System

Abstract

A system is provided.SOLUTION: A system including a means for receiving a content posted by a user, a means for analyzing the received posted content using a natural language processing model, a means for determining whether the posted content is malicious based on an analysis result, a means for generating background information or additional information using a generative artificial intelligence when the posted content is determined to be malicious, a means for creating a description note based on the generated background information or the additional information, and a means for adding the description note to the original posted content and displaying the description note for the user.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] Malicious posts are common in online communications and can affect the credibility of companies and individuals. Such posts can damage a company's brand value and lead to misunderstandings among consumers. In particular, inappropriate posts directed at a specific company can significantly damage that company's reputation. However, simply deleting or reporting posts is often insufficient in terms of preventing recurrence or educating users. Therefore, a system is needed that can automatically detect malicious posts and immediately provide background information and corrections to accurately convey information to users. [Means for solving the problem]

[0005] The present invention solves the above problems by the following means. First, a means for receiving content posted by a user is provided. Next, a means for analyzing the received posted content using a natural language processing model is provided. Next, a means for determining whether the posted content is malicious based on the analysis results is provided. If the posted content is determined to be malicious, a means for generating background information and additional information using generative artificial intelligence is provided. A means for creating an explanatory note based on the generated background information and additional information is provided. Finally, a system is provided that includes a means for adding the explanatory note to the original post and displaying it to the user. This system allows users to obtain accurate information, prevents the recurrence of malicious posts, and helps create a healthy communication environment.

[0006] "User" means any person or entity that uses the System to enter and submit Submissions.

[0007] "Posted Content" refers to a text message that a user enters into the system and sends.

[0008] "Means of receiving" refers to the function of receiving the content posted by the user and incorporating it into the system.

[0009] "Natural language processing model" refers to a machine learning model used to understand and analyze human language.

[0010] "Means of analysis" refers to the ability to analyze the context and sentiment of posts using natural language processing models.

[0011] "Means of determination" refers to the function of determining whether the content of a post is malicious based on the analysis results.

[0012] "Generative AI" refers to AI that automatically generates background and additional information based on specific input.

[0013] "Background information" refers to information that provides fact-checking or context related to the content of a post.

[0014] "Additional information" refers to information used to correct or supplement the posted content.

[0015] An "explanatory note" is a document containing background or additional information generated by generative artificial intelligence.

[0016] "Means for displaying" refers to the ability to add explanatory notes to the original post and visually present them to the user. [Brief explanation of the drawings]

[0017] [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

[0018] 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.

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

[0020] 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).

[0021] 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.

[0022] 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.

[0023] 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.

[0024] 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."

[0025] [First embodiment]

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

[0027] 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.

[0028] 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).

[0029] 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.

[0030] 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.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form 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.

[0032] 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.

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

[0034] 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.

[0035] 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.

[0036] 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.

[0037] 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."

[0038] The present invention provides a system that analyzes posts sent by users, detects malicious content, and adds background information and correction information. The overall configuration of the system and each processing step are described in detail below.

[0039] First, the user inputs the content of the post through their terminal and sends it to the server. The server receives the content and temporarily stores it in preparation for the next step. At this stage, the user directly interacts with the system.

[0040] The server then sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context, word choice, and emotional expression of the post to determine whether the post is malicious. The results of this analysis are returned to the server.

[0041] If the server determines that a post is malicious based on the analysis results, it invokes a generative artificial intelligence (AI) to generate background information and corrections. The generative AI gathers relevant information from the internet and databases and automatically creates appropriate corrections and additional information for the post. This information may include specific fact-checking and statistical data.

[0042] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model will determine that the post is malicious. The generative AI will generate background information, such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0043] The server creates an "explanatory note" based on this generated background information. The explanatory note contains specific information obtained from the generative AI. The explanatory note is provided to the user as an attachment to the original post.

[0044] Finally, the corrected post and explanatory notes are sent to the device and displayed to the user, who can then review the displayed explanatory notes to obtain accurate information. Through this process, users can correct malicious posts and spread more accurate information.

