completely pointless. It's a waste of time."

[0202] User B: "No one is listening to your opinion. Shut up."

[0203] Prompt Sentence Examples

[0204] "Rewrite the following offensive comment politely: This video is completely pointless. A waste of time."

[0205] "Create a neutral and empathetic response to the following opinion: No one is listening to your opinion. Shut up."

[0206] "Provide a constructive advice to resolve an online argument."

[0207] The server generates these prompts and inputs them into a generative AI model, which rewrites offensive comments into polite expressions and generates empathetic comments and constructive advice. The rewritten comments and advice are then displayed on the user's smart device screen, promoting healthy communication.

[0208] In this way, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[0209] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0210] Step 1:

[0211] The server monitors the comment section of an online community in real time. As input, comment data is provided from the online community. As output, offensive comments or comments that indicate signs of conflict are extracted. Specifically, the server uses a natural language processing (NLP) model to analyze the sentiment of the comments, for example, using Hugging Face's "sentiment-analysis" pipeline.

[0212] Step 2:

[0213] The server analyzes the extracted offensive comments. As input, it uses the offensive comments extracted in step 1. As output, it identifies the offensive elements and provides information to translate them into polite language. Specifically, the server runs an algorithm to identify negative words and phrases in the comments.

[0214] Step 3:

[0215] The server selects a generative AI model. As input, it uses the comment content and its characteristic information analyzed in step 2. As output, it selects an appropriate generative AI model (e.g., GPT model). Specifically, the server applies rule-based logic to select an appropriate AI model based on the comment content.

[0216] Step 4:

[0217] The server rewrites offensive comments into polite language. The selected generative AI model and the offensive comment are used as input. The rewritten, polite comment is generated as output. Specifically, the server generates a prompt sentence: "Rewrite the following offensive comment politely: [offensive comment]" and inputs it into the generative AI model.

[0218] Step 5:

[0219] The server sorts out the conflicting opinions and generates empathetic comments. The rewritten comments and the content of the conflict are used as input. The output is an empathetic comment. Specifically, the server generates a prompt statement: "Create a neutral and empathetic response to the following opinion: [content of the conflict]" and inputs this into the generative AI model.

[0220] Step 6:

[0221] The server generates neutral and constructive advice. Empathetic comments and conflict resolution information are used as input. Constructive advice is generated as output. Specifically, the server generates a prompt statement, "Provide a constructive advice to resolve an online argument," and inputs this into the generative AI model.

[0222] Step 7:

[0223] The server sends the generated rewrite comments, empathy comments, and constructive advice to the terminal. As input, the comments generated in steps 4, 5, and 6 are used. As output, rewrite comments, empathy comments, and constructive advice are generated and displayed on the terminal. Specifically, the server uses a protocol to combine each comment and send it to the user's screen.

[0224] 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.

[0225] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, and is further enhanced by combining it with an emotion engine. This system is implemented using the following process:

[0226] System configuration

[0227] 1. Monitoring and Detection

[0228] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect and quantify comments containing offensive words or phrases.

[0229] 2. Certification and Selection

[0230] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[0231] 3. Emotion recognition

[0232] The emotion engine detects emotions from the user's text messages. It also uses physiological indicators (e.g., heart rate, sweat rate, blink rate) to recognize the user's emotions. For example, if it detects negative emotions (e.g., anger, sadness, irritation), it uses that information for the next step.

[0233] 4. Automatic Rewrite

[0234] The arbitration GPT receives offensive comments and rewrites them in a more polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[0235] 5. Facilitate

[0236] The arbitration GPT sorts through the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This is also sent to the device via the server and displayed on the user's screen.

[0237] 6. Reconciliation Advice

[0238] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0239] Specific examples

[0240] Case

[0241] If you encounter technical conflicts such as:

[0242] User A: This code is completely unusable. It's full of bugs.

[0243] User B: You're the one who doesn't understand anything.

[0244] Example

[0245] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0246] 2. The server selects a technically-specialized arbitrator, the GPT.

[0247] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[0248] 4. Arbitration GPT rewrites the above comment as follows:

[0249] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0250] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0251] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0252] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[0253] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0254] 7. The server sends the empathy comment to the terminal and it is displayed on the user's screen.

[0255] 8. Arbitration GPT generates constructive advice towards a resolution.

[0256] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0257] 9. The server sends the advice comment to the terminal and it is displayed on the user's screen.

[0258] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[0259] The processing flow will be explained below.

[0260] Step 1:

[0261] The server monitors all conversation threads in online communities in real time, using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[0262] Step 2:

[0263] The server analyzes the detected offensive comments and quantifies their offensiveness. Specifically, if a score exceeding a certain threshold is assigned, the comment is deemed to be definitively offensive.

[0264] Step 3:

[0265] The server analyzes the content of detected offensive comments and identifies the subject of the conflict, categorizing them into categories such as technical arguments, personal attacks, and political conflicts.

[0266] Step 4:

[0267] Based on the identified subject, the server selects a generative AI model (GPT) with domain expertise. For example, in the case of a technical discussion, a technical GPT is selected.

[0268] Step 5:

[0269] The emotion engine analyzes the user's text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.) to recognize the user's emotions. For example, it can determine that User A is feeling "anger" based on text analysis.

[0270] Step 6:

[0271] The arbitration GPT applies an algorithm to rewrite offensive comments to a more polite style, taking into account the results of the sentiment engine and rewriting appropriately. For example, it transforms a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?"

[0272] Step 7:

[0273] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[0274] Step 8:

[0275] The arbitration GPT identifies the root cause of the conflict based on the content of the comments and the results of the sentiment engine. For example, in a technical conflict, it will equally explain the advantages and disadvantages of each solution.

[0276] Step 9:

[0277] The arbitration GPT generates comments to foster empathy. Based on the emotional information detected by the emotion engine, it generates more appropriate empathetic comments. For example, it might say, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?"

[0278] Step 10:

[0279] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[0280] Step 11:

[0281] The arbitration GPT generates neutral and constructive advice, taking into account the information detected by the emotion engine and proposing a solution that both parties can agree on. For example, it may provide advice such as, "How about setting up an opportunity to discuss your respective opinions in more detail?"

[0282] Step 12:

[0283] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[0284] Example 2

[0285] 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."

[0286] In online communities, aggressive comments and arguments are common and often impede smooth communication. This increases the psychological burden on users and worsens the overall atmosphere of the community. Furthermore, emotional conflicts often deepen, preventing constructive discussions. Conventional automated moderation systems are limited to simply filtering or deleting words, and do not fundamentally address the problem. Therefore, this invention aims to resolve conflicts quickly and effectively by recognizing emotions and promoting appropriate communication.

[0287] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for monitoring an online community to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them in a polite style; means for selecting a generative AI model with specialized knowledge; means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model; means for generating neutral and constructive advice; means for recognizing emotions using a user's text message and physiological indicators; and means for rewriting offensive comments based on the emotion recognition results. This makes it possible to rewrite offensive comments and generate appropriate mediating comments taking emotions into consideration, thereby enabling disputes in online communities to be resolved quickly and effectively.

[0288] An "online community" is a virtual space where people can interact using the Internet.

[0289] "Offensive comments" are comments that contain words or phrases intended to insult or harm others.

[0290] A "fight" is a dispute that arises from a clash of opinions or feelings within an online community.

[0291] "Emotion recognition means" refers to technology that analyzes a user's text messages and physiological indicators to detect their emotional state.

[0292] "Polite rewriting" refers to an algorithm that automatically converts offensive comments into more readable and polite language.

[0293] A "generative AI model" is a model that uses artificial intelligence to generate text, particularly using deep learning techniques.

[0294] "Means for sorting out conflicting opinions and generating empathetic comments" refers to the process of analyzing conflicting opinions and generating comments that show empathy for those opinions.

[0295] "Means for generating neutral and constructive advice" refers to technology that automatically generates neutral and helpful suggestions to defuse conflicts.

[0296] "Physiological indicators" refer to data that indicate the user's physical condition, such as heart rate, sweat rate, and blink rate.

[0297] This invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. Specifically, it is more effective when combined with an emotion engine. This system is implemented using the following process:

[0298] Monitoring and Detection

[0299] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) technology. A specific implementation example uses Apache Kafka to process large volumes of messages in real time and performs analysis using the SpaCy library. It detects comments containing offensive words and phrases and quantifies them. This process allows problematic comments to be identified immediately.

[0300] Certification and Selection

[0301] The server analyzes the content of detected offensive comments to identify the subject of the conflict (technical discussion, personal attack, political conflict, etc.). This analysis uses a topic modeling algorithm (e.g., LDA) to classify the content of the comment. Based on the identified subject, the server selects the most appropriate generative AI model (e.g., GPT-3 (registered trademark)).

[0302] emotion recognition

[0303] The emotion engine detects emotions from users' text messages. This process uses the Sentiment Analysis API to classify emotions from text. Additionally, physiological indicators (heart rate, sweat rate, blink rate, etc.) are also collected using sensors such as Arduino and Raspberry Pi.

[0304] Automatic Rewrite

[0305] The arbitration GPT receives offensive comments and rewrites them in a polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[0306] Facilitate

[0307] The arbitration GPT organizes the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This empathetic comment is also sent to the device via the server and displayed on the user's screen.

[0308] Reconciliation advice

[0309] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0310] Specific examples

[0311] An example of a system in place in the event of a technical conflict

[0312] User A: This code is completely unusable. It's full of bugs.

[0313] User B: You're the one who doesn't understand anything.

[0314] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0315] 2. The server selects a technically specialized generative AI model.

[0316] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[0317] 4. Arbitration GPT rewrites the above comment as follows:

[0318] User A (after rewrite): This code looks a bit problematic. It might need some fixing.

[0319] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0320] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0321] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[0322] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0323] 7. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0324] 8. Arbitration GPT generates constructive advice towards a resolution.

[0325] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0326] 9. The server sends the advice comment to the terminal and displays it on the user's screen.

[0327] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[0328] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0329] Step 1:

[0330] The server collects all conversation threads of an online community.

[0331] Input: New message in the community

[0332] Processing: Stream messages in real time using Apache Kafka and store them in a database.

[0333] Output: All saved message logs

[0334] Step 2:

[0335] The server analyzes the collected messages using natural language processing (NLP) technology to detect offensive words and phrases.

[0336] Input: Saved message log

[0337] Processing: Parse the message text using the SpaCy library and match it against a list of negative words

[0338] How it works: NLP analysis detects comments like "Are you stupid?" and records them as a numerical value

[0339] Output: List of offensive comments

[0340] Step 3:

[0341] The server analyzes the content of the detected offensive comments to identify the subject of the conflict.

[0342] Input: offensive comment list

[0343] Processing: Classify comments using a topic modeling algorithm (e.g. LDA)

[0344] Behavior: Categorized into categories such as "technical debate," "personal attacks," and "political conflict"

[0345] Output: Categorized offensive comments

[0346] Step 4:

[0347] Based on the identified subject, the server selects a generative AI model in that field.

[0348] Input: Categorized offensive comment

[0349] Processing: Select a generative AI model (e.g., GPT-3) appropriate for each subject

[0350] How it works: For technical discussions, we select a technically specialized GPT-3 model.

[0351] Output: The selected generative AI model

[0352] Step 5:

[0353] The emotion engine uses the user's text messages and physiological indicators to recognize emotions.

[0354] Input: User text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.)

[0355] Processing: Detect emotional states using Sentiment Analysis API and physiological data sensors (Arduino, Raspberry Pi, etc.)

[0356] Behavior: Classifying emotions such as "anger" and "sadness" from text messages and integrating them with physiological indicators

[0357] Output: Emotion recognition result

[0358] Step 6:

[0359] The server uses a generative AI model to rewrite offensive comments based on the emotional information detected by the emotion engine.

[0360] Input: offensive comments and emotion recognition results

[0361] Processing: Generate prompts for rewriting and send them to the generative AI model

[0362] Behavior: Transform "Are you stupid?" into a more polite statement such as "Can you explain your point of view in more detail?"

[0363] Output: Rewritten comments

[0364] Step 7:

[0365] The server sends the rewritten comment to the terminal and displays it on the user's screen.

[0366] Input: Rewritten comment

[0367] Processing: Send the rewritten comment to the device using the REST API

[0368] How it works: Rewritten comments are displayed in real time on the chat screen.

[0369] Output: Rewrite comment displayed to the user

[0370] Step 8:

[0371] The server uses a generative AI model to generate empathetic comments based on the information provided by the emotion engine.

[0372] Input: Emotion recognition results

[0373] Processing: Generate prompts for generating empathetic comments and send them to the generative AI model

[0374] What it does: Generates empathetic comments like, "It's really important to calmly discuss bugs in your code. Can you share your perspectives on the issues?"

[0375] Output: Empathetic comments

[0376] Step 9:

[0377] The server sends the sympathetic comments to the terminal and displays them on the user's screen.

[0378] Input: Sympathetic comment

[0379] Processing: Send sympathetic comments to the device using the REST API

[0380] How it works: Show empathetic comments in real time on the chat screen

[0381] Output: Empathy comments displayed on the user's screen

[0382] Step 10:

[0383] The server uses arbitration GPT to generate neutral and constructive advice.

[0384] Input: Emotion recognition results and conflict content

[0385] Processing: Generate prompts for advice generation and send them to the generative AI model

[0386] Behavior: Generates constructive advice such as, "How about we set up an opportunity to discuss each other's opinions in more detail?"

[0387] Output: Advisory comments

[0388] Step 11:

[0389] The server sends the advice comment to the terminal and displays it on the user's screen.

[0390] Input: Advisory Comment

[0391] Processing: Sends advice to the device using the REST API

[0392] How it works: Real-time advice is displayed on the chat screen.

[0393] Output: Advisory comments displayed on the user's screen

[0394] This processing flow enables aggressive comments and disputes in online communities to be resolved quickly and effectively, providing a healthy communication environment that also takes into consideration the feelings of users.

[0395] (Application example 2)

[0396] 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."

[0397] In traditional online communities, aggressive comments and arguments are common, often damaging the user experience. Even in brick-and-mortar stores, communication between customers and staff, or between staff members, can be disrupted, leading to lower customer satisfaction and reduced staff efficiency. A system that can quickly and effectively resolve these issues and promote healthy communication is needed.

[0398] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for applying the monitored communication content to customer service and staff dialogue in physical stores, means for providing real-time filtering and rewriting of offensive comments via a smartphone, means for visualizing the emotions of customers and staff using an emotion engine, and means for displaying empathetic comments and constructive advice in the physical store environment in real time. This makes it possible to provide a healthy communication environment in online communities and physical stores.

[0399] An "online community" is a place where users can interact with each other and share information via the Internet.

[0400] "Monitoring" is the act of observing and monitoring specific situations or data in real time.

[0401] "Offensive comments" are statements that contain the intent to insult or attack others.

[0402] A "quarrel" refers to a clash caused by an argument or a difference of opinion.

[0403] "Rewriting" is the act of reconstructing a sentence and rewriting it in a different way.

[0404] A "generative AI model" is an artificial intelligence model that is trained to perform generative tasks.

[0405] "Selection" is the act of choosing an appropriate subject.

[0406] An "empathetic comment" is a statement made to show empathy for the feelings or opinions of others.

[0407] "Neutral and constructive advice" is advice that is unbiased and aims to lead in a positive direction.

[0408] A "system" is a collection of elements organized to achieve a particular purpose.

[0409] A "physical store" is a store or service facility that exists in a physical location.

[0410] "Customer service" refers to the act of responding to customer requests and questions.

[0411] "Inter-staff dialogue" refers to communication between employees who belong to the same workplace.

[0412] A "smartphone" is a mobile device that has similar communication capabilities to a mobile phone, but with more advanced computing capabilities.

[0413] An "emotion engine" is a technology that analyzes and recognizes a user's emotions from text and physiological data.

[0414] "Real-time filtering" is the process of instantly analyzing data or information and sorting it based on specific criteria.

[0415] "Visualization" is the process of expressing abstract data or information in a visually easy-to-understand form.

[0416] The present invention provides a system for monitoring communications in online communities and brick-and-mortar stores and effectively resolving offensive comments and conflicts. Hereinafter, embodiments of the present invention will be described in detail.