[0045] The system of the present invention contributes to the prevention of inappropriate posts by users and the provision of accurate information. As a result, it is possible to improve the quality of online communication and protect the reputations of companies and individuals. This system gives users the opportunity to reconsider their posts and encourages them to learn how to avoid misleading information.

[0046] The processing flow will be explained below.

[0047] Step 1:

[0048] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0049] Step 2:

[0050] The server receives the posted content sent by the user and temporarily stores it in a database.

[0051] Step 3:

[0052] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the context and sentiment of the posts.

[0053] Step 4:

[0054] The analysis results from the natural language processing model are returned to the server, which then receives the analysis results and determines whether the post is malicious.

[0055] Step 5:

[0056] If a post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and corrections related to the post.

[0057] Step 6:

[0058] The background information and corrections generated by the generative AI are returned to the server, which then creates an "explanatory note" based on this information.

[0059] Step 7:

[0060] The server appends the generated explanatory note to the original post to create a modified post.

[0061] Step 8:

[0062] The server sends the modified post and explanatory notes to the device.

[0063] Step 9:

[0064] The terminal displays the received corrected post and explanatory notes to the user, who can then check the displayed explanatory notes to obtain accurate information.

[0065] This process allows the entire system to cooperate, curb malicious postings, and contribute to the provision of accurate information.

[0066] Example 1

[0067] 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."

[0068] In recent years, the amount of communication via user posts on the Internet has increased dramatically, but inappropriate posts and malicious information have become more prominent. Such posts not only damage the reputations of companies and individuals, but also risk causing misunderstandings and confusion. Therefore, there is a growing need for a system that can automatically detect malicious posts and add appropriate background information and corrections.

[0069] 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.

[0070] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and additional information using a generative artificial intelligence if the posted content is determined to be malicious, means for creating an explanatory note based on the generated background information and additional information, means for adding the explanatory note to the original post and displaying it to the user, and means for sending a prompt message to the generative artificial intelligence to generate appropriate background information and additional information. This makes it possible to automatically detect malicious posts and add appropriate information to present to the user.

[0071] A "user" is an operator of a terminal who uses the system to input and send posting content.

[0072] "Posted content" is text data that a user inputs via a terminal and sends to the server.

[0073] "Server" refers to a central computer system that receives, stores, analyzes, and provides generated information to users.

[0074] A "natural language processing model" is an algorithm or machine learning model for analyzing text data and understanding sentiment and context. Examples include high-performance models such as BERT and GPT-3 (registered trademark).

[0075] The "analysis results" are the evaluation results of the posted content generated by the natural language processing model, and include information on whether the post is malicious or not.

[0076] "Generative AI" refers to algorithms or machine learning models that gather information from the internet or databases and generate necessary background information or corrections.

[0077] "Background information" is supplementary information related to the content of a post, and is information that complements the context and situation of the post.

[0078] "Additional information" is information used to correct or supplement the posted content, including accurate data and statistical information.

[0079] An "explanatory note" is a document created based on background information and additional information generated by generative artificial intelligence, and is intended to provide explanations and corrections to the user.

[0080] A "prompt sentence" is an instruction sentence that requests the generative artificial intelligence to generate background information or additional information.

[0081] The present invention provides a system for analyzing posts sent by users, detecting malicious content, and adding background information and correction information. Specific implementation methods of the system will be described in detail below.

[0082] First, the user enters the content of the post through their device and sends it to the server. The device can be a PC, smartphone, tablet, etc. The user enters the content of the post into the text box using the device's browser or a dedicated app, and presses the send button. For example, the user might enter, "The service is so bad! I can't connect at all!"

[0083] Next, the server receives the post sent by the user and temporarily stores it. The server is a high-performance computer, and the database is MySQL (registered trademark) or PostgreSQL. The text data included in the body of the post request, "The service is terrible! I can't connect at all!", is stored.