[0417] System Program

[0418] The system of the present invention is implemented using the following hardware and software:

[0419] Hardware: Smartphones, servers

[0420] Software: Natural language processing (NLP) libraries (e.g., spaCy), sentiment analysis tools (e.g., AWS® Comprehend, Google Cloud Natural Language API), generative AI models (e.g., OpenAI® GPT)

[0421] Process Overview

[0422] Monitoring and Detection

[0423] The server monitors communication content in physical stores and online communities in real time. This monitoring includes using NLP technology to detect offensive words and phrases. For example, if a user types an offensive comment such as "it's unusable" into their smartphone, the content will be detected immediately.

[0424] Comment Rewrite

[0425] Detected offensive comments are automatically rewritten to a more polite style using a generative AI model. The server inputs appropriate prompts to the generative AI model and receives the rewritten comments. For example, a comment such as "The service on this product was poor!" is rewritten to "There seems to be some room for improvement in the service on this product."

[0426] Visualizing emotions

[0427] The server uses an emotion engine to analyze the user's emotions and notify customers and staff of negative emotions. The results of the emotion analysis are visualized and displayed in real time on smartphones and other devices. For example, if "anger" is detected based on a customer's text and physiological data, that information is displayed on the staff's device.

[0428] Empathy comment generation

[0429] The selected generative AI model generates empathetic comments based on the detected emotional information, which promotes conflict resolution. The server reads the output of the emotion engine and inputs a prompt sentence to generate an empathetic comment. For example, the prompt sentence used is "Generate an empathetic comment for the following message: The service on this product was poor!"

[0430] Providing constructive advice

[0431] Neutral and constructive advice is generated by the generative AI model and provided to store staff and customers in real time, enabling all parties involved to find a better solution. The server inputs prompts offering constructive advice based on the content of the conflict, and the generated advice is displayed on the device.

[0432] Specific examples

[0433] Handling customer complaints

[0434] If a customer says, "The service on this product was poor!"

[0435] 1. The server detects this offensive comment.

[0436] 2. Using the generative AI model, rewrite the sentence to "There seems to be some room for improvement in how this product is handled."

[0437] 3. The emotion engine detects anger and notifies staff.

[0438] 4. The server inputs the prompt text "Generate an empathetic comment for the following message: The service on this product was poor!" and generates an empathetic comment.

[0439] 5. Provide the staff with a sympathetic comment saying, "We take your feedback seriously and will strive to improve," along with constructive advice saying, "We would appreciate it if you could tell us the specific issues you had."

[0440] This will enable us to provide a healthy communication environment in our physical stores and online communities.

[0441] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0442] Step 1:

[0443] The server monitors communication content in physical stores and online communities. The input includes user text messages, and the data is analyzed using NLP techniques. The output detects messages containing offensive comments or negative language.

[0444] Step 2:

[0445] The server inputs the detected offensive comments into a generative AI model, which rewrites them in a polite style. The input includes an offensive text message, which is passed to the generative AI model along with a prompt. The output is a rewritten, polite comment. For example, a comment like "The service on this product was poor!" is changed to "There seems to be some room for improvement in the service on this product."

[0446] Step 3:

[0447] The server uses an emotion engine to analyze emotions from the user's text message and physiological indicators (e.g., heart rate, sweat rate, blink rate). The input includes the rewritten text message and physiological data. The output is the emotion analysis result, which detects emotions such as "anger" or "sadness."

[0448] Step 4:

[0449] The server generates an empathetic comment based on the analysis results of the emotion engine. The input includes the emotion analysis results and the rewritten text message, which are passed to the generative AI model along with a prompt. The output is a comment expressing empathy. For example, a prompt such as "Generate an empathetic comment for the following message: The service on this product was poor!" is used.

[0450] Step 5:

[0451] The server generates neutral and constructive advice. The input includes the sentiment analysis results and empathetic comments, which are passed to the generative AI model along with a prompt. The output is constructive advice. For example, the generated advice might be, "We would appreciate it if you could tell us the specific problem you are having."

[0452] Step 6:

[0453] The terminal displays the rewritten comments, empathetic comments, and constructive advice sent from the server to the user. The input includes the rewritten comments, empathetic comments, and advice from the server. The output is displayed on the user's screen, promoting healthy communication.

[0454] Step 7:

[0455] The user can smoothly communicate based on the rewritten comments, empathetic comments, and constructive advice displayed on the terminal. The input includes the content displayed on the terminal, and the user's response and behavior are obtained as the output.

[0456] 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.

[0457] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0458] 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.

[0459] [Second embodiment]

[0460] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0461] 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.

[0462] 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).

[0463] 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.

[0464] 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.

[0465] 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).

[0466] 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.

[0467] 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.

[0468] 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.

[0469] 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.

[0470] 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.

[0471] 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."

[0472] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, which is implemented through the following process.

[0473] System configuration

[0474] 1. Monitoring and Detection

[0475] The server monitors all conversation threads in online communities in real time and automatically detects offensive comments and signs of conflict using natural language processing (NLP) technology, specifically by detecting negative language patterns (e.g., "stupid" or "useless").

[0476] 2. Certification and Selection

[0477] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict. Based on this, it selects a generative AI model with specific expertise and selects it as the appropriate arbitrator. For example, in the case of a technical discussion, a technically specialized GPT (generative AI model) is selected.

[0478] 3. Automatic Rewrite

[0479] The arbitration GPT rewrites offensive comments into polite language. For example, it converts a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?" The rewritten comment is sent to the terminal via the server and displayed on the user's screen.

[0480] 4. Facilitate

[0481] The arbitration GPT organizes the content of comments and identifies the root cause of the conflict. For example, in the case of a technical conflict, it will evenly summarize the advantages and disadvantages of the solution. It also generates empathetic comments to satisfy the user's desire for approval. For example, it generates a comment such as, "Your opinion is very interesting. I was particularly impressed with the part about XX." This is also sent to the device via the server and displayed on the user's screen.

[0482] 5. Reconciliation Advice

[0483] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0484] Specific examples

[0485] Case

[0486] If you encounter technical conflicts such as:

[0487] User A: This code is completely unusable. It's full of bugs.

[0488] User B: You're the one who doesn't understand anything.

[0489] Example

[0490] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0491] 2. The server selects a technically-specialized arbitrator, the GPT.

[0492] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0493] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0494] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0495] 4. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0496] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments.

[0497] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0498] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0499] 7. Arbitration GPT generates constructive advice towards a resolution.

[0500] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0501] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[0502] In this way, the system of the present invention can quickly and effectively resolve disputes and conflicts in online communities, providing a healthy communication environment.

[0503] The processing flow will be explained below.

[0504] Step 1:

[0505] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[0506] Step 2:

[0507] The server analyzes the detected offensive comments and uses a comment scoring algorithm to quantify their offensiveness. For example, if a score above a certain threshold is assigned, the comment is deemed definitively offensive.

[0508] Step 3:

[0509] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[0510] Step 4:

[0511] The arbitration GPT takes offensive comments and applies an algorithm to rewrite them in a more polite style, for example rewriting a comment like "Are you stupid?" to "Could you please elaborate a bit more on your point of view?"

[0512] Step 5:

[0513] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[0514] Step 6:

[0515] Arbitration GPT analyzes the background and arguments of the conflicting opinions and identifies the fundamental conflicts. For example, in the case of technical conflicts, it evenly summarizes the advantages and disadvantages of the conflicting solutions.

[0516] Step 7:

[0517] Arbitration GPT generates comments to foster empathy, such as "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?"

[0518] Step 8:

[0519] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[0520] Step 9:

[0521] Arbitration GPT generates neutral and constructive advice, such as "How about creating an opportunity to discuss each other's opinions in more detail?"

[0522] Step 10:

[0523] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[0524] Example 1

[0525] 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."

[0526] As online communities grow, the number of offensive comments and conflicts between users increases, leading to a deterioration in the communication environment. Conventional methods rely on manual moderation, making it difficult to respond quickly and effectively, so an automated system is needed to maintain the health of online communities.

[0527] 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.

[0528] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and conflicts, means for analyzing the detected comments to identify the subject of the conflict and select an appropriate generative AI model, means for automatically rewriting the offensive comments to a polite style using the selected generative AI model, means for the generative AI model to organize the conflicting opinions and generate empathetic comments, and means for generating neutral and constructive advice. This makes it possible to quickly and effectively resolve offensive comments and conflicts in online communities and provide a healthy communication environment.

[0529] An "online community" is a virtual space where people can exchange information and interact via the Internet.

[0530] "Monitoring" means to observe a specific object periodically or continuously to check its trends and status.

[0531] "Offensive comments" are statements that contain negative, insulting, or provocative content toward others.

[0532] "Conflict" refers to a clash or dispute that arises due to different opinions or values.

[0533] "Automatic detection" refers to the ability of a system to detect specific events based on certain conditions or patterns without human intervention.

[0534] A "generative AI model" is an artificial intelligence system that can perform generative tasks (such as text generation or translation).

[0535] "Polite writing style" is a characteristic of writing that uses gentle expressions that convey respect and consideration.

[0536] "Rewriting" means changing the form or wording of an original text while preserving its content.

[0537] An "empathetic comment" is a statement that shows understanding or agreement with the feelings or opinions of others.

[0538] "Constructive advice" refers to making specific and positive suggestions for solving problems or making improvements.

[0539] "Natural language processing technology" refers to technology that allows computers to understand, analyze, and generate human language.

[0540] This invention is a system for quickly and effectively resolving offensive comments and conflicts in online communities, which mainly involves the cooperation of three entities: a server, a terminal, and a user.

[0541] System configuration

[0542] Hardware and software used

[0543] Server: A high-performance server system for monitoring online communities in real time. Data is collected using a communication API.

[0544] Natural Language Processing (NLP): Analyze comment data using Google Cloud Natural Language and SpaCy.

[0545] Generative AI model: An AI system used for generative tasks. The model uses the technically specialized GPT (Generative Pre-trained Transformer).

[0546] Implementation method

[0547] 1. Monitoring

[0548] The server collects comment data from online communities via specific APIs (e.g., Twitter API or Reddit API). The collected data is periodically sent to the server and analyzed in real time using NLP technology.

[0549] 2. Detecting offensive comments

[0550] The server analyzes the collected comment data, searching for negative language patterns and specific keywords (e.g., "stupid," "useless," etc.), using a pre-trained machine learning model (e.g., the BERT model).

[0551] 3. Selecting a Generative AI Model

[0552] The server further analyzes the content of the detected comments to identify the topic of conflict, and then selects the most appropriate generative AI model (e.g., technically specialized GPT) based on the identified topic.

[0553] 4. Rewrite offensive comments

[0554] The server generates a prompt sentence that inputs the detected offensive comment into the arbitration GPT model. For example, a prompt sentence of the form "Please rewrite the following offensive comment in a polite style: 'Are you stupid?'" is created. The arbitration GPT generates the rewritten comment and sends it to the terminal (the user's display device) via the server.

[0555] 5. Generating empathetic comments and advice

[0556] The arbitration GPT analyzes the cause of the conflict and generates empathetic comments and neutral advice. For example, it might create a comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?" These comments are also sent to the terminal via the server and displayed to the user.

[0557] Specific examples

[0558] Case

[0559] If you encounter technical conflicts such as:

[0560] User A: This code is completely unusable. It's full of bugs.

[0561] User B: You're the one who doesn't understand anything.

[0562] Example

[0563] 1. The server monitors comments and detects offensive comments ("useless" or "you don't understand").

[0564] 2. The server selects a technically-specialized arbitrator GPT for the detected comment.

[0565] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0566] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0567] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0568] 4. The server sends the rewritten comments to the terminals and displays them on the screens of User A and User B.

[0569] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments:

[0570] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0571] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0572] 7. Arbitration GPT generates constructive advice towards resolution:

[0573] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0574] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[0575] As a result, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[0576] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0577] Step 1:

[0578] The server collects comment data from online communities. Specifically, it retrieves comment data in real time using a specific API (e.g., Twitter API or Reddit API). The input is the comment data retrieved from the API, and the output is the raw comment data to be analyzed.

[0579] Step 2:

[0580] The server applies natural language processing (NLP) techniques to the collected comment data to detect offensive or conflicting comments. Specifically, it performs text analysis using NLP tools (such as Google Cloud Natural Language or SpaCy). The input is raw comment data, and the output is comment data flagged as offensive or conflicting.

[0581] Step 3:

[0582] The server performs a detailed analysis of the content of the detected offensive comments and identifies the subject of the conflict. Specifically, it uses topic modeling techniques (such as LDA and CLIP) to identify what the conflict is about. The input is flagged comment data, and the output is data identifying the subject of the conflict.

[0583] Step 4:

[0584] The server selects an appropriate generative AI model (arbitration GPT) based on the identified subject of the conflict. The input is data identifying the subject of the conflict, and the output is information about the selected generative AI model.

[0585] Step 5:

[0586] The server generates a prompt for the selected generative AI model to rewrite the offensive comment. An example of a specific prompt is "Please rewrite the following offensive comment in a polite manner: 'Are you stupid?'" The input is the offensive comment and information about the selected generative AI model, and the output is the prompt.

[0587] Step 6:

[0588] The arbitration GPT receives the generated prompt and rewrites the offensive comments to a more polite style. The input is the prompt, and the output is the rewritten comment.

[0589] Step 7:

[0590] The server sends the rewritten comment to the terminal and displays it on the user's screen. The input is the rewritten comment, and the output is the rewritten comment displayed on the user's screen.

[0591] Step 8:

[0592] The arbitration GPT generates empathetic comments based on the rewritten comments and conflict analysis data. Specifically, it generates phrases that ease the conflict and deepen understanding between users. The inputs are the rewritten comments and conflict analysis data, and the output is empathetic comments.

[0593] Step 9:

[0594] The server sends the generated empathy comments to the terminal and displays them to the user. The input is the empathy comments, and the output is the empathy comments displayed on the user's screen.

[0595] Step 10:

[0596] The arbitration GPT ultimately generates constructive advice for resolving the conflict. Specifically, it suggests something like, "How about setting up an opportunity to discuss each other's opinions in more detail?" The inputs are conflict analysis data and the context of the dialogue, and the output is constructive advice.

[0597] Step 11:

[0598] The server sends the generated advice comment to the terminal and displays it on the user's screen. The advice comment is the input, and the advice comment displayed on the user's screen is obtained as the output.

[0599] (Application example 1)

[0600] 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."

[0601] Currently, offensive comments and conflicts are common in online community comment sections, hindering healthy communication. Applications used on smart devices, in particular, lack the ability to detect and appropriately address these issues in real time, resulting in a poor user experience. This invention aims to provide a system that quickly and effectively detects offensive comments and promotes constructive communication.

[0602] 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.

[0603] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for monitoring the comment section of an application used on a smart device to detect offensive comments in real time, and means for rewriting offensive comments to polite expressions and generating empathetic comments and constructive advice. This makes it possible to effectively suppress offensive behavior in the comment section and promote healthy and constructive communication.

[0604] An "online community" is a virtual space on the Internet where users can exchange opinions and share information.

[0605] "Monitoring" is the process by which a system continuously monitors all conversation threads in an online community to detect problematic comments in real time.

[0606] "Offensive comments" are comments that contain expressions or words intended to harm other users or that may cause conflict.

[0607] A "fight" refers to an emotional or aggressive conflict between users within an online community.

[0608] "Polite writing style" refers to the transformation of aggressive comments into expressions that respect others and allow for polite communication.

[0609] A "generative AI model" is an algorithm or system that has the ability to generate natural language using artificial intelligence technology.

[0610] An "empathetic comment" is a comment that includes expressions that empathize with the opinions and feelings of others and show positive and understanding.

[0611] "Neutral and constructive advice" is practical advice given from an impartial standpoint to resolve conflicts.

[0612] A "smart device" is a computerized portable device that can connect to the Internet and perform multiple functions. Examples include smartphones, smart glasses, and head-mounted displays.

[0613] "Real-time" means that data and information are processed almost immediately and results are obtained without any time lag.

[0614] "Polite expression" is language that is polite and carefully chosen so as not to offend others.

[0615] MODE FOR CARRYING OUT THE INVENTION

[0616] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. The system includes an application and a server that are run on smart devices.