[0084] The server then sends the received post content to a natural language processing (NLP) model for analysis. NLP models such as BERT and GPT-3 are used. The server calls the model's API and analyzes the post content. For example, a post such as "The service is so bad! I can't connect at all!" is judged to be "malicious."

[0085] Based on the analysis results, the server uses generative artificial intelligence (AI) to generate background and additional information. An example of a generative AI model is GPT-3. By sending specific prompts to this model, it automatically generates relevant information. An example of a prompt is, "When a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service."

[0086] The server then creates an "explanatory note" based on the generated background information and additional information. The explanatory note contains specific details obtained from the generative AI and is formatted as a text file or HTML. For example, the note might include information such as, "We have recently installed many new base stations and are working to improve the quality of our service."

[0087] Finally, the server sends the created explanatory note and the corrections to the original post to the user's device, and displays them in the browser or dedicated app. For example, the original post "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0088] This system allows users to check the background information and corrections of posted content and obtain accurate information. This will help prevent users from posting inappropriate content and provide accurate information, improving the quality of online communication. It will also help protect the reputations of companies and individuals and promote learning to avoid misunderstandings.

[0089] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0090] Step 1:

[0091] The user enters the content of the post and sends it. Using a device (PC or smartphone), the user enters the content of the post into a text box on a browser or dedicated app. For example, the user might enter "The service is so bad! I can't connect at all!" and click the "Send" button. The data entered is in text format, and the device sends this input data to the server as a post request.

[0092] Step 2:

[0093] The server receives the posted content and temporarily stores it. The server catches the incoming request and extracts the posted content. The received data is text data saying "The service is so bad! I can't connect at all!" The server temporarily stores this data in a database (e.g. MySQL, PostgreSQL). Here, data processing is performed, converting the format so that it can be stored appropriately in the database. The output is the temporarily stored posted data.

[0094] Step 3:

[0095] The server analyzes the post content using a natural language processing (NLP) model. The server retrieves the saved post content and sends it to an NLP model (e.g., BERT, GPT-3). The input includes the text "The service is so bad! It doesn't connect at all!" The model analyzes this text and infers whether the post is malicious based on sentiment and keywords. The output is an analysis result such as "Malicious." Specifically, the server calls the model's API and sends an analysis request in JSON format.

[0096] Step 4:

[0097] Based on the analysis results, the server requests the generative AI model to generate background information and correction information. The server receives the analysis result of "malicious" obtained from the NLP model and sends a prompt message to the generative AI model based on that result. The input sent is the prompt message, "If a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service." The AI ​​model generates background information in response, and the output is the information, "We have recently installed many new base stations and are working to improve the quality of our service."

[0098] Step 5:

[0099] The server creates an "explanatory note" based on the generated information. The server creates an "explanatory note" based on background information and additional information obtained from the generative AI model. The input includes the generated information (e.g., "We have recently installed many new base stations and are working to improve the quality of our service") and formats it in a format that is easy for the user to understand. The output is an explanatory note formatted as a text file or HTML, and specific operations include format conversion and the use of document generation tools.

[0100] Step 6:

[0101] The server sends the corrected post and explanatory note to the user's device. The server combines the created explanatory note with the corrections to the original post and sends it to the user's device. The input includes the corrected post and explanatory note. Specifically, the server generates a response and sends it to the browser or dedicated app. The output is the corrected post and explanatory note displayed on the user's device. For example, a post saying "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0102] Through these processing steps, users can see how their posts have been analyzed and what contextual information has been added, which is expected to improve the quality of online communication.

[0103] (Application example 1)

[0104] 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."

[0105] Malicious and inappropriate posts are frequently found on social media and message boards on the Internet. Such posts not only damage the reputations of individuals and companies, but also encourage the spread of misinformation and degrade the quality of online communication. A system is needed to detect malicious posts in real time and provide appropriate background information and corrections to clear up users' misunderstandings and prevent the spread of misinformation.

[0106] 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.