[0617] Overall system configuration

[0618] The server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them to a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, and means for generating neutral and constructive advice. The server also includes means for monitoring comment sections in an application used on a smart device and detecting offensive comments in real time.

[0619] Hardware and Software Used

[0620] Hardware:

[0621] Smartphone

[0622] Smart Glasses

[0623] head-mounted display

[0624] software:

[0625] Natural language processing (NLP) models for monitoring (e.g., parsers such as Hugging Face's BERT)

[0626] Generative AI models (e.g., GPT models) for detecting and rewriting offensive comments

[0627] Data processing and calculation

[0628] The server monitors online community comments in real time and uses NLP technology, such as Hugging Face's "sentiment-analysis" pipeline, to detect offensive comments. Detected offensive comments generate appropriate prompts, which are then fed into a generative AI model (e.g., a GPT model). This generative AI model rewrites offensive comments into politer expressions and also generates empathetic comments and constructive advice.

[0629] Specific examples

[0630] Below is a concrete example of the process for detecting offensive comments, rewriting them, and generating empathetic comments and constructive advice.

[0631] Comment example

[0632] User A: "This video is completely pointless. It's a waste of time."

[0633] User B: "No one is listening to your opinion. Shut up."

[0634] Prompt Sentence Examples

[0635] "Rewrite the following offensive comment politely: This video is completely pointless. A waste of time."

[0636] "Create a neutral and empathetic response to the following opinion: No one is listening to your opinion. Shut up."

[0637] "Provide a constructive advice to resolve an online argument."

[0638] The server generates these prompts and inputs them into a generative AI model, which rewrites offensive comments into polite expressions and generates empathetic comments and constructive advice. The rewritten comments and advice are then displayed on the user's smart device screen, promoting healthy communication.

[0639] In this way, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[0640] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0641] Step 1:

[0642] The server monitors the comment section of an online community in real time. As input, comment data is provided from the online community. As output, offensive comments or comments that indicate signs of conflict are extracted. Specifically, the server uses a natural language processing (NLP) model to analyze the sentiment of the comments, for example, using Hugging Face's "sentiment-analysis" pipeline.

[0643] Step 2:

[0644] The server analyzes the extracted offensive comments. As input, it uses the offensive comments extracted in step 1. As output, it identifies the offensive elements and provides information to translate them into polite language. Specifically, the server runs an algorithm to identify negative words and phrases in the comments.

[0645] Step 3:

[0646] The server selects a generative AI model. As input, it uses the comment content and its characteristic information analyzed in step 2. As output, it selects an appropriate generative AI model (e.g., GPT model). Specifically, the server applies rule-based logic to select an appropriate AI model based on the comment content.

[0647] Step 4:

[0648] The server rewrites offensive comments into polite language. The selected generative AI model and the offensive comment are used as input. The rewritten, polite comment is generated as output. Specifically, the server generates a prompt sentence: "Rewrite the following offensive comment politely: [offensive comment]" and inputs it into the generative AI model.

[0649] Step 5:

[0650] The server sorts out the conflicting opinions and generates empathetic comments. The rewritten comments and the content of the conflict are used as input. The output is an empathetic comment. Specifically, the server generates a prompt statement: "Create a neutral and empathetic response to the following opinion: [content of the conflict]" and inputs this into the generative AI model.

[0651] Step 6:

[0652] The server generates neutral and constructive advice. Empathetic comments and conflict resolution information are used as input. Constructive advice is generated as output. Specifically, the server generates a prompt statement, "Provide a constructive advice to resolve an online argument," and inputs this into the generative AI model.

[0653] Step 7:

[0654] The server sends the generated rewrite comments, empathy comments, and constructive advice to the terminal. As input, the comments generated in steps 4, 5, and 6 are used. As output, rewrite comments, empathy comments, and constructive advice are generated and displayed on the terminal. Specifically, the server uses a protocol to combine each comment and send it to the user's screen.

[0655] 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.

[0656] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, and is further enhanced by combining it with an emotion engine. This system is implemented using the following process:

[0657] System configuration

[0658] 1. Monitoring and Detection

[0659] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect and quantify comments containing offensive words or phrases.

[0660] 2. Certification and Selection

[0661] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[0662] 3. Emotion recognition

[0663] The emotion engine detects emotions from the user's text messages. It also uses physiological indicators (e.g., heart rate, sweat rate, blink rate) to recognize the user's emotions. For example, if it detects negative emotions (e.g., anger, sadness, irritation), it uses that information for the next step.

[0664] 4. Automatic Rewrite

[0665] The arbitration GPT receives offensive comments and rewrites them in a more polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[0666] 5. Facilitate

[0667] The arbitration GPT sorts through the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This is also sent to the device via the server and displayed on the user's screen.

[0668] 6. Reconciliation Advice

[0669] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0670] Specific examples

[0671] Case

[0672] If you encounter technical conflicts such as:

[0673] User A: This code is completely unusable. It's full of bugs.

[0674] User B: You're the one who doesn't understand anything.

[0675] Example

[0676] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0677] 2. The server selects a technically-specialized arbitrator, the GPT.

[0678] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[0679] 4. Arbitration GPT rewrites the above comment as follows:

[0680] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0681] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0682] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0683] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[0684] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0685] 7. The server sends the empathy comment to the terminal and it is displayed on the user's screen.

[0686] 8. Arbitration GPT generates constructive advice towards a resolution.

[0687] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0688] 9. The server sends the advice comment to the terminal and it is displayed on the user's screen.

[0689] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[0690] The processing flow will be explained below.

[0691] Step 1:

[0692] The server monitors all conversation threads in online communities in real time, using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[0693] Step 2:

[0694] The server analyzes the detected offensive comments and quantifies their offensiveness. Specifically, if a score exceeding a certain threshold is assigned, the comment is deemed to be definitively offensive.

[0695] Step 3:

[0696] The server analyzes the content of detected offensive comments and identifies the subject of the conflict, categorizing them into categories such as technical arguments, personal attacks, and political conflicts.

[0697] Step 4:

[0698] Based on the identified subject, the server selects a generative AI model (GPT) with domain expertise. For example, in the case of a technical discussion, a technical GPT is selected.

[0699] Step 5:

[0700] The emotion engine analyzes the user's text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.) to recognize the user's emotions. For example, it can determine that User A is feeling "anger" based on text analysis.

[0701] Step 6:

[0702] The arbitration GPT applies an algorithm to rewrite offensive comments to a more polite style, taking into account the results of the sentiment engine and rewriting appropriately. For example, it transforms a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?"

[0703] Step 7:

[0704] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[0705] Step 8:

[0706] The arbitration GPT identifies the root cause of the conflict based on the content of the comments and the results of the sentiment engine. For example, in a technical conflict, it will equally explain the advantages and disadvantages of each solution.

[0707] Step 9:

[0708] The arbitration GPT generates comments to foster empathy. Based on the emotional information detected by the emotion engine, it generates more appropriate empathetic comments. For example, it might say, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?"

[0709] Step 10:

[0710] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[0711] Step 11:

[0712] The arbitration GPT generates neutral and constructive advice, taking into account the information detected by the emotion engine and proposing a solution that both parties can agree on. For example, it may provide advice such as, "How about setting up an opportunity to discuss your respective opinions in more detail?"

[0713] Step 12:

[0714] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[0715] Example 2

[0716] 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."

[0717] In online communities, aggressive comments and arguments are common and often impede smooth communication. This increases the psychological burden on users and worsens the overall atmosphere of the community. Furthermore, emotional conflicts often deepen, preventing constructive discussions. Conventional automated moderation systems are limited to simply filtering or deleting words, and do not fundamentally address the problem. Therefore, this invention aims to resolve conflicts quickly and effectively by recognizing emotions and promoting appropriate communication.

[0718] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for monitoring an online community to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them in a polite style; means for selecting a generative AI model with specialized knowledge; means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model; means for generating neutral and constructive advice; means for recognizing emotions using a user's text message and physiological indicators; and means for rewriting offensive comments based on the emotion recognition results. This makes it possible to rewrite offensive comments and generate appropriate mediating comments taking emotions into consideration, thereby enabling disputes in online communities to be resolved quickly and effectively.

[0719] An "online community" is a virtual space where people can interact using the Internet.

[0720] "Offensive comments" are comments that contain words or phrases intended to insult or harm others.

[0721] A "fight" is a dispute that arises from a clash of opinions or feelings within an online community.

[0722] "Emotion recognition means" refers to technology that analyzes a user's text messages and physiological indicators to detect their emotional state.

[0723] "Polite rewriting" refers to an algorithm that automatically converts offensive comments into more readable and polite language.

[0724] A "generative AI model" is a model that uses artificial intelligence to generate text, particularly using deep learning techniques.

[0725] "Means for sorting out conflicting opinions and generating empathetic comments" refers to the process of analyzing conflicting opinions and generating comments that show empathy for those opinions.

[0726] "Means for generating neutral and constructive advice" refers to technology that automatically generates neutral and helpful suggestions to defuse conflicts.

[0727] "Physiological indicators" refer to data that indicate the user's physical condition, such as heart rate, sweat rate, and blink rate.

[0728] This invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. Specifically, it is more effective when combined with an emotion engine. This system is implemented using the following process:

[0729] Monitoring and Detection

[0730] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques. A specific implementation example uses Apache Kafka to process large volumes of messages in real time and performs analysis using the SpaCy library. It detects comments containing offensive words and phrases and quantifies them. This process allows problematic comments to be identified immediately.

[0731] Certification and Selection

[0732] The server analyzes the content of detected offensive comments to identify the subject of the conflict (technical discussion, personal attack, political conflict, etc.). This analysis uses a topic modeling algorithm (e.g., LDA) to classify the content of the comment. Based on the identified subject, the server selects the most appropriate generative AI model (e.g., GPT-3).

[0733] emotion recognition

[0734] The emotion engine detects emotions from users' text messages. This process uses the Sentiment Analysis API to classify emotions from text. Additionally, physiological indicators (heart rate, sweat rate, blink rate, etc.) are also collected using sensors such as Arduino and Raspberry Pi.

[0735] Automatic Rewrite

[0736] The arbitration GPT receives offensive comments and rewrites them in a polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[0737] Facilitate

[0738] The arbitration GPT organizes the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This empathetic comment is also sent to the device via the server and displayed on the user's screen.

[0739] Reconciliation advice

[0740] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0741] Specific examples

[0742] An example of a system in place in the event of a technical conflict

[0743] User A: This code is completely unusable. It's full of bugs.

[0744] User B: You're the one who doesn't understand anything.

[0745] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0746] 2. The server selects a technically specialized generative AI model.

[0747] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[0748] 4. Arbitration GPT rewrites the above comment as follows:

[0749] User A (after rewrite): This code looks a bit problematic. It might need some fixing.

[0750] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0751] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0752] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[0753] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0754] 7. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0755] 8. Arbitration GPT generates constructive advice towards a resolution.

[0756] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0757] 9. The server sends the advice comment to the terminal and displays it on the user's screen.

[0758] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[0759] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0760] Step 1:

[0761] The server collects all conversation threads of an online community.

[0762] Input: New message in the community

[0763] Processing: Stream messages in real time using Apache Kafka and store them in a database.

[0764] Output: All saved message logs

[0765] Step 2:

[0766] The server analyzes the collected messages using natural language processing (NLP) technology to detect offensive words and phrases.

[0767] Input: Saved message log

[0768] Processing: Parse the message text using the SpaCy library and match it against a list of negative words

[0769] How it works: NLP analysis detects comments like "Are you stupid?" and records them as a numerical value

[0770] Output: List of offensive comments

[0771] Step 3:

[0772] The server analyzes the content of the detected offensive comments to identify the subject of the conflict.

[0773] Input: offensive comment list

[0774] Processing: Classify comments using a topic modeling algorithm (e.g. LDA)

[0775] Behavior: Categorized into categories such as "technical debate," "personal attacks," and "political conflict"

[0776] Output: Categorized offensive comments

[0777] Step 4:

[0778] Based on the identified subject, the server selects a generative AI model in that field.

[0779] Input: Categorized offensive comment

[0780] Processing: Select a generative AI model (e.g., GPT-3) appropriate for each subject

[0781] How it works: For technical discussions, we select a technically specialized GPT-3 model.

[0782] Output: The selected generative AI model

[0783] Step 5:

[0784] The emotion engine uses the user's text messages and physiological indicators to recognize emotions.

[0785] Input: User text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.)

[0786] Processing: Detect emotional states using Sentiment Analysis API and physiological data sensors (Arduino, Raspberry Pi, etc.)

[0787] Behavior: Classifying emotions such as "anger" and "sadness" from text messages and integrating them with physiological indicators

[0788] Output: Emotion recognition result

[0789] Step 6:

[0790] The server uses a generative AI model to rewrite offensive comments based on the emotional information detected by the emotion engine.

[0791] Input: offensive comments and emotion recognition results

[0792] Processing: Generate prompts for rewriting and send them to the generative AI model

[0793] Behavior: Transform "Are you stupid?" into a more polite statement such as "Can you explain your point of view in more detail?"

[0794] Output: Rewritten comments

[0795] Step 7:

[0796] The server sends the rewritten comment to the terminal and displays it on the user's screen.

[0797] Input: Rewritten comment

[0798] Processing: Send the rewritten comment to the device using the REST API

[0799] How it works: Rewritten comments are displayed in real time on the chat screen.

[0800] Output: Rewrite comment displayed to the user

[0801] Step 8:

[0802] The server uses a generative AI model to generate empathetic comments based on the information provided by the emotion engine.

[0803] Input: Emotion recognition results

[0804] Processing: Generate prompts for generating empathetic comments and send them to the generative AI model

[0805] What it does: Generates empathetic comments like, "It's really important to calmly discuss bugs in your code. Can you share your perspectives on the issues?"

[0806] Output: Empathetic comments

[0807] Step 9:

[0808] The server sends the sympathetic comments to the terminal and displays them on the user's screen.

[0809] Input: Sympathetic comment

[0810] Processing: Send sympathetic comments to the device using the REST API

[0811] How it works: Show empathetic comments in real time on the chat screen

[0812] Output: Empathy comments displayed on the user's screen

[0813] Step 10:

[0814] The server uses arbitration GPT to generate neutral and constructive advice.

[0815] Input: Emotion recognition results and conflict content

[0816] Processing: Generate prompts for advice generation and send them to the generative AI model

[0817] Behavior: Generates constructive advice such as, "How about we set up an opportunity to discuss each other's opinions in more detail?"

[0818] Output: Advisory comments

[0819] Step 11:

[0820] The server sends the advice comment to the terminal and displays it on the user's screen.

[0821] Input: Advisory Comment

[0822] Processing: Sends advice to the device using the REST API

[0823] How it works: Real-time advice is displayed on the chat screen.

[0824] Output: Advisory comments displayed on the user's screen

[0825] This processing flow enables aggressive comments and disputes in online communities to be resolved quickly and effectively, providing a healthy communication environment that also takes into consideration the feelings of users.

[0826] (Application example 2)

[0827] 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."

[0828] In traditional online communities, aggressive comments and arguments are common, often damaging the user experience. Even in brick-and-mortar stores, communication between customers and staff, or between staff members, can be disrupted, leading to lower customer satisfaction and reduced staff efficiency. A system that can quickly and effectively resolve these issues and promote healthy communication is needed.

[0829] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for applying the monitored communication content to customer service and staff dialogue in physical stores, means for providing real-time filtering and rewriting of offensive comments via a smartphone, means for visualizing the emotions of customers and staff using an emotion engine, and means for displaying empathetic comments and constructive advice in the physical store environment in real time. This makes it possible to provide a healthy communication environment in online communities and physical stores.

[0830] An "online community" is a place where users can interact with each other and share information via the Internet.

[0831] "Monitoring" is the act of observing and monitoring specific situations or data in real time.

[0832] "Offensive comments" are statements that contain the intent to insult or attack others.

[0833] A "quarrel" refers to a clash caused by an argument or a difference of opinion.

[0834] "Rewriting" is the act of reconstructing a sentence and rewriting it in a different way.

[0835] A "generative AI model" is an artificial intelligence model that is trained to perform generative tasks.

[0836] "Selection" is the act of choosing an appropriate subject.