[0107] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and correction information using generative artificial intelligence if the posted content is determined to be malicious, means for generating a reminder based on the generated background information and correction information, means for adding the reminder to the original post and displaying it to the user, and means for analyzing the posted content in real time, thereby enabling the detection and correction of inappropriate posted content in real time.

[0108] "User" means an individual or organization that posts using the system.

[0109] "Postings" means text or messages that users send through the system.

[0110] A "natural language processing model" is a collection of algorithms and software that analyzes text data and understands its meaning and sentiment.

[0111] "Analysis results" refers to the information and data obtained after the natural language processing model analyzes the content of a post.

[0112] "Hateful posts" are inappropriate posts that contain negative emotions or intentions.

[0113] "Generative artificial intelligence" refers to an AI system that has the ability to collect appropriate background information and correction information from the internet or databases and generate new information.

[0114] "Background information" refers to factual and evidence-based explanations or information related to the content of a post.

[0115] "Correction information" means accurate information to correct a misunderstanding or error.

[0116] A "reminder" is a notice or statement provided to encourage the user to review or correct the generated background information or correction information.

[0117] "Analyzing in real time" means that the content is analyzed immediately after the user inputs and sends the content to be posted, and the corresponding processing is carried out.

[0118] An "explanatory note" is a document that contains a collection of generated background information and corrections.

[0119] "System" means the combination of equipment and software that performs a series of processes to receive, analyze, determine, and provide correction information for user posts.

[0120] To implement this invention, a series of processes involving a server, a terminal, and a user are required. Each step proceeds as follows:

[0121] First, when a user posts on social media or a message board, the user device sends the post to a server, which then temporarily stores the received post.

[0122] The server then sends the received content to a natural language processing (NLP) model to analyze it. The NLP model is designed to analyze the context, keywords, and sentiment of the text data to determine whether the content is malicious. This analysis is performed using libraries such as the spacy and transformers libraries.

[0123] Once the analysis results are returned, the server uses them to determine whether the post is malicious. If it is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and corrections for the post. The generative AI collects relevant information from external databases and the Internet to provide accurate and appropriate information. For example, if a user posts, "The service is so bad! I can't connect at all!", the AI ​​generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0124] Based on the generated background and correction information, the server generates a reminder and displays it to the user along with the original post, allowing the user to review their post and make corrections if necessary.

[0125] To achieve real-time post analysis, the server must process data quickly and efficiently. This requires a high-performance CPU, large memory capacity, and a high-speed network connection. In addition to the aforementioned natural language processing model and generative artificial intelligence, the server also uses a database management system.

[0126] For example, here's a prompt that might be used when a user posts, "The service is so bad! I can't get through at all!":

[0127] User Post: "The service is terrible! I can't connect at all!"

[0128] Generative AI prompt: "Please provide information about recent improvements in the quality of your telecommunications services."

[0129] This allows the generative AI to provide appropriate context and give users the opportunity to correct misinformation, improving the quality of online communication and clearing up misunderstandings.

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

[0131] Step 1: The user enters the content to post on a social media platform or bulletin board and sends it from the device to the server. The input is the user's post (text data), and the output is request data including the post content.

[0132] Step 2: The server temporarily saves the received post. The input is the request data containing the post, and the output is the save operation result in the session database. Specifically, the server stores the post in the database.

[0133] Step 3: The server sends the saved post content to a natural language processing model (NLP model) for analysis. The input is the saved post content, and the output is the analysis result. Specifically, the server calls the NLP model and performs text analysis on the post content.

[0134] Step 4: The NLP model analyzes the post's context, keywords, and sentiment and returns the results. The input is the text data of the post, and the output is the context analysis, keyword matching, and sentiment analysis results. Specifically, the natural language processing algorithm analyzes the text data and scores it for sentiment and keywords.

[0135] Step 5: Based on the analysis results, the server determines whether the post is malicious. The input is the analysis results, and the output is the judgment result (whether malicious or not). Specifically, the server performs a threshold judgment based on the sentiment analysis results and keyword scores.