[0837] An "empathetic comment" is a statement made to show empathy for the feelings or opinions of others.

[0838] "Neutral and constructive advice" is advice that is unbiased and aims to lead in a positive direction.

[0839] A "system" is a collection of elements organized to achieve a particular purpose.

[0840] A "physical store" is a store or service facility that exists in a physical location.

[0841] "Customer service" refers to the act of responding to customer requests and questions.

[0842] "Inter-staff dialogue" refers to communication between employees who belong to the same workplace.

[0843] A "smartphone" is a mobile device that has similar communication capabilities to a mobile phone, but with more advanced computing capabilities.

[0844] An "emotion engine" is a technology that analyzes and recognizes a user's emotions from text and physiological data.

[0845] "Real-time filtering" is the process of instantly analyzing data or information and sorting it based on specific criteria.

[0846] "Visualization" is the process of expressing abstract data or information in a visually easy-to-understand form.

[0847] The present invention provides a system for monitoring communications in online communities and brick-and-mortar stores and effectively resolving offensive comments and conflicts. Hereinafter, embodiments of the present invention will be described in detail.

[0848] System Program

[0849] The system of the present invention is implemented using the following hardware and software:

[0850] Hardware: Smartphones, servers

[0851] Software: Natural language processing (NLP) libraries (e.g., spaCy), sentiment analysis tools (e.g., AWS Comprehend, Google Cloud Natural Language API), generative AI models (e.g., OpenAI GPT)

[0852] Process Overview

[0853] Monitoring and Detection

[0854] The server monitors communication content in physical stores and online communities in real time. This monitoring includes using NLP technology to detect offensive words and phrases. For example, if a user types an offensive comment such as "it's unusable" into their smartphone, the content will be detected immediately.

[0855] Comment Rewrite

[0856] Detected offensive comments are automatically rewritten to a more polite style using a generative AI model. The server inputs appropriate prompts to the generative AI model and receives the rewritten comments. For example, a comment such as "The service on this product was poor!" is rewritten to "There seems to be some room for improvement in the service on this product."

[0857] Visualizing emotions

[0858] The server uses an emotion engine to analyze the user's emotions and notify customers and staff of negative emotions. The results of the emotion analysis are visualized and displayed in real time on smartphones and other devices. For example, if "anger" is detected based on a customer's text and physiological data, that information is displayed on the staff's device.

[0859] Empathy comment generation

[0860] The selected generative AI model generates empathetic comments based on the detected emotional information, which promotes conflict resolution. The server reads the output of the emotion engine and inputs a prompt sentence to generate an empathetic comment. For example, the prompt sentence used is "Generate an empathetic comment for the following message: The service on this product was poor!"

[0861] Providing constructive advice

[0862] Neutral and constructive advice is generated by the generative AI model and provided to store staff and customers in real time, enabling all parties involved to find a better solution. The server inputs prompts offering constructive advice based on the content of the conflict, and the generated advice is displayed on the device.

[0863] Specific examples

[0864] Handling customer complaints

[0865] If a customer says, "The service on this product was poor!"

[0866] 1. The server detects this offensive comment.

[0867] 2. Using the generative AI model, rewrite the sentence to "There seems to be some room for improvement in how this product is handled."

[0868] 3. The emotion engine detects anger and notifies staff.

[0869] 4. The server inputs the prompt text "Generate an empathetic comment for the following message: The service on this product was poor!" and generates an empathetic comment.

[0870] 5. Provide the staff with a sympathetic comment saying, "We take your feedback seriously and will strive to improve," along with constructive advice saying, "We would appreciate it if you could tell us the specific issues you had."

[0871] This will enable us to provide a healthy communication environment in our physical stores and online communities.

[0872] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0873] Step 1:

[0874] The server monitors communication content in physical stores and online communities. The input includes user text messages, and the data is analyzed using NLP techniques. The output detects messages containing offensive comments or negative language.

[0875] Step 2:

[0876] The server inputs the detected offensive comments into a generative AI model, which rewrites them in a polite style. The input includes an offensive text message, which is passed to the generative AI model along with a prompt. The output is a rewritten, polite comment. For example, a comment like "The service on this product was poor!" is changed to "There seems to be some room for improvement in the service on this product."

[0877] Step 3:

[0878] The server uses an emotion engine to analyze emotions from the user's text message and physiological indicators (e.g., heart rate, sweat rate, blink rate). The input includes the rewritten text message and physiological data. The output is the emotion analysis result, which detects emotions such as "anger" or "sadness."

[0879] Step 4:

[0880] The server generates an empathetic comment based on the analysis results of the emotion engine. The input includes the emotion analysis results and the rewritten text message, which are passed to the generative AI model along with a prompt. The output is a comment expressing empathy. For example, a prompt such as "Generate an empathetic comment for the following message: The service on this product was poor!" is used.

[0881] Step 5:

[0882] The server generates neutral and constructive advice. The input includes the sentiment analysis results and empathetic comments, which are passed to the generative AI model along with a prompt. The output is constructive advice. For example, the generated advice might be, "We would appreciate it if you could tell us the specific problem you are having."

[0883] Step 6:

[0884] The terminal displays the rewritten comments, empathetic comments, and constructive advice sent from the server to the user. The input includes the rewritten comments, empathetic comments, and advice from the server. The output is displayed on the user's screen, promoting healthy communication.

[0885] Step 7:

[0886] The user can smoothly communicate based on the rewritten comments, empathetic comments, and constructive advice displayed on the terminal. The input includes the content displayed on the terminal, and the user's response and behavior are obtained as the output.

[0887] 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.

[0888] 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.

[0889] 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.

[0890] [Third embodiment]

[0891] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0892] 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.

[0893] 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).

[0894] 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.

[0895] 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.

[0896] 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).

[0897] 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.

[0898] 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.

[0899] 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.

[0900] 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.

[0901] 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.

[0902] 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."

[0903] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, which is implemented through the following process.

[0904] System configuration

[0905] 1. Monitoring and Detection

[0906] The server monitors all conversation threads in online communities in real time and automatically detects offensive comments and signs of conflict using natural language processing (NLP) technology, specifically by detecting negative language patterns (e.g., "stupid" or "useless").

[0907] 2. Certification and Selection

[0908] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict. Based on this, it selects a generative AI model with specific expertise and selects it as the appropriate arbitrator. For example, in the case of a technical discussion, a technically specialized GPT (generative AI model) is selected.

[0909] 3. Automatic Rewrite

[0910] The arbitration GPT rewrites offensive comments into polite language. For example, it converts a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?" The rewritten comment is sent to the terminal via the server and displayed on the user's screen.

[0911] 4. Facilitate

[0912] The arbitration GPT organizes the content of comments and identifies the root cause of the conflict. For example, in the case of a technical conflict, it will evenly summarize the advantages and disadvantages of the solution. It also generates empathetic comments to satisfy the user's desire for approval. For example, it generates a comment such as, "Your opinion is very interesting. I was particularly impressed with the part about XX." This is also sent to the device via the server and displayed on the user's screen.

[0913] 5. Reconciliation Advice

[0914] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[0915] Specific examples

[0916] Case

[0917] If you encounter technical conflicts such as:

[0918] User A: This code is completely unusable. It's full of bugs.

[0919] User B: You're the one who doesn't understand anything.

[0920] Example

[0921] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[0922] 2. The server selects a technically-specialized arbitrator, the GPT.

[0923] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0924] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0925] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0926] 4. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[0927] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments.

[0928] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[0929] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[0930] 7. Arbitration GPT generates constructive advice towards a resolution.

[0931] "How about we create an opportunity to discuss each other's opinions in more detail?"

[0932] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[0933] In this way, the system of the present invention can quickly and effectively resolve disputes and conflicts in online communities, providing a healthy communication environment.

[0934] The processing flow will be explained below.

[0935] Step 1:

[0936] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[0937] Step 2:

[0938] The server analyzes the detected offensive comments and uses a comment scoring algorithm to quantify their offensiveness. For example, if a score above a certain threshold is assigned, the comment is deemed definitively offensive.

[0939] Step 3:

[0940] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[0941] Step 4:

[0942] The arbitration GPT takes offensive comments and applies an algorithm to rewrite them in a more polite style, for example rewriting a comment like "Are you stupid?" to "Could you please elaborate a bit more on your point of view?"

[0943] Step 5:

[0944] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[0945] Step 6:

[0946] Arbitration GPT analyzes the background and arguments of the conflicting opinions and identifies the fundamental conflicts. For example, in the case of technical conflicts, it evenly summarizes the advantages and disadvantages of the conflicting solutions.

[0947] Step 7:

[0948] Arbitration GPT generates comments to foster empathy, such as "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?"

[0949] Step 8:

[0950] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[0951] Step 9:

[0952] Arbitration GPT generates neutral and constructive advice, such as "How about creating an opportunity to discuss each other's opinions in more detail?"

[0953] Step 10:

[0954] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[0955] Example 1

[0956] 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."

[0957] As online communities grow, the number of offensive comments and conflicts between users increases, leading to a deterioration in the communication environment. Conventional methods rely on manual moderation, making it difficult to respond quickly and effectively, so an automated system is needed to maintain the health of online communities.

[0958] 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.

[0959] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and conflicts, means for analyzing the detected comments to identify the subject of the conflict and select an appropriate generative AI model, means for automatically rewriting the offensive comments to a polite style using the selected generative AI model, means for the generative AI model to organize the conflicting opinions and generate empathetic comments, and means for generating neutral and constructive advice. This makes it possible to quickly and effectively resolve offensive comments and conflicts in online communities and provide a healthy communication environment.

[0960] An "online community" is a virtual space where people can exchange information and interact via the Internet.

[0961] "Monitoring" means to observe a specific object periodically or continuously to check its trends and status.

[0962] "Offensive comments" are statements that contain negative, insulting, or provocative content toward others.

[0963] "Conflict" refers to a clash or dispute that arises due to different opinions or values.

[0964] "Automatic detection" refers to the ability of a system to detect specific events based on certain conditions or patterns without human intervention.

[0965] A "generative AI model" is an artificial intelligence system that can perform generative tasks (such as text generation or translation).

[0966] "Polite writing style" is a characteristic of writing that uses gentle expressions that convey respect and consideration.

[0967] "Rewriting" means changing the form or wording of an original text while preserving its content.

[0968] An "empathetic comment" is a statement that shows understanding or agreement with the feelings or opinions of others.

[0969] "Constructive advice" refers to making specific and positive suggestions for solving problems or making improvements.

[0970] "Natural language processing technology" refers to technology that allows computers to understand, analyze, and generate human language.

[0971] This invention is a system for quickly and effectively resolving offensive comments and conflicts in online communities, which mainly involves the cooperation of three entities: a server, a terminal, and a user.

[0972] System configuration

[0973] Hardware and software used

[0974] Server: A high-performance server system for monitoring online communities in real time. Data is collected using a communication API.

[0975] Natural Language Processing (NLP): Analyze comment data using Google Cloud Natural Language and SpaCy.

[0976] Generative AI model: An AI system used for generative tasks. The model uses the technically specialized GPT (Generative Pre-trained Transformer).

[0977] Implementation method

[0978] 1. Monitoring

[0979] The server collects comment data from online communities via specific APIs (e.g., Twitter API or Reddit API). The collected data is periodically sent to the server and analyzed in real time using NLP technology.

[0980] 2. Detecting offensive comments

[0981] The server analyzes the collected comment data, searching for negative language patterns and specific keywords (e.g., "stupid," "useless," etc.), using a pre-trained machine learning model (e.g., the BERT model).

[0982] 3. Selecting a Generative AI Model

[0983] The server further analyzes the content of the detected comments to identify the topic of conflict, and then selects the most appropriate generative AI model (e.g., technically specialized GPT) based on the identified topic.

[0984] 4. Rewrite offensive comments

[0985] The server generates a prompt sentence that inputs the detected offensive comment into the arbitration GPT model. For example, a prompt sentence of the form "Please rewrite the following offensive comment in a polite style: 'Are you stupid?'" is created. The arbitration GPT generates the rewritten comment and sends it to the terminal (the user's display device) via the server.

[0986] 5. Generating empathetic comments and advice

[0987] The arbitration GPT analyzes the cause of the conflict and generates empathetic comments and neutral advice. For example, it might create a comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?" These comments are also sent to the terminal via the server and displayed to the user.

[0988] Specific examples

[0989] Case

[0990] If you encounter technical conflicts such as:

[0991] User A: This code is completely unusable. It's full of bugs.

[0992] User B: You're the one who doesn't understand anything.

[0993] Example

[0994] 1. The server monitors comments and detects offensive comments ("useless" or "you don't understand").

[0995] 2. The server selects a technically-specialized arbitrator GPT for the detected comment.

[0996] 3. Arbitration GPT automatically rewrites the above comment as follows:

[0997] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[0998] User B (after rewrite): Why don't we both take a look at the code a bit more?

[0999] 4. The server sends the rewritten comments to the terminals and displays them on the screens of User A and User B.

[1000] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments:

[1001] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1002] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[1003] 7. Arbitration GPT generates constructive advice towards resolution:

[1004] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1005] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[1006] As a result, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[1007] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1008] Step 1:

[1009] The server collects comment data from online communities. Specifically, it retrieves comment data in real time using a specific API (e.g., Twitter API or Reddit API). The input is the comment data retrieved from the API, and the output is the raw comment data to be analyzed.

[1010] Step 2:

[1011] The server applies natural language processing (NLP) techniques to the collected comment data to detect offensive or conflicting comments. Specifically, it performs text analysis using NLP tools (such as Google Cloud Natural Language or SpaCy). The input is raw comment data, and the output is comment data flagged as offensive or conflicting.

[1012] Step 3:

[1013] The server performs a detailed analysis of the content of the detected offensive comments and identifies the subject of the conflict. Specifically, it uses topic modeling techniques (such as LDA and CLIP) to identify what the conflict is about. The input is flagged comment data, and the output is data identifying the subject of the conflict.

[1014] Step 4:

[1015] The server selects an appropriate generative AI model (arbitration GPT) based on the identified subject of the conflict. The input is data identifying the subject of the conflict, and the output is information about the selected generative AI model.

[1016] Step 5:

[1017] The server generates a prompt for the selected generative AI model to rewrite the offensive comment. An example of a specific prompt is "Please rewrite the following offensive comment in a polite manner: 'Are you stupid?'" The input is the offensive comment and information about the selected generative AI model, and the output is the prompt.

[1018] Step 6:

[1019] The arbitration GPT receives the generated prompt and rewrites the offensive comments to a more polite style. The input is the prompt, and the output is the rewritten comment.

[1020] Step 7:

[1021] The server sends the rewritten comment to the terminal and displays it on the user's screen. The input is the rewritten comment, and the output is the rewritten comment displayed on the user's screen.

[1022] Step 8:

[1023] The arbitration GPT generates empathetic comments based on the rewritten comments and conflict analysis data. Specifically, it generates phrases that ease the conflict and deepen understanding between users. The inputs are the rewritten comments and conflict analysis data, and the output is empathetic comments.

[1024] Step 9:

[1025] The server sends the generated empathy comments to the terminal and displays them to the user. The input is the empathy comments, and the output is the empathy comments displayed on the user's screen.

[1026] Step 10:

[1027] The arbitration GPT ultimately generates constructive advice for resolving the conflict. Specifically, it suggests something like, "How about setting up an opportunity to discuss each other's opinions in more detail?" The inputs are conflict analysis data and the context of the dialogue, and the output is constructive advice.

[1028] Step 11:

[1029] The server sends the generated advice comment to the terminal and displays it on the user's screen. The advice comment is the input, and the advice comment displayed on the user's screen is obtained as the output.

[1030] (Application example 1)

[1031] 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."

[1032] Currently, offensive comments and conflicts are common in online community comment sections, hindering healthy communication. Applications used on smart devices, in particular, lack the ability to detect and appropriately address these issues in real time, resulting in a poor user experience. This invention aims to provide a system that quickly and effectively detects offensive comments and promotes constructive communication.

[1033] 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.

[1034] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for monitoring the comment section of an application used on a smart device to detect offensive comments in real time, and means for rewriting offensive comments to polite expressions and generating empathetic comments and constructive advice. This makes it possible to effectively suppress offensive behavior in the comment section and promote healthy and constructive communication.

[1035] An "online community" is a virtual space on the Internet where users can exchange opinions and share information.

[1036] "Monitoring" is the process by which a system continuously monitors all conversation threads in an online community to detect problematic comments in real time.

[1037] "Offensive comments" are comments that contain expressions or words intended to harm other users or that may cause conflict.

[1038] A "fight" refers to an emotional or aggressive conflict between users within an online community.

[1039] "Polite writing style" refers to the transformation of aggressive comments into expressions that respect others and allow for polite communication.

[1040] A "generative AI model" is an algorithm or system that has the ability to generate natural language using artificial intelligence technology.

[1041] An "empathetic comment" is a comment that includes expressions that empathize with the opinions and feelings of others and show positive and understanding.

[1042] "Neutral and constructive advice" is practical advice given from an impartial standpoint to resolve conflicts.

[1043] A "smart device" is a computerized portable device that can connect to the Internet and perform multiple functions. Examples include smartphones, smart glasses, and head-mounted displays.

[1044] "Real-time" means that data and information are processed almost immediately and results are obtained without any time lag.

[1045] "Polite expression" is language that is polite and carefully chosen so as not to offend others.

[1046] MODE FOR CARRYING OUT THE INVENTION

[1047] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. The system includes an application and a server that are run on smart devices.

[1048] Overall system configuration

[1049] The server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them to a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, and means for generating neutral and constructive advice. The server also includes means for monitoring comment sections in an application used on a smart device and detecting offensive comments in real time.

[1050] Hardware and Software Used

[1051] Hardware:

[1052] Smartphone

[1053] Smart Glasses

[1054] head-mounted display

[1055] software:

[1056] Natural language processing (NLP) models for monitoring (e.g., parsers such as Hugging Face's BERT)

[1057] Generative AI models (e.g., GPT models) for detecting and rewriting offensive comments

[1058] Data processing and calculation

[1059] The server monitors online community comments in real time and uses NLP technology, such as Hugging Face's "sentiment-analysis" pipeline, to detect offensive comments. Detected offensive comments generate appropriate prompts, which are then fed into a generative AI model (e.g., a GPT model). This generative AI model rewrites offensive comments into politer expressions and also generates empathetic comments and constructive advice.

[1060] Specific examples

[1061] Below is a concrete example of the process for detecting offensive comments, rewriting them, and generating empathetic comments and constructive advice.

[1062] Comment example

[1063] User A: "This video is completely pointless. It's a waste of time."

[1064] User B: "No one is listening to your opinion. Shut up."

[1065] Prompt Sentence Examples

[1066] "Rewrite the following offensive comment politely: This video is completely pointless. A waste of time."

[1067] "Create a neutral and empathetic response to the following opinion: No one is listening to your opinion. Shut up."

[1068] "Provide a constructive advice to resolve an online argument."

[1069] The server generates these prompts and inputs them into a generative AI model, which rewrites offensive comments into polite expressions and generates empathetic comments and constructive advice. The rewritten comments and advice are then displayed on the user's smart device screen, promoting healthy communication.

[1070] In this way, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[1071] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1072] Step 1:

[1073] The server monitors the comment section of an online community in real time. As input, comment data is provided from the online community. As output, offensive comments or comments that indicate signs of conflict are extracted. Specifically, the server uses a natural language processing (NLP) model to analyze the sentiment of the comments, for example, using Hugging Face's "sentiment-analysis" pipeline.

[1074] Step 2:

[1075] The server analyzes the extracted offensive comments. As input, it uses the offensive comments extracted in step 1. As output, it identifies the offensive elements and provides information to translate them into polite language. Specifically, the server runs an algorithm to identify negative words and phrases in the comments.

[1076] Step 3:

[1077] The server selects a generative AI model. As input, it uses the comment content and its characteristic information analyzed in step 2. As output, it selects an appropriate generative AI model (e.g., GPT model). Specifically, the server applies rule-based logic to select an appropriate AI model based on the comment content.

[1078] Step 4:

[1079] The server rewrites offensive comments into polite language. The selected generative AI model and the offensive comment are used as input. The rewritten, polite comment is generated as output. Specifically, the server generates a prompt sentence: "Rewrite the following offensive comment politely: [offensive comment]" and inputs it into the generative AI model.

[1080] Step 5:

[1081] The server sorts out the conflicting opinions and generates empathetic comments. The rewritten comments and the content of the conflict are used as input. The output is an empathetic comment. Specifically, the server generates a prompt statement: "Create a neutral and empathetic response to the following opinion: [content of the conflict]" and inputs this into the generative AI model.

[1082] Step 6:

[1083] The server generates neutral and constructive advice. Empathetic comments and conflict resolution information are used as input. Constructive advice is generated as output. Specifically, the server generates a prompt statement, "Provide a constructive advice to resolve an online argument," and inputs this into the generative AI model.

[1084] Step 7:

[1085] The server sends the generated rewrite comments, empathy comments, and constructive advice to the terminal. As input, the comments generated in steps 4, 5, and 6 are used. As output, rewrite comments, empathy comments, and constructive advice are generated and displayed on the terminal. Specifically, the server uses a protocol to combine each comment and send it to the user's screen.

[1086] 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.

[1087] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, and is further enhanced by combining it with an emotion engine. This system is implemented using the following process:

[1088] System configuration

[1089] 1. Monitoring and Detection

[1090] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect and quantify comments containing offensive words or phrases.

[1091] 2. Certification and Selection

[1092] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[1093] 3. Emotion recognition

[1094] The emotion engine detects emotions from the user's text messages. It also uses physiological indicators (e.g., heart rate, sweat rate, blink rate) to recognize the user's emotions. For example, if it detects negative emotions (e.g., anger, sadness, irritation), it uses that information for the next step.

[1095] 4. Automatic Rewrite

[1096] The arbitration GPT receives offensive comments and rewrites them in a more polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[1097] 5. Facilitate

[1098] The arbitration GPT sorts through the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This is also sent to the device via the server and displayed on the user's screen.

[1099] 6. Reconciliation Advice

[1100] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[1101] Specific examples

[1102] Case

[1103] If you encounter technical conflicts such as:

[1104] User A: This code is completely unusable. It's full of bugs.

[1105] User B: You're the one who doesn't understand anything.

[1106] Example

[1107] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[1108] 2. The server selects a technically-specialized arbitrator, the GPT.

[1109] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[1110] 4. Arbitration GPT rewrites the above comment as follows:

[1111] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[1112] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1113] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[1114] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[1115] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1116] 7. The server sends the empathy comment to the terminal and it is displayed on the user's screen.

[1117] 8. Arbitration GPT generates constructive advice towards a resolution.

[1118] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1119] 9. The server sends the advice comment to the terminal and it is displayed on the user's screen.

[1120] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[1121] The processing flow will be explained below.

[1122] Step 1:

[1123] The server monitors all conversation threads in online communities in real time, using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[1124] Step 2:

[1125] The server analyzes the detected offensive comments and quantifies their offensiveness. Specifically, if a score exceeding a certain threshold is assigned, the comment is deemed to be definitively offensive.

[1126] Step 3:

[1127] The server analyzes the content of detected offensive comments and identifies the subject of the conflict, categorizing them into categories such as technical arguments, personal attacks, and political conflicts.

[1128] Step 4:

[1129] Based on the identified subject, the server selects a generative AI model (GPT) with domain expertise. For example, in the case of a technical discussion, a technical GPT is selected.

[1130] Step 5:

[1131] The emotion engine analyzes the user's text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.) to recognize the user's emotions. For example, it can determine that User A is feeling "anger" based on text analysis.

[1132] Step 6:

[1133] The arbitration GPT applies an algorithm to rewrite offensive comments to a more polite style, taking into account the results of the sentiment engine and rewriting appropriately. For example, it transforms a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?"

[1134] Step 7:

[1135] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[1136] Step 8:

[1137] The arbitration GPT identifies the root cause of the conflict based on the content of the comments and the results of the sentiment engine. For example, in a technical conflict, it will equally explain the advantages and disadvantages of each solution.

[1138] Step 9:

[1139] The arbitration GPT generates comments to foster empathy. Based on the emotional information detected by the emotion engine, it generates more appropriate empathetic comments. For example, it might say, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?"

[1140] Step 10:

[1141] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[1142] Step 11:

[1143] The arbitration GPT generates neutral and constructive advice, taking into account the information detected by the emotion engine and proposing a solution that both parties can agree on. For example, it may provide advice such as, "How about setting up an opportunity to discuss your respective opinions in more detail?"

[1144] Step 12:

[1145] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[1146] Example 2

[1147] 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."

[1148] In online communities, aggressive comments and arguments are common and often impede smooth communication. This increases the psychological burden on users and worsens the overall atmosphere of the community. Furthermore, emotional conflicts often deepen, preventing constructive discussions. Conventional automated moderation systems are limited to simply filtering or deleting words, and do not fundamentally address the problem. Therefore, this invention aims to resolve conflicts quickly and effectively by recognizing emotions and promoting appropriate communication.

[1149] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for monitoring an online community to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them in a polite style; means for selecting a generative AI model with specialized knowledge; means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model; means for generating neutral and constructive advice; means for recognizing emotions using a user's text message and physiological indicators; and means for rewriting offensive comments based on the emotion recognition results. This makes it possible to rewrite offensive comments and generate appropriate mediating comments taking emotions into consideration, thereby enabling disputes in online communities to be resolved quickly and effectively.

[1150] An "online community" is a virtual space where people can interact using the Internet.

[1151] "Offensive comments" are comments that contain words or phrases intended to insult or harm others.

[1152] A "fight" is a dispute that arises from a clash of opinions or feelings within an online community.

[1153] "Emotion recognition means" refers to technology that analyzes a user's text messages and physiological indicators to detect their emotional state.

[1154] "Polite rewriting" refers to an algorithm that automatically converts offensive comments into more readable and polite language.

[1155] A "generative AI model" is a model that uses artificial intelligence to generate text, particularly using deep learning techniques.

[1156] "Means for sorting out conflicting opinions and generating empathetic comments" refers to the process of analyzing conflicting opinions and generating comments that show empathy for those opinions.

[1157] "Means for generating neutral and constructive advice" refers to technology that automatically generates neutral and helpful suggestions to defuse conflicts.

[1158] "Physiological indicators" refer to data that indicate the user's physical condition, such as heart rate, sweat rate, and blink rate.

[1159] This invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. Specifically, it is more effective when combined with an emotion engine. This system is implemented using the following process:

[1160] Monitoring and Detection

[1161] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques. A specific implementation example uses Apache Kafka to process large volumes of messages in real time and performs analysis using the SpaCy library. It detects comments containing offensive words and phrases and quantifies them. This process allows problematic comments to be identified immediately.

[1162] Certification and Selection

[1163] The server analyzes the content of detected offensive comments to identify the subject of the conflict (technical discussion, personal attack, political conflict, etc.). This analysis uses a topic modeling algorithm (e.g., LDA) to classify the content of the comment. Based on the identified subject, the server selects the most appropriate generative AI model (e.g., GPT-3).

[1164] emotion recognition

[1165] The emotion engine detects emotions from users' text messages. This process uses the Sentiment Analysis API to classify emotions from text. Additionally, physiological indicators (heart rate, sweat rate, blink rate, etc.) are also collected using sensors such as Arduino and Raspberry Pi.

[1166] Automatic Rewrite

[1167] The arbitration GPT receives offensive comments and rewrites them in a polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[1168] Facilitate

[1169] The arbitration GPT organizes the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This empathetic comment is also sent to the device via the server and displayed on the user's screen.

[1170] Reconciliation advice

[1171] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[1172] Specific examples

[1173] An example of a system in place in the event of a technical conflict

[1174] User A: This code is completely unusable. It's full of bugs.

[1175] User B: You're the one who doesn't understand anything.

[1176] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[1177] 2. The server selects a technically specialized generative AI model.

[1178] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[1179] 4. Arbitration GPT rewrites the above comment as follows:

[1180] User A (after rewrite): This code looks a bit problematic. It might need some fixing.

[1181] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1182] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[1183] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[1184] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1185] 7. The server sends the empathy comment to the terminal and displays it on the user's screen.

[1186] 8. Arbitration GPT generates constructive advice towards a resolution.

[1187] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1188] 9. The server sends the advice comment to the terminal and displays it on the user's screen.

[1189] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[1190] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1191] Step 1:

[1192] The server collects all conversation threads of an online community.

[1193] Input: New message in the community

[1194] Processing: Stream messages in real time using Apache Kafka and store them in a database.

[1195] Output: All saved message logs

[1196] Step 2:

[1197] The server analyzes the collected messages using natural language processing (NLP) technology to detect offensive words and phrases.

[1198] Input: Saved message log

[1199] Processing: Parse the message text using the SpaCy library and match it against a list of negative words

[1200] How it works: NLP analysis detects comments like "Are you stupid?" and records them as a numerical value

[1201] Output: List of offensive comments

[1202] Step 3:

[1203] The server analyzes the content of the detected offensive comments to identify the subject of the conflict.

[1204] Input: offensive comment list

[1205] Processing: Classify comments using a topic modeling algorithm (e.g. LDA)

[1206] Behavior: Categorized into categories such as "technical debate," "personal attacks," and "political conflict"

[1207] Output: Categorized offensive comments

[1208] Step 4:

[1209] Based on the identified subject, the server selects a generative AI model in that field.

[1210] Input: Categorized offensive comment

[1211] Processing: Select a generative AI model (e.g., GPT-3) appropriate for each subject

[1212] How it works: For technical discussions, we select a technically specialized GPT-3 model.

[1213] Output: The selected generative AI model

[1214] Step 5:

[1215] The emotion engine uses the user's text messages and physiological indicators to recognize emotions.

[1216] Input: User text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.)

[1217] Processing: Detect emotional states using Sentiment Analysis API and physiological data sensors (Arduino, Raspberry Pi, etc.)

[1218] Behavior: Classifying emotions such as "anger" and "sadness" from text messages and integrating them with physiological indicators

[1219] Output: Emotion recognition result

[1220] Step 6:

[1221] The server uses a generative AI model to rewrite offensive comments based on the emotional information detected by the emotion engine.

[1222] Input: offensive comments and emotion recognition results

[1223] Processing: Generate prompts for rewriting and send them to the generative AI model

[1224] Behavior: Transform "Are you stupid?" into a more polite statement such as "Can you explain your point of view in more detail?"

[1225] Output: Rewritten comments

[1226] Step 7:

[1227] The server sends the rewritten comment to the terminal and displays it on the user's screen.

[1228] Input: Rewritten comment

[1229] Processing: Send the rewritten comment to the device using the REST API

[1230] How it works: Rewritten comments are displayed in real time on the chat screen.

[1231] Output: Rewrite comment displayed to the user

[1232] Step 8:

[1233] The server uses a generative AI model to generate empathetic comments based on the information provided by the emotion engine.

[1234] Input: Emotion recognition results

[1235] Processing: Generate prompts for generating empathetic comments and send them to the generative AI model

[1236] What it does: Generates empathetic comments like, "It's really important to calmly discuss bugs in your code. Can you share your perspectives on the issues?"

[1237] Output: Empathetic comments

[1238] Step 9:

[1239] The server sends the sympathetic comments to the terminal and displays them on the user's screen.

[1240] Input: Sympathetic comment

[1241] Processing: Send sympathetic comments to the device using the REST API

[1242] How it works: Show empathetic comments in real time on the chat screen

[1243] Output: Empathy comments displayed on the user's screen

[1244] Step 10:

[1245] The server uses arbitration GPT to generate neutral and constructive advice.

[1246] Input: Emotion recognition results and conflict content

[1247] Processing: Generate prompts for advice generation and send them to the generative AI model

[1248] Behavior: Generates constructive advice such as, "How about we set up an opportunity to discuss each other's opinions in more detail?"

[1249] Output: Advisory comments

[1250] Step 11:

[1251] The server sends the advice comment to the terminal and displays it on the user's screen.

[1252] Input: Advisory Comment

[1253] Processing: Sends advice to the device using the REST API

[1254] How it works: Real-time advice is displayed on the chat screen.