[0136] Step 6: If the post is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and correction information. The input is the malicious post content and a prompt to the generative AI, and the output is background information and correction information. Specifically, the server sends the prompt to the AI ​​model and obtains relevant information.

[0137] Step 7: The generative AI generates background information and correction information and returns it to the server. The input is the prompt text, and the output is the generated information (background information and correction information). Specifically, the generative AI collects data from the internet and databases and generates appropriate information.

[0138] Step 8: The server generates a reminder based on the generated background information and correction information. The input is the information received from the generation AI, and the output is the reminder (explanatory note). Specifically, the server constructs an explanatory note from the generated information.

[0139] Step 9: Attach the reminder to the original post and display it to the user. The input is the reminder and the original post, and the output is the information displayed on the device. Specifically, the server integrates the reminder into the post and sends it to the user's device.

[0140] Step 10: The user checks the displayed reminder and reconsiders or modifies the post. The input is the displayed reminder and the original post, and the output is the new post by the user. Specifically, the user decides whether to repost or modify the post based on the displayed information. This process allows the user to reconsider the content of their post and, if necessary, modify it to the appropriate information.

[0141] 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.

[0142] This invention is a system that detects malicious posts by analyzing the content of posts entered by users and recognizing emotions. This system combines a natural language processing model with an emotion engine and also utilizes generative artificial intelligence to provide background information and corrections, allowing users to receive accurate information.

[0143] First, the user enters the content of the post through their device and presses the send button. The device then sends this content to the server. The server receives the content sent by the user and temporarily stores it in a database. Next, the server sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context and word choice of the post and passes the content to the system's analysis engine.

[0144] Furthermore, the system includes an emotion engine that recognizes users' emotions from the content of their posts. The emotion engine analyzes keywords and expressions contained in the posts to determine emotions such as anger, dissatisfaction, and sadness. This information is fed back to the analysis engine, which generates comprehensive analysis results.

[0145] The server receives the analysis results from the analysis engine and determines whether the post content is malicious. If the post content is determined to be malicious based on the results of the analysis, it calls on generative artificial intelligence (AI) to generate background information and correction information. The generative AI generates background information and correction information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. This allows information to be provided in a format that is more acceptable to the user.

[0146] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model analyzes the post and the emotion engine recognizes the emotion of anger. Based on this information, the generative AI generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service," and writes an explanatory note in a calm tone.

[0147] Next, the server creates an "explanatory note" based on the generated background information. The explanatory note contains specific information obtained from the generative AI and is provided to the user as an attachment to the original post. The corrected post and explanatory note are sent to the device, which displays them to the user. The user can then check the explanatory note and receive accurate information.

[0148] This invention provides users with an opportunity to reconsider their posts and encourages them to learn how to avoid misleading others. This system contributes to the suppression of inappropriate posts and the provision of accurate information, thereby improving the quality of online communication and protecting the reputations of companies and individuals.

[0149] The processing flow will be explained below.

[0150] Step 1:

[0151] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0152] Step 2:

[0153] The server receives the posted content sent by the user and temporarily stores it in a database.

[0154] Step 3:

[0155] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the post's context and word choice.

[0156] Step 4:

[0157] The server sends the analyzed post content to the emotion engine, which analyzes the keywords and expressions contained in the post content to determine the user's emotion (anger, dissatisfaction, sadness, etc.).

[0158] Step 5:

[0159] The server then integrates the analysis results again based on the user's emotional information recognized by the emotion engine, and receives the integrated analysis results to determine whether the posted content is malicious.

[0160] Step 6:

[0161] If the post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and additional information based on the post's content and emotional information.

[0162] Step 7:

[0163] The generative AI generates background and additional information, which is then returned to the server. The server then uses this information to create an "explanatory note," which is written in a tone that takes into account the user's emotions.

[0164] Step 8:

[0165] The server appends the generated explanatory note to the original post and creates the revised post, preparing it for delivery to the user.