[1255] Output: Advisory comments displayed on the user's screen

[1256] This processing flow enables aggressive comments and disputes in online communities to be resolved quickly and effectively, providing a healthy communication environment that also takes into consideration the feelings of users.

[1257] (Application example 2)

[1258] 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."

[1259] In traditional online communities, aggressive comments and arguments are common, often damaging the user experience. Even in brick-and-mortar stores, communication between customers and staff, or between staff members, can be disrupted, leading to lower customer satisfaction and reduced staff efficiency. A system that can quickly and effectively resolve these issues and promote healthy communication is needed.

[1260] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for applying the monitored communication content to customer service and staff dialogue in physical stores, means for providing real-time filtering and rewriting of offensive comments via a smartphone, means for visualizing the emotions of customers and staff using an emotion engine, and means for displaying empathetic comments and constructive advice in the physical store environment in real time. This makes it possible to provide a healthy communication environment in online communities and physical stores.

[1261] An "online community" is a place where users can interact with each other and share information via the Internet.

[1262] "Monitoring" is the act of observing and monitoring specific situations or data in real time.

[1263] "Offensive comments" are statements that contain the intent to insult or attack others.

[1264] A "quarrel" refers to a clash caused by an argument or a difference of opinion.

[1265] "Rewriting" is the act of reconstructing a sentence and rewriting it in a different way.

[1266] A "generative AI model" is an artificial intelligence model that is trained to perform generative tasks.

[1267] "Selection" is the act of choosing an appropriate subject.

[1268] An "empathetic comment" is a statement made to show empathy for the feelings or opinions of others.

[1269] "Neutral and constructive advice" is advice that is unbiased and aims to lead in a positive direction.

[1270] A "system" is a collection of elements organized to achieve a particular purpose.

[1271] A "physical store" is a store or service facility that exists in a physical location.

[1272] "Customer service" refers to the act of responding to customer requests and questions.

[1273] "Inter-staff dialogue" refers to communication between employees who belong to the same workplace.

[1274] A "smartphone" is a mobile device that has similar communication capabilities to a mobile phone, but with more advanced computing capabilities.

[1275] An "emotion engine" is a technology that analyzes and recognizes a user's emotions from text and physiological data.

[1276] "Real-time filtering" is the process of instantly analyzing data or information and sorting it based on specific criteria.

[1277] "Visualization" is the process of expressing abstract data or information in a visually easy-to-understand form.

[1278] The present invention provides a system for monitoring communications in online communities and brick-and-mortar stores and effectively resolving offensive comments and conflicts. Hereinafter, embodiments of the present invention will be described in detail.

[1279] System Program

[1280] The system of the present invention is implemented using the following hardware and software:

[1281] Hardware: Smartphones, servers

[1282] Software: Natural language processing (NLP) libraries (e.g., spaCy), sentiment analysis tools (e.g., AWS Comprehend, Google Cloud Natural Language API), generative AI models (e.g., OpenAI GPT)

[1283] Process Overview

[1284] Monitoring and Detection

[1285] The server monitors communication content in physical stores and online communities in real time. This monitoring includes using NLP technology to detect offensive words and phrases. For example, if a user types an offensive comment such as "it's unusable" into their smartphone, the content will be detected immediately.

[1286] Comment Rewrite

[1287] Detected offensive comments are automatically rewritten to a more polite style using a generative AI model. The server inputs appropriate prompts to the generative AI model and receives the rewritten comments. For example, a comment such as "The service on this product was poor!" is rewritten to "There seems to be some room for improvement in the service on this product."

[1288] Visualizing emotions

[1289] The server uses an emotion engine to analyze the user's emotions and notify customers and staff of negative emotions. The results of the emotion analysis are visualized and displayed in real time on smartphones and other devices. For example, if "anger" is detected based on a customer's text and physiological data, that information is displayed on the staff's device.

[1290] Empathy comment generation

[1291] The selected generative AI model generates empathetic comments based on the detected emotional information, which promotes conflict resolution. The server reads the output of the emotion engine and inputs a prompt sentence to generate an empathetic comment. For example, the prompt sentence used is "Generate an empathetic comment for the following message: The service on this product was poor!"

[1292] Providing constructive advice

[1293] Neutral and constructive advice is generated by the generative AI model and provided to store staff and customers in real time, enabling all parties involved to find a better solution. The server inputs prompts offering constructive advice based on the content of the conflict, and the generated advice is displayed on the device.

[1294] Specific examples

[1295] Handling customer complaints

[1296] If a customer says, "The service on this product was poor!"

[1297] 1. The server detects this offensive comment.

[1298] 2. Using the generative AI model, rewrite the sentence to "There seems to be some room for improvement in how this product is handled."

[1299] 3. The emotion engine detects anger and notifies staff.

[1300] 4. The server inputs the prompt text "Generate an empathetic comment for the following message: The service on this product was poor!" and generates an empathetic comment.

[1301] 5. Provide the staff with a sympathetic comment saying, "We take your feedback seriously and will strive to improve," along with constructive advice saying, "We would appreciate it if you could tell us the specific issues you had."

[1302] This will enable us to provide a healthy communication environment in our physical stores and online communities.

[1303] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1304] Step 1:

[1305] The server monitors communication content in physical stores and online communities. The input includes user text messages, and the data is analyzed using NLP techniques. The output detects messages containing offensive comments or negative language.

[1306] Step 2:

[1307] The server inputs the detected offensive comments into a generative AI model, which rewrites them in a polite style. The input includes an offensive text message, which is passed to the generative AI model along with a prompt. The output is a rewritten, polite comment. For example, a comment like "The service on this product was poor!" is changed to "There seems to be some room for improvement in the service on this product."

[1308] Step 3:

[1309] The server uses an emotion engine to analyze emotions from the user's text message and physiological indicators (e.g., heart rate, sweat rate, blink rate). The input includes the rewritten text message and physiological data. The output is the emotion analysis result, which detects emotions such as "anger" or "sadness."

[1310] Step 4:

[1311] The server generates an empathetic comment based on the analysis results of the emotion engine. The input includes the emotion analysis results and the rewritten text message, which are passed to the generative AI model along with a prompt. The output is a comment expressing empathy. For example, a prompt such as "Generate an empathetic comment for the following message: The service on this product was poor!" is used.

[1312] Step 5:

[1313] The server generates neutral and constructive advice. The input includes the sentiment analysis results and empathetic comments, which are passed to the generative AI model along with a prompt. The output is constructive advice. For example, the generated advice might be, "We would appreciate it if you could tell us the specific problem you are having."

[1314] Step 6:

[1315] The terminal displays the rewritten comments, empathetic comments, and constructive advice sent from the server to the user. The input includes the rewritten comments, empathetic comments, and advice from the server. The output is displayed on the user's screen, promoting healthy communication.

[1316] Step 7:

[1317] The user can smoothly communicate based on the rewritten comments, empathetic comments, and constructive advice displayed on the terminal. The input includes the content displayed on the terminal, and the user's response and behavior are obtained as the output.

[1318] 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.

[1319] 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.

[1320] 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.

[1321] [Fourth embodiment]

[1322] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1323] 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.

[1324] 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).

[1325] 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.

[1326] 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.

[1327] 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).

[1328] 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.

[1329] 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.

[1330] 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.

[1331] 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.

[1332] 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.

[1333] 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.

[1334] 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."

[1335] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, which is implemented through the following process.

[1336] System configuration

[1337] 1. Monitoring and Detection

[1338] The server monitors all conversation threads in online communities in real time and automatically detects offensive comments and signs of conflict using natural language processing (NLP) technology, specifically by detecting negative language patterns (e.g., "stupid" or "useless").

[1339] 2. Certification and Selection

[1340] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict. Based on this, it selects a generative AI model with specific expertise and selects it as the appropriate arbitrator. For example, in the case of a technical discussion, a technically specialized GPT (generative AI model) is selected.

[1341] 3. Automatic Rewrite

[1342] The arbitration GPT rewrites offensive comments into polite language. For example, it converts a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?" The rewritten comment is sent to the terminal via the server and displayed on the user's screen.

[1343] 4. Facilitate

[1344] The arbitration GPT organizes the content of comments and identifies the root cause of the conflict. For example, in the case of a technical conflict, it will evenly summarize the advantages and disadvantages of the solution. It also generates empathetic comments to satisfy the user's desire for approval. For example, it generates a comment such as, "Your opinion is very interesting. I was particularly impressed with the part about XX." This is also sent to the device via the server and displayed on the user's screen.

[1345] 5. Reconciliation Advice

[1346] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[1347] Specific examples

[1348] Case

[1349] If you encounter technical conflicts such as:

[1350] User A: This code is completely unusable. It's full of bugs.

[1351] User B: You're the one who doesn't understand anything.

[1352] Example

[1353] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[1354] 2. The server selects a technically-specialized arbitrator, the GPT.

[1355] 3. Arbitration GPT automatically rewrites the above comment as follows:

[1356] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[1357] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1358] 4. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[1359] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments.

[1360] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1361] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[1362] 7. Arbitration GPT generates constructive advice towards a resolution.

[1363] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1364] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[1365] In this way, the system of the present invention can quickly and effectively resolve disputes and conflicts in online communities, providing a healthy communication environment.

[1366] The processing flow will be explained below.

[1367] Step 1:

[1368] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[1369] Step 2:

[1370] The server analyzes the detected offensive comments and uses a comment scoring algorithm to quantify their offensiveness. For example, if a score above a certain threshold is assigned, the comment is deemed definitively offensive.

[1371] Step 3:

[1372] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[1373] Step 4:

[1374] The arbitration GPT takes offensive comments and applies an algorithm to rewrite them in a more polite style, for example rewriting a comment like "Are you stupid?" to "Could you please elaborate a bit more on your point of view?"

[1375] Step 5:

[1376] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[1377] Step 6:

[1378] Arbitration GPT analyzes the background and arguments of the conflicting opinions and identifies the fundamental conflicts. For example, in the case of technical conflicts, it evenly summarizes the advantages and disadvantages of the conflicting solutions.

[1379] Step 7:

[1380] Arbitration GPT generates comments to foster empathy, such as "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?"

[1381] Step 8:

[1382] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[1383] Step 9:

[1384] Arbitration GPT generates neutral and constructive advice, such as "How about creating an opportunity to discuss each other's opinions in more detail?"

[1385] Step 10:

[1386] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[1387] Example 1

[1388] 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."

[1389] As online communities grow, the number of offensive comments and conflicts between users increases, leading to a deterioration in the communication environment. Conventional methods rely on manual moderation, making it difficult to respond quickly and effectively, so an automated system is needed to maintain the health of online communities.

[1390] 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.

[1391] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and conflicts, means for analyzing the detected comments to identify the subject of the conflict and select an appropriate generative AI model, means for automatically rewriting the offensive comments to a polite style using the selected generative AI model, means for the generative AI model to organize the conflicting opinions and generate empathetic comments, and means for generating neutral and constructive advice. This makes it possible to quickly and effectively resolve offensive comments and conflicts in online communities and provide a healthy communication environment.

[1392] An "online community" is a virtual space where people can exchange information and interact via the Internet.

[1393] "Monitoring" means to observe a specific object periodically or continuously to check its trends and status.

[1394] "Offensive comments" are statements that contain negative, insulting, or provocative content toward others.

[1395] "Conflict" refers to a clash or dispute that arises due to different opinions or values.

[1396] "Automatic detection" refers to the ability of a system to detect specific events based on certain conditions or patterns without human intervention.

[1397] A "generative AI model" is an artificial intelligence system that can perform generative tasks (such as text generation or translation).

[1398] "Polite writing style" is a characteristic of writing that uses gentle expressions that convey respect and consideration.

[1399] "Rewriting" means changing the form or wording of an original text while preserving its content.

[1400] An "empathetic comment" is a statement that shows understanding or agreement with the feelings or opinions of others.

[1401] "Constructive advice" refers to making specific and positive suggestions for solving problems or making improvements.

[1402] "Natural language processing technology" refers to technology that allows computers to understand, analyze, and generate human language.

[1403] This invention is a system for quickly and effectively resolving offensive comments and conflicts in online communities, which mainly involves the cooperation of three entities: a server, a terminal, and a user.

[1404] System configuration

[1405] Hardware and software used

[1406] Server: A high-performance server system for monitoring online communities in real time. Data is collected using a communication API.

[1407] Natural Language Processing (NLP): Analyze comment data using Google Cloud Natural Language and SpaCy.

[1408] Generative AI model: An AI system used for generative tasks. The model uses the technically specialized GPT (Generative Pre-trained Transformer).

[1409] Implementation method

[1410] 1. Monitoring

[1411] The server collects comment data from online communities via specific APIs (e.g., Twitter API or Reddit API). The collected data is periodically sent to the server and analyzed in real time using NLP technology.

[1412] 2. Detecting offensive comments

[1413] The server analyzes the collected comment data, searching for negative language patterns and specific keywords (e.g., "stupid," "useless," etc.), using a pre-trained machine learning model (e.g., the BERT model).

[1414] 3. Selecting a Generative AI Model

[1415] The server further analyzes the content of the detected comments to identify the topic of conflict, and then selects the most appropriate generative AI model (e.g., technically specialized GPT) based on the identified topic.

[1416] 4. Rewrite offensive comments

[1417] The server generates a prompt sentence that inputs the detected offensive comment into the arbitration GPT model. For example, a prompt sentence of the form "Please rewrite the following offensive comment in a polite style: 'Are you stupid?'" is created. The arbitration GPT generates the rewritten comment and sends it to the terminal (the user's display device) via the server.

[1418] 5. Generating empathetic comments and advice

[1419] The arbitration GPT analyzes the cause of the conflict and generates empathetic comments and neutral advice. For example, it might create a comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the issues?" These comments are also sent to the terminal via the server and displayed to the user.

[1420] Specific examples

[1421] Case

[1422] If you encounter technical conflicts such as:

[1423] User A: This code is completely unusable. It's full of bugs.

[1424] User B: You're the one who doesn't understand anything.

[1425] Example

[1426] 1. The server monitors comments and detects offensive comments ("useless" or "you don't understand").

[1427] 2. The server selects a technically-specialized arbitrator GPT for the detected comment.

[1428] 3. Arbitration GPT automatically rewrites the above comment as follows:

[1429] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[1430] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1431] 4. The server sends the rewritten comments to the terminals and displays them on the screens of User A and User B.

[1432] 5. Arbitration GPT analyzes the cause of the conflict and generates empathetic comments:

[1433] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1434] 6. The server sends the empathy comment to the terminal and displays it on the user's screen.

[1435] 7. Arbitration GPT generates constructive advice towards resolution:

[1436] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1437] 8. The server sends the advice comment to the terminal and displays it on the user's screen.

[1438] As a result, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[1439] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1440] Step 1:

[1441] The server collects comment data from online communities. Specifically, it retrieves comment data in real time using a specific API (e.g., Twitter API or Reddit API). The input is the comment data retrieved from the API, and the output is the raw comment data to be analyzed.

[1442] Step 2:

[1443] The server applies natural language processing (NLP) techniques to the collected comment data to detect offensive or conflicting comments. Specifically, it performs text analysis using NLP tools (such as Google Cloud Natural Language or SpaCy). The input is raw comment data, and the output is comment data flagged as offensive or conflicting.

[1444] Step 3:

[1445] The server performs a detailed analysis of the content of the detected offensive comments and identifies the subject of the conflict. Specifically, it uses topic modeling techniques (such as LDA and CLIP) to identify what the conflict is about. The input is flagged comment data, and the output is data identifying the subject of the conflict.

[1446] Step 4:

[1447] The server selects an appropriate generative AI model (arbitration GPT) based on the identified subject of the conflict. The input is data identifying the subject of the conflict, and the output is information about the selected generative AI model.

[1448] Step 5:

[1449] The server generates a prompt for the selected generative AI model to rewrite the offensive comment. An example of a specific prompt is "Please rewrite the following offensive comment in a polite manner: 'Are you stupid?'" The input is the offensive comment and information about the selected generative AI model, and the output is the prompt.

[1450] Step 6:

[1451] The arbitration GPT receives the generated prompt and rewrites the offensive comments to a more polite style. The input is the prompt, and the output is the rewritten comment.