[0166] Step 9:

[0167] The server transmits the corrected post and explanatory notes to the terminal, which displays the received corrected post and explanatory notes to the user.

[0168] Step 10:

[0169] The user checks the explanatory notes displayed on the terminal, and can receive and understand accurate information and appropriate background explanations from the explanatory notes.

[0170] This process allows users to review their posts and spread accurate information, and also helps curb inappropriate posts to protect the reputations of businesses and individuals.

[0171] Example 2

[0172] 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."

[0173] In today's online communication environment, content posted by users often contains negative and malicious language. Such posts not only damage the reputation of companies and individuals, but also risk misleading other users. Furthermore, simply deleting negative posts may result in ignoring users' opinions and lowering their satisfaction. To solve these issues, it is necessary to properly analyze the content of posts and provide users with accurate and precise information.

[0174] 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 a means for receiving content posted by a user, a means for analyzing the received posted content using a natural language processing model, a means for determining whether the posted content is malicious based on the analysis result, a means for generating background information and supplemental information using generative artificial intelligence if the posted content is determined to be malicious, a means for creating an explanatory text based on the generated background information and supplemental information, and a means for adding the explanatory text to the original post and displaying it to the user. This allows the posted content to be appropriately analyzed, enabling the user to receive accurate and precise information.

[0175] A "User" is a person who accesses the Service or System and creates and inputs Posted Content.

[0176] "Posted content" refers to information such as messages, opinions, and feedback sent by users through the system.

[0177] The "receiving means" is a function or process by which the server acquires the posted content sent by the user.

[0178] A "natural language processing model" is an algorithm or software that enables a computer to understand and analyze human language.

[0179] The "analysis means" is a function for analyzing received posted content using a natural language processing model and understanding its content and sentiment.

[0180] The "determination means" is a system function for determining whether posted content is malicious based on the analysis results.

[0181] "Generative artificial intelligence" refers to algorithms or processes that use generative AI models to create background or supplemental information.

[0182] "Generation means" is a function that uses generative artificial intelligence to create background information and supplementary information in a form that is easy for users to understand.

[0183] A "description" is information that is created based on the generated background information and supplementary information and added to the original post.

[0184] The "display means" is a function for adding the generated explanatory text to the original post and displaying it to the user.

[0185] "Emotion recognition" is the process by which a natural language processing model or emotion engine analyzes and identifies emotions in posted content.

[0186] The present invention is a system that analyzes content posted by users, determines whether the content is malicious, and uses generative artificial intelligence to provide background and supplemental information. This system uses a combination of a natural language processing model and an emotion engine to deeply understand the posted content and generate appropriate feedback. Specific techniques for implementing the present invention are described in detail below.

[0187] Hardware and software used

[0188] This system uses the following hardware and software:

[0189] 1. Server: Use a high-performance server computer. You can use MySQL or PostgreSQL as the database server, and Apache (registered trademark) or Nginx as the web server.

[0190] 2. Terminal: The device through which the user accesses the system, including PCs, smartphones, tablets, etc.

[0191] 3. Natural Language Processing Models (NLP): Use advanced NLP models such as BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer).

[0192] 4. Generative AI: Use GPT-3 and other generative AI models to generate relevant context and supporting information.

[0193] Data processing and calculation

[0194] This system processes and calculates data in the following procedure.

[0195] 1. Receiving and storing your submissions:

[0196] The user uses the terminal to input the content to be posted and presses the send button.

[0197] The terminal sends the posted content to the server via an HTTP request.

[0198] The server stores the received post content in a database such as MySQL.

[0199] 2. Analysis using Natural Language Processing (NLP) models:

[0200] The server sends the received posted content to an NLP model for linguistic analysis.

[0201] The NLP model analyzes the context of the post and word choice and returns the analysis results.

[0202] 3. Emotion Recognition with Emotion Engine:

[0203] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions.

[0204] The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness.

[0205] 4. Malicious intent detection and AI generation:

[0206] The server determines whether the posted content is malicious based on the analysis results from the analysis engine.