[1452] Step 7:

[1453] The server sends the rewritten comment to the terminal and displays it on the user's screen. The input is the rewritten comment, and the output is the rewritten comment displayed on the user's screen.

[1454] Step 8:

[1455] The arbitration GPT generates empathetic comments based on the rewritten comments and conflict analysis data. Specifically, it generates phrases that ease the conflict and deepen understanding between users. The inputs are the rewritten comments and conflict analysis data, and the output is empathetic comments.

[1456] Step 9:

[1457] The server sends the generated empathy comments to the terminal and displays them to the user. The input is the empathy comments, and the output is the empathy comments displayed on the user's screen.

[1458] Step 10:

[1459] The arbitration GPT ultimately generates constructive advice for resolving the conflict. Specifically, it suggests something like, "How about setting up an opportunity to discuss each other's opinions in more detail?" The inputs are conflict analysis data and the context of the dialogue, and the output is constructive advice.

[1460] Step 11:

[1461] The server sends the generated advice comment to the terminal and displays it on the user's screen. The advice comment is the input, and the advice comment displayed on the user's screen is obtained as the output.

[1462] (Application example 1)

[1463] 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."

[1464] Currently, offensive comments and conflicts are common in online community comment sections, hindering healthy communication. Applications used on smart devices, in particular, lack the ability to detect and appropriately address these issues in real time, resulting in a poor user experience. This invention aims to provide a system that quickly and effectively detects offensive comments and promotes constructive communication.

[1465] 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.

[1466] In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for monitoring the comment section of an application used on a smart device to detect offensive comments in real time, and means for rewriting offensive comments to polite expressions and generating empathetic comments and constructive advice. This makes it possible to effectively suppress offensive behavior in the comment section and promote healthy and constructive communication.

[1467] An "online community" is a virtual space on the Internet where users can exchange opinions and share information.

[1468] "Monitoring" is the process by which a system continuously monitors all conversation threads in an online community to detect problematic comments in real time.

[1469] "Offensive comments" are comments that contain expressions or words intended to harm other users or that may cause conflict.

[1470] A "fight" refers to an emotional or aggressive conflict between users within an online community.

[1471] "Polite writing style" refers to the transformation of aggressive comments into expressions that respect others and allow for polite communication.

[1472] A "generative AI model" is an algorithm or system that has the ability to generate natural language using artificial intelligence technology.

[1473] An "empathetic comment" is a comment that includes expressions that empathize with the opinions and feelings of others and show positive and understanding.

[1474] "Neutral and constructive advice" is practical advice given from an impartial standpoint to resolve conflicts.

[1475] A "smart device" is a computerized portable device that can connect to the Internet and perform multiple functions. Examples include smartphones, smart glasses, and head-mounted displays.

[1476] "Real-time" means that data and information are processed almost immediately and results are obtained without any time lag.

[1477] "Polite expression" is language that is polite and carefully chosen so as not to offend others.

[1478] MODE FOR CARRYING OUT THE INVENTION

[1479] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. The system includes an application and a server that are run on smart devices.

[1480] Overall system configuration

[1481] The server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them to a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, and means for generating neutral and constructive advice. The server also includes means for monitoring comment sections in an application used on a smart device and detecting offensive comments in real time.

[1482] Hardware and Software Used

[1483] Hardware:

[1484] Smartphone

[1485] Smart Glasses

[1486] head-mounted display

[1487] software:

[1488] Natural language processing (NLP) models for monitoring (e.g., parsers such as Hugging Face's BERT)

[1489] Generative AI models (e.g., GPT models) for detecting and rewriting offensive comments

[1490] Data processing and calculation

[1491] The server monitors online community comments in real time and uses NLP technology, such as Hugging Face's "sentiment-analysis" pipeline, to detect offensive comments. Detected offensive comments generate appropriate prompts, which are then fed into a generative AI model (e.g., a GPT model). This generative AI model rewrites offensive comments into politer expressions and also generates empathetic comments and constructive advice.

[1492] Specific examples

[1493] Below is a concrete example of the process for detecting offensive comments, rewriting them, and generating empathetic comments and constructive advice.

[1494] Comment example

[1495] User A: "This video is completely pointless. It's a waste of time."

[1496] User B: "No one is listening to your opinion. Shut up."

[1497] Prompt Sentence Examples

[1498] "Rewrite the following offensive comment politely: This video is completely pointless. A waste of time."

[1499] "Create a neutral and empathetic response to the following opinion: No one is listening to your opinion. Shut up."

[1500] "Provide a constructive advice to resolve an online argument."

[1501] The server generates these prompts and inputs them into a generative AI model, which rewrites offensive comments into polite expressions and generates empathetic comments and constructive advice. The rewritten comments and advice are then displayed on the user's smart device screen, promoting healthy communication.

[1502] In this way, the system of the present invention can quickly and effectively resolve offensive comments and conflicts in online communities, providing a healthy communication environment.

[1503] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1504] Step 1:

[1505] The server monitors the comment section of an online community in real time. As input, comment data is provided from the online community. As output, offensive comments or comments that indicate signs of conflict are extracted. Specifically, the server uses a natural language processing (NLP) model to analyze the sentiment of the comments, for example, using Hugging Face's "sentiment-analysis" pipeline.

[1506] Step 2:

[1507] The server analyzes the extracted offensive comments. As input, it uses the offensive comments extracted in step 1. As output, it identifies the offensive elements and provides information to translate them into polite language. Specifically, the server runs an algorithm to identify negative words and phrases in the comments.

[1508] Step 3:

[1509] The server selects a generative AI model. As input, it uses the comment content and its characteristic information analyzed in step 2. As output, it selects an appropriate generative AI model (e.g., GPT model). Specifically, the server applies rule-based logic to select an appropriate AI model based on the comment content.

[1510] Step 4:

[1511] The server rewrites offensive comments into polite language. The selected generative AI model and the offensive comment are used as input. The rewritten, polite comment is generated as output. Specifically, the server generates a prompt sentence: "Rewrite the following offensive comment politely: [offensive comment]" and inputs it into the generative AI model.

[1512] Step 5:

[1513] The server sorts out the conflicting opinions and generates empathetic comments. The rewritten comments and the content of the conflict are used as input. The output is an empathetic comment. Specifically, the server generates a prompt statement: "Create a neutral and empathetic response to the following opinion: [content of the conflict]" and inputs this into the generative AI model.

[1514] Step 6:

[1515] The server generates neutral and constructive advice. Empathetic comments and conflict resolution information are used as input. Constructive advice is generated as output. Specifically, the server generates a prompt statement, "Provide a constructive advice to resolve an online argument," and inputs this into the generative AI model.

[1516] Step 7:

[1517] The server sends the generated rewrite comments, empathy comments, and constructive advice to the terminal. As input, the comments generated in steps 4, 5, and 6 are used. As output, rewrite comments, empathy comments, and constructive advice are generated and displayed on the terminal. Specifically, the server uses a protocol to combine each comment and send it to the user's screen.

[1518] 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.

[1519] The present invention is a system for quickly and effectively resolving offensive comments and disputes in online communities, and is further enhanced by combining it with an emotion engine. This system is implemented using the following process:

[1520] System configuration

[1521] 1. Monitoring and Detection

[1522] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques to detect and quantify comments containing offensive words or phrases.

[1523] 2. Certification and Selection

[1524] The server analyzes the content of the detected offensive comments and identifies the subject of the conflict (technical discussion, personal attack, political conflict, etc.) Based on this identified subject, it selects a generative AI model (GPT) with expertise in the relevant field.

[1525] 3. Emotion recognition

[1526] The emotion engine detects emotions from the user's text messages. It also uses physiological indicators (e.g., heart rate, sweat rate, blink rate) to recognize the user's emotions. For example, if it detects negative emotions (e.g., anger, sadness, irritation), it uses that information for the next step.

[1527] 4. Automatic Rewrite

[1528] The arbitration GPT receives offensive comments and rewrites them in a more polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[1529] 5. Facilitate

[1530] The arbitration GPT sorts through the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This is also sent to the device via the server and displayed on the user's screen.

[1531] 6. Reconciliation Advice

[1532] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[1533] Specific examples

[1534] Case

[1535] If you encounter technical conflicts such as:

[1536] User A: This code is completely unusable. It's full of bugs.

[1537] User B: You're the one who doesn't understand anything.

[1538] Example

[1539] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[1540] 2. The server selects a technically-specialized arbitrator, the GPT.

[1541] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[1542] 4. Arbitration GPT rewrites the above comment as follows:

[1543] User A (after rewrite): This code looks a bit problematic. It might need some modification.

[1544] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1545] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[1546] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[1547] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1548] 7. The server sends the empathy comment to the terminal and it is displayed on the user's screen.

[1549] 8. Arbitration GPT generates constructive advice towards a resolution.

[1550] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1551] 9. The server sends the advice comment to the terminal and it is displayed on the user's screen.

[1552] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[1553] The processing flow will be explained below.

[1554] Step 1:

[1555] The server monitors all conversation threads in online communities in real time, using natural language processing (NLP) techniques to detect comments containing offensive words or phrases.

[1556] Step 2:

[1557] The server analyzes the detected offensive comments and quantifies their offensiveness. Specifically, if a score exceeding a certain threshold is assigned, the comment is deemed to be definitively offensive.

[1558] Step 3:

[1559] The server analyzes the content of detected offensive comments and identifies the subject of the conflict, categorizing them into categories such as technical arguments, personal attacks, and political conflicts.

[1560] Step 4:

[1561] Based on the identified subject, the server selects a generative AI model (GPT) with domain expertise. For example, in the case of a technical discussion, a technical GPT is selected.

[1562] Step 5:

[1563] The emotion engine analyzes the user's text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.) to recognize the user's emotions. For example, it can determine that User A is feeling "anger" based on text analysis.

[1564] Step 6:

[1565] The arbitration GPT applies an algorithm to rewrite offensive comments to a more polite style, taking into account the results of the sentiment engine and rewriting appropriately. For example, it transforms a comment like "Are you stupid?" into "Could you please explain your point of view in more detail?"

[1566] Step 7:

[1567] The server receives the rewritten comment and sends it to the terminal, which displays the rewritten comment on the user's screen.

[1568] Step 8:

[1569] The arbitration GPT identifies the root cause of the conflict based on the content of the comments and the results of the sentiment engine. For example, in a technical conflict, it will equally explain the advantages and disadvantages of each solution.

[1570] Step 9:

[1571] The arbitration GPT generates comments to foster empathy. Based on the emotional information detected by the emotion engine, it generates more appropriate empathetic comments. For example, it might say, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?"

[1572] Step 10:

[1573] The server transmits the generated sympathy comment to the terminal, which displays it on the user's screen.

[1574] Step 11:

[1575] The arbitration GPT generates neutral and constructive advice, taking into account the information detected by the emotion engine and proposing a solution that both parties can agree on. For example, it may provide advice such as, "How about setting up an opportunity to discuss your respective opinions in more detail?"

[1576] Step 12:

[1577] The server sends the generated constructive advice to the terminal, which displays it on the user's screen.

[1578] Example 2

[1579] 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."

[1580] In online communities, aggressive comments and arguments are common and often impede smooth communication. This increases the psychological burden on users and worsens the overall atmosphere of the community. Furthermore, emotional conflicts often deepen, preventing constructive discussions. Conventional automated moderation systems are limited to simply filtering or deleting words, and do not fundamentally address the problem. Therefore, this invention aims to resolve conflicts quickly and effectively by recognizing emotions and promoting appropriate communication.

[1581] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes: means for monitoring an online community to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them in a polite style; means for selecting a generative AI model with specialized knowledge; means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model; means for generating neutral and constructive advice; means for recognizing emotions using a user's text message and physiological indicators; and means for rewriting offensive comments based on the emotion recognition results. This makes it possible to rewrite offensive comments and generate appropriate mediating comments taking emotions into consideration, thereby enabling disputes in online communities to be resolved quickly and effectively.

[1582] An "online community" is a virtual space where people can interact using the Internet.

[1583] "Offensive comments" are comments that contain words or phrases intended to insult or harm others.

[1584] A "fight" is a dispute that arises from a clash of opinions or feelings within an online community.

[1585] "Emotion recognition means" refers to technology that analyzes a user's text messages and physiological indicators to detect their emotional state.

[1586] "Polite rewriting" refers to an algorithm that automatically converts offensive comments into more readable and polite language.

[1587] A "generative AI model" is a model that uses artificial intelligence to generate text, particularly using deep learning techniques.

[1588] "Means for sorting out conflicting opinions and generating empathetic comments" refers to the process of analyzing conflicting opinions and generating comments that show empathy for those opinions.

[1589] "Means for generating neutral and constructive advice" refers to technology that automatically generates neutral and helpful suggestions to defuse conflicts.

[1590] "Physiological indicators" refer to data that indicate the user's physical condition, such as heart rate, sweat rate, and blink rate.

[1591] This invention is a system for quickly and effectively resolving offensive comments and disputes in online communities. Specifically, it is more effective when combined with an emotion engine. This system is implemented using the following process:

[1592] Monitoring and Detection

[1593] The server monitors all conversation threads in online communities in real time and analyzes messages using natural language processing (NLP) techniques. A specific implementation example uses Apache Kafka to process large volumes of messages in real time and performs analysis using the SpaCy library. It detects comments containing offensive words and phrases and quantifies them. This process allows problematic comments to be identified immediately.

[1594] Certification and Selection

[1595] The server analyzes the content of detected offensive comments to identify the subject of the conflict (technical discussion, personal attack, political conflict, etc.). This analysis uses a topic modeling algorithm (e.g., LDA) to classify the content of the comment. Based on the identified subject, the server selects the most appropriate generative AI model (e.g., GPT-3).

[1596] emotion recognition

[1597] The emotion engine detects emotions from users' text messages. This process uses the Sentiment Analysis API to classify emotions from text. Additionally, physiological indicators (heart rate, sweat rate, blink rate, etc.) are also collected using sensors such as Arduino and Raspberry Pi.

[1598] Automatic Rewrite

[1599] The arbitration GPT receives offensive comments and rewrites them in a polite style, taking into account the emotional information detected by the emotion engine. For example, it rewrites a comment like "Are you stupid?" to "Could you please explain your point of view in more detail?" The rewritten comment is sent to the device via the server and displayed on the user's screen.

[1600] Facilitate

[1601] The arbitration GPT organizes the content of the comments and identifies the root cause of the conflict. It generates empathetic comments based on the emotional information detected by the emotion engine. For example, it generates an empathetic comment such as, "It's very important to calmly discuss code bugs. Could you please share your perspectives on the problem?" This empathetic comment is also sent to the device via the server and displayed on the user's screen.

[1602] Reconciliation advice

[1603] The arbitration GPT provides neutral and constructive advice, such as suggesting, "How about creating an opportunity to discuss each other's opinions in more detail?" This advice is also sent to the terminal via the server and displayed on the user's screen.

[1604] Specific examples

[1605] An example of a system in place in the event of a technical conflict

[1606] User A: This code is completely unusable. It's full of bugs.

[1607] User B: You're the one who doesn't understand anything.

[1608] 1. The server monitors comments and detects offensive remarks ("useless" or "you don't understand").

[1609] 2. The server selects a technically specialized generative AI model.

[1610] 3. The emotion engine analyzes the text messages and physiological indicators of User A and User B to detect negative emotions, for example, if User A is feeling angry.

[1611] 4. Arbitration GPT rewrites the above comment as follows:

[1612] User A (after rewrite): This code looks a bit problematic. It might need some fixing.

[1613] User B (after rewrite): Why don't we both take a look at the code a bit more?

[1614] 5. The server sends the rewritten comments to the terminal and displays them on the screens of User A and User B.

[1615] 6. Based on the information from the emotion engine, the arbitration GPT generates empathetic comments.

[1616] "It's really important to calmly discuss bugs in the code. Can you share your perspectives on the issues?"

[1617] 7. The server sends the empathy comment to the terminal and displays it on the user's screen.

[1618] 8. Arbitration GPT generates constructive advice towards a resolution.

[1619] "How about we create an opportunity to discuss each other's opinions in more detail?"

[1620] 9. The server sends the advice comment to the terminal and displays it on the user's screen.

[1621] In this way, the system of the present invention can quickly resolve disputes and conflicts in online communities and provide a healthy communication environment that also takes into consideration the feelings of users.