[0207] If the content of a post is determined to be malicious, the server calls on generative artificial intelligence to generate background and supplementary information.

[0208] The generative AI generates background and supplementary information with appropriate tone and content based on the user's emotions recognized by the emotion engine.

[0209] 5. Creating and providing explanatory notes:

[0210] The server creates a description based on the generated background information.

[0211] The explanation is added to the original post and provided to the user.

[0212] The revised post and description are sent to the device, which displays it to the user.

[0213] Examples and prompts

[0214] For example, if a user posts, "The service is terrible! I can't connect at all!", the following processing will occur:

[0215] User comments: "The service is terrible! I can't connect at all!"

[0216] Prompt for generative AI model: "An angry user has posted a complaint about your service. Please respond in a calm tone with context."

[0217] The NLP model analyzes the post, and the emotion engine recognizes the emotion of "anger." The server invokes the generative AI to generate background information such as, "We have recently installed many new base stations and are working to improve the quality of our service." The server then creates a description based on the generated background information, and the revised post and description are displayed on the user's device.

[0218] In this way, the present invention allows users to receive information in a manner that is free from misunderstanding, thereby improving the quality of online communication.

[0219] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0220] Step 1:

[0221] The user uses the device to input posted content that represents their own opinion or feedback. The input posted content is the input in step 1. Specifically, the user uses the keyboard or touch screen to input "The service is so bad! I can't connect at all!" This posted content is stored in the text input field on the device.

[0222] Step 2:

[0223] When the user presses the send button, the entered post content is sent from the terminal to the server. The click event of the send button generates an HTTP request and the posted content is sent to the server. The input is the posted content entered by the user, and the output is the posted content received by the server.

[0224] Step 3:

[0225] The server receives the posted content sent by the user and temporarily stores it in a database. Specifically, it receives an HTTP request and inserts the content into a database such as MySQL or PostgreSQL. The input is the posted content sent, and the output is the posted content stored in the database.

[0226] Step 4:

[0227] The server sends the received post content to a natural language processing (NLP) model and begins analysis. Specifically, the server-side program sends an HTTP request to the NLP model via a REST API to obtain the analysis results. The input is the post content stored in the database, and the output is the analysis results returned by the NLP model.

[0228] Step 5:

[0229] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions. The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness. Specifically, the emotion engine receives the analysis results from the NLP model and performs emotion analysis. The input is the analysis results of the NLP model, and the output is the emotion recognition results from the emotion engine.

[0230] Step 6:

[0231] The server determines whether the posted content is malicious based on the analysis results from the analysis engine. If the posted content is determined to be malicious, it calls on a generative artificial intelligence (generative AI model) to generate background information and supplementary information. The generative AI generates background information and supplementary information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. The input is the emotion recognition result, and the output is the generated result from the generative AI model.

[0232] Step 7:

[0233] The server creates a description based on the generated background information. For example, using the background information obtained from the generative AI model, it creates a description such as, "We have recently installed many new base stations and are working to improve the quality of our service." The input is the result generated by the generative AI model, and the output is the created description.

[0234] Step 8:

[0235] The modified post and description are sent from the server to the user's device, which then displays them to the user. The server sends the modified post and description as an HTTP response. The input is the created description and the modified post content, and the output is what is displayed on the user's device. The user can check the displayed description and receive accurate information.

[0236] (Application example 2)

[0237] 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."

[0238] On modern online shopping sites, customer reviews are an important source of information for other customers. However, some customer reviews are emotional, overly negative, and contain factually incorrect information. This can lead to the spread of inaccurate information, misrepresenting product ratings, and damaging a company's reputation. To solve this problem, a system is needed that can appropriately address inaccurate or emotional posts and provide background information or explanatory notes.