[1622] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1623] Step 1:

[1624] The server collects all conversation threads of an online community.

[1625] Input: New message in the community

[1626] Processing: Stream messages in real time using Apache Kafka and store them in a database.

[1627] Output: All saved message logs

[1628] Step 2:

[1629] The server analyzes the collected messages using natural language processing (NLP) technology to detect offensive words and phrases.

[1630] Input: Saved message log

[1631] Processing: Parse the message text using the SpaCy library and match it against a list of negative words

[1632] How it works: NLP analysis detects comments like "Are you stupid?" and records them as a numerical value

[1633] Output: List of offensive comments

[1634] Step 3:

[1635] The server analyzes the content of the detected offensive comments to identify the subject of the conflict.

[1636] Input: offensive comment list

[1637] Processing: Classify comments using a topic modeling algorithm (e.g. LDA)

[1638] Behavior: Categorized into categories such as "technical debate," "personal attacks," and "political conflict"

[1639] Output: Categorized offensive comments

[1640] Step 4:

[1641] Based on the identified subject, the server selects a generative AI model in that field.

[1642] Input: Categorized offensive comment

[1643] Processing: Select a generative AI model (e.g., GPT-3) appropriate for each subject

[1644] How it works: For technical discussions, we select a technically specialized GPT-3 model.

[1645] Output: The selected generative AI model

[1646] Step 5:

[1647] The emotion engine uses the user's text messages and physiological indicators to recognize emotions.

[1648] Input: User text messages and physiological indicators (heart rate, sweat rate, blink rate, etc.)

[1649] Processing: Detect emotional states using Sentiment Analysis API and physiological data sensors (Arduino, Raspberry Pi, etc.)

[1650] Behavior: Classifying emotions such as "anger" and "sadness" from text messages and integrating them with physiological indicators

[1651] Output: Emotion recognition result

[1652] Step 6:

[1653] The server uses a generative AI model to rewrite offensive comments based on the emotional information detected by the emotion engine.

[1654] Input: offensive comments and emotion recognition results

[1655] Processing: Generate prompts for rewriting and send them to the generative AI model

[1656] Behavior: Transform "Are you stupid?" into a more polite statement such as "Can you explain your point of view in more detail?"

[1657] Output: Rewritten comments

[1658] Step 7:

[1659] The server sends the rewritten comment to the terminal and displays it on the user's screen.

[1660] Input: Rewritten comment

[1661] Processing: Send the rewritten comment to the device using the REST API

[1662] How it works: Rewritten comments are displayed in real time on the chat screen.

[1663] Output: Rewrite comment displayed to the user

[1664] Step 8:

[1665] The server uses a generative AI model to generate empathetic comments based on the information provided by the emotion engine.

[1666] Input: Emotion recognition results

[1667] Processing: Generate prompts for generating empathetic comments and send them to the generative AI model

[1668] What it does: Generates empathetic comments like, "It's really important to calmly discuss bugs in your code. Can you share your perspectives on the issues?"

[1669] Output: Empathetic comments

[1670] Step 9:

[1671] The server sends the sympathetic comments to the terminal and displays them on the user's screen.

[1672] Input: Sympathetic comment

[1673] Processing: Send sympathetic comments to the device using the REST API

[1674] How it works: Show empathetic comments in real time on the chat screen

[1675] Output: Empathy comments displayed on the user's screen

[1676] Step 10:

[1677] The server uses arbitration GPT to generate neutral and constructive advice.

[1678] Input: Emotion recognition results and conflict content

[1679] Processing: Generate prompts for advice generation and send them to the generative AI model

[1680] Behavior: Generates constructive advice such as, "How about we set up an opportunity to discuss each other's opinions in more detail?"

[1681] Output: Advisory comments

[1682] Step 11:

[1683] The server sends the advice comment to the terminal and displays it on the user's screen.

[1684] Input: Advisory Comment

[1685] Processing: Sends advice to the device using the REST API

[1686] How it works: Real-time advice is displayed on the chat screen.

[1687] Output: Advisory comments displayed on the user's screen

[1688] This processing flow enables aggressive comments and disputes in online communities to be resolved quickly and effectively, providing a healthy communication environment that also takes into consideration the feelings of users.

[1689] (Application example 2)

[1690] 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."

[1691] In traditional online communities, aggressive comments and arguments are common, often damaging the user experience. Even in brick-and-mortar stores, communication between customers and staff, or between staff members, can be disrupted, leading to lower customer satisfaction and reduced staff efficiency. A system that can quickly and effectively resolve these issues and promote healthy communication is needed.

[1692] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for monitoring online communities to automatically detect offensive comments and arguments, means for analyzing the detected comments and rewriting them in a polite style, means for selecting a generative AI model with specialized knowledge, means for sorting out conflicts of opinion and generating empathetic comments using the selected generative AI model, means for generating neutral and constructive advice, means for applying the monitored communication content to customer service and staff dialogue in physical stores, means for providing real-time filtering and rewriting of offensive comments via a smartphone, means for visualizing the emotions of customers and staff using an emotion engine, and means for displaying empathetic comments and constructive advice in the physical store environment in real time. This makes it possible to provide a healthy communication environment in online communities and physical stores.

[1693] An "online community" is a place where users can interact with each other and share information via the Internet.

[1694] "Monitoring" is the act of observing and monitoring specific situations or data in real time.

[1695] "Offensive comments" are statements that contain the intent to insult or attack others.

[1696] A "quarrel" refers to a clash caused by an argument or a difference of opinion.

[1697] "Rewriting" is the act of reconstructing a sentence and rewriting it in a different way.

[1698] A "generative AI model" is an artificial intelligence model that is trained to perform generative tasks.

[1699] "Selection" is the act of choosing an appropriate subject.

[1700] An "empathetic comment" is a statement made to show empathy for the feelings or opinions of others.

[1701] "Neutral and constructive advice" is advice that is unbiased and aims to lead in a positive direction.

[1702] A "system" is a collection of elements organized to achieve a particular purpose.

[1703] A "physical store" is a store or service facility that exists in a physical location.

[1704] "Customer service" refers to the act of responding to customer requests and questions.

[1705] "Inter-staff dialogue" refers to communication between employees who belong to the same workplace.

[1706] A "smartphone" is a mobile device that has similar communication capabilities to a mobile phone, but with more advanced computing capabilities.

[1707] An "emotion engine" is a technology that analyzes and recognizes a user's emotions from text and physiological data.

[1708] "Real-time filtering" is the process of instantly analyzing data or information and sorting it based on specific criteria.

[1709] "Visualization" is the process of expressing abstract data or information in a visually easy-to-understand form.

[1710] The present invention provides a system for monitoring communications in online communities and brick-and-mortar stores and effectively resolving offensive comments and conflicts. Hereinafter, embodiments of the present invention will be described in detail.

[1711] System Program

[1712] The system of the present invention is implemented using the following hardware and software:

[1713] Hardware: Smartphones, servers

[1714] Software: Natural language processing (NLP) libraries (e.g., spaCy), sentiment analysis tools (e.g., AWS Comprehend, Google Cloud Natural Language API), generative AI models (e.g., OpenAI GPT)

[1715] Process Overview

[1716] Monitoring and Detection

[1717] The server monitors communication content in physical stores and online communities in real time. This monitoring includes using NLP technology to detect offensive words and phrases. For example, if a user types an offensive comment such as "it's unusable" into their smartphone, the content will be detected immediately.

[1718] Comment Rewrite

[1719] Detected offensive comments are automatically rewritten to a more polite style using a generative AI model. The server inputs appropriate prompts to the generative AI model and receives the rewritten comments. For example, a comment such as "The service on this product was poor!" is rewritten to "There seems to be some room for improvement in the service on this product."

[1720] Visualizing emotions

[1721] The server uses an emotion engine to analyze the user's emotions and notify customers and staff of negative emotions. The results of the emotion analysis are visualized and displayed in real time on smartphones and other devices. For example, if "anger" is detected based on a customer's text and physiological data, that information is displayed on the staff's device.

[1722] Empathy comment generation

[1723] The selected generative AI model generates empathetic comments based on the detected emotional information, which promotes conflict resolution. The server reads the output of the emotion engine and inputs a prompt sentence to generate an empathetic comment. For example, the prompt sentence used is "Generate an empathetic comment for the following message: The service on this product was poor!"

[1724] Providing constructive advice

[1725] Neutral and constructive advice is generated by the generative AI model and provided to store staff and customers in real time, enabling all parties involved to find a better solution. The server inputs prompts offering constructive advice based on the content of the conflict, and the generated advice is displayed on the device.

[1726] Specific examples

[1727] Handling customer complaints

[1728] If a customer says, "The service on this product was poor!"

[1729] 1. The server detects this offensive comment.

[1730] 2. Using the generative AI model, rewrite the sentence to "There seems to be some room for improvement in how this product is handled."

[1731] 3. The emotion engine detects anger and notifies staff.

[1732] 4. The server inputs the prompt text "Generate an empathetic comment for the following message: The service on this product was poor!" and generates an empathetic comment.

[1733] 5. Provide the staff with a sympathetic comment saying, "We take your feedback seriously and will strive to improve," along with constructive advice saying, "We would appreciate it if you could tell us the specific issues you had."

[1734] This will enable us to provide a healthy communication environment in our physical stores and online communities.

[1735] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1736] Step 1:

[1737] The server monitors communication content in physical stores and online communities. The input includes user text messages, and the data is analyzed using NLP techniques. The output detects messages containing offensive comments or negative language.

[1738] Step 2:

[1739] The server inputs the detected offensive comments into a generative AI model, which rewrites them in a polite style. The input includes an offensive text message, which is passed to the generative AI model along with a prompt. The output is a rewritten, polite comment. For example, a comment like "The service on this product was poor!" is changed to "There seems to be some room for improvement in the service on this product."

[1740] Step 3:

[1741] The server uses an emotion engine to analyze emotions from the user's text message and physiological indicators (e.g., heart rate, sweat rate, blink rate). The input includes the rewritten text message and physiological data. The output is the emotion analysis result, which detects emotions such as "anger" or "sadness."

[1742] Step 4:

[1743] The server generates an empathetic comment based on the analysis results of the emotion engine. The input includes the emotion analysis results and the rewritten text message, which are passed to the generative AI model along with a prompt. The output is a comment expressing empathy. For example, a prompt such as "Generate an empathetic comment for the following message: The service on this product was poor!" is used.

[1744] Step 5:

[1745] The server generates neutral and constructive advice. The input includes the sentiment analysis results and empathetic comments, which are passed to the generative AI model along with a prompt. The output is constructive advice. For example, the generated advice might be, "We would appreciate it if you could tell us the specific problem you are having."

[1746] Step 6:

[1747] The terminal displays the rewritten comments, empathetic comments, and constructive advice sent from the server to the user. The input includes the rewritten comments, empathetic comments, and advice from the server. The output is displayed on the user's screen, promoting healthy communication.

[1748] Step 7:

[1749] The user can smoothly communicate based on the rewritten comments, empathetic comments, and constructive advice displayed on the terminal. The input includes the content displayed on the terminal, and the user's response and behavior are obtained as the output.

[1750] 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.

[1751] 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.

[1752] 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.

[1753] 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.

[1754] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1755] 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.

[1756] 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).

[1757] 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.

[1758] 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."

[1759] 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.

[1760] 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).

[1761] 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.

[1762] 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.

[1763] 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.

[1764] 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.

[1765] 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.

[1766] 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.

[1767] 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.

[1768] 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.

[1769] 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.

[1770] 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.

[1771] The following is further disclosed regarding the above embodiment.

[1772] (Claim 1)

[1773] A means of monitoring online communities to automatically detect offensive comments and arguments;

[1774] means for analyzing the detected comments and rewriting them in a polite style;

[1775] A means for selecting a generative AI model with specialized knowledge;

[1776] A means to sort out conflicting opinions and generate empathetic comments using a selected generative AI model;

[1777] A system that includes a means for generating neutral and constructive advice.

[1778] (Claim 2)

[1779] 10. The system of claim 1, wherein the monitoring means uses natural language processing techniques.

[1780] (Claim 3)

[1781] 2. The system of claim 1, wherein the rewriting means uses an algorithm to automatically convert offensive comments into polite language.

[1782] "Example 1"

[1783] (Claim 1)

[1784] Automated means of monitoring online communities to detect offensive comments and conflicts;

[1785] A means for analyzing the detected comments to identify a subject of conflict and select an appropriate generative AI model;

[1786] A means to automatically rewrite offensive comments into polite language using selected generative AI models, and

[1787] A means for the generative AI model to organize conflicting opinions and generate empathetic comments;

[1788] A system that includes a means for generating neutral and constructive advice.

[1789] (Claim 2)

[1790] 10. The system of claim 1, wherein the monitoring means uses natural language processing techniques.

[1791] (Claim 3)

[1792] The system of claim 1, wherein the rewriting means uses a generative AI model that automatically transforms offensive comments into polite writing.

[1793] "Application Example 1"

[1794] (Claim 1)

[1795] A means of monitoring online communities to automatically detect offensive comments and arguments;

[1796] means for analyzing the detected comments and rewriting them in a polite style;

[1797] A means for selecting a generative AI model with specialized knowledge;

[1798] A means to sort out conflicting opinions and generate empathetic comments using a selected generative AI model;

[1799] A means of generating neutral and constructive advice;

[1800] A means for monitoring comment sections in applications used on smart devices to detect offensive comments in real time;

[1801] A system that includes a means to rewrite offensive comments into polite expressions and generate empathetic comments and constructive advice.

[1802] (Claim 2)

[1803] 10. The system of claim 1, wherein the monitoring means uses natural language processing techniques.

[1804] (Claim 3)

[1805] 2. The system of claim 1, wherein the rewriting means uses an algorithm to automatically convert offensive comments into polite language.

[1806] "Example 2: Combining Emotion Engines"

[1807] (Claim 1)

[1808] A means of monitoring online communities to automatically detect offensive comments and arguments;

[1809] means for analyzing the detected comments and rewriting them in a polite style;

[1810] A means for selecting a generative AI model with specialized knowledge;

[1811] A means to sort out conflicting opinions and generate empathetic comments using a selected generative AI model;

[1812] A means of generating neutral and constructive advice;

[1813] means for recognizing emotions using text messages and physiological indicators of a user;

[1814] A system that includes a means to rewrite offensive comments based on emotion recognition results.

[1815] (Claim 2)

[1816] 10. The system of claim 1, wherein the monitoring means uses natural language processing techniques.

[1817] (Claim 3)

[1818] 2. The system of claim 1, wherein the rewriting means uses an algorithm to automatically convert offensive comments into polite language.

[1819] "Application example 2 when combining emotion engines"

[1820] (Claim 1)

[1821] A means of monitoring online communities to automatically detect offensive comments and arguments;

[1822] means for analyzing the detected comments and rewriting them in a polite style;

[1823] A means for selecting a generative AI model with specialized knowledge;

[1824] A means to sort out conflicting opinions and generate empathetic comments using a selected generative AI model;

[1825] A means of generating neutral and constructive advice;

[1826] A means to apply the monitored communication content to customer service and staff interactions in physical stores,

[1827] A means to provide real-time filtering and rewriting of offensive comments via smartphone;

[1828] A means for visualizing customer and staff emotions using an emotion engine;

[1829] A means of displaying empathetic comments and constructive advice in real time in a brick-and-mortar store environment

[1830] A system including:

[1831] (Claim 2)

[1832] 10. The system of claim 1, wherein the monitoring means uses natural language processing techniques.

[1833] (Claim 3)

[1834] 2. The system of claim 1, wherein the rewriting means uses an algorithm to automatically convert offensive comments into polite language. [Explanation of symbols]

[1835] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. A means of monitoring online communities to automatically detect offensive comments and arguments; means for analyzing the detected comments and rewriting them in a polite style; A means for selecting a generative AI model with specialized knowledge; A means for sorting out conflicting opinions and generating empathetic comments using the selected generation AI model; A system that includes a means for generating neutral and constructive advice.

2. The system of claim 1 , wherein the monitoring means uses natural language processing techniques.

3. 2. The system of claim 1, wherein the rewriting means uses an algorithm to automatically convert offensive comments into polite language.

Images