[0239] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving content posted by a user; means for analyzing the received content using a natural language processing model; means for determining whether the content of the post is malicious based on the analysis result; means for generating background information and additional information using generative artificial intelligence if the content is determined to be malicious; means for creating an explanatory note based on the generated background information and additional information; means for adding the explanatory note to the original post and displaying it to the user; means for monitoring reviews on the online shopping site in real time and determining whether the content of reviews posted by users is appropriate; and means for generating background information and correction information for reviews determined to be inappropriate using generative artificial intelligence and presenting the explanatory note to the user. This makes it possible to prevent the spread of inaccurate information on the online shopping site and provide customers with accurate and reliable information.

[0240] The "means for receiving content posted by users" refers to a mechanism for transmitting text data entered by users via their electronic devices to a server and receiving it.

[0241] "Means for analyzing received posts using a natural language processing model" refers to a mechanism in which the server uses a computer program to analyze the text data it receives and understand the meaning and context of the text.

[0242] "Means for determining whether the content of a post is malicious based on the analysis results" refers to a mechanism for evaluating the analysis data obtained from a natural language processing model to determine whether the content of a post is offensive, insulting, or misleading.

[0243] "Means of using generative AI to generate background and additional information when a post is determined to be malicious" refers to a mechanism that uses generative AI to automatically create information to supplement the situation for posts that are detected to be malicious.

[0244] The "means for creating explanatory notes based on the generated background information and additional information" is a mechanism for organizing the generated supplementary information and creating a document to present it to the user in an easy-to-understand format.

[0245] The "means for adding an explanatory note to the original post and displaying it to the user" is a mechanism for associating the created explanatory note with the content of the original post and displaying it in a format that is easy for the user to view.

[0246] "Means for monitoring reviews on online shopping sites in real time and determining whether the content of reviews posted by users is appropriate" is a system for constantly observing customer reviews on online shopping sites and instantly evaluating the appropriateness of the posts.

[0247] "Means of using generative AI to generate background information and correction information for reviews that are judged to be inappropriate and presenting it to the user as an explanatory note" refers to a mechanism that, when an inappropriate review is detected, uses generative AI to create information and background explanations for correction and presents them to the user as an explanatory note.

[0248] The present invention is a system that analyzes the content of reviews posted by users through an integrated system via electronic devices in real time, generates appropriate background information and correction information as needed, and provides it to the user as explanatory notes.

[0249] First, the review content posted by the user is sent to the server. The server receives the text data using a means for receiving content posted by the user. Next, the received post content is analyzed using a natural language processing (NLP) model. Examples of models used here include spaCy and TextBlob. This analysis analyzes the context and sentiment of the review.

[0250] The server uses an emotion engine to determine whether the post is malicious based on the analysis results. The emotion engine analyzes keywords and expressions contained in the post to determine emotions such as anger, dissatisfaction, and sadness. This determines whether the post is appropriate, and only if it is not is it used to generate background information or corrections using generative artificial intelligence (AI).

[0251] The generative AI used is OpenAI's GPT-3. This generative AI uses user reviews and analysis results as prompts to generate appropriate background information and corrections. An example of a generative prompt is as follows:

[0252] User Review: The quality of this product is terrible!

[0253] Sentiment score: -0.8

[0254] Generate background information.

[0255] The server then uses the generated background and additional information to create explanatory notes. These notes contain information obtained from the generative AI and are presented in a format that is easy for the user to understand. For example, the notes might state, "To improve product quality, we are strengthening our inspection system and introducing a new quality control process."

[0256] Finally, the server adds explanatory notes to the original post and displays them to the user. This system prevents the spread of inaccurate information on online shopping sites and allows customers to obtain reliable information. The entire system operates in real time and aims to improve the quality of reviews on online shopping sites.

[0257] The above is a specific embodiment of the system according to the present invention.

[0258] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0259] Step 1:

[0260] Users post reviews using their electronic devices. The posted content is sent from the user's device to the server. The input is text data entered by the user, and the output is text data sent to the server.

[0261] Step 2:

[0262] The server receives the content posted by the user. The input is text data sent from the terminal, and the received data is temporarily stored in a database. The output is the stored text data.

[0263] Step 3:

[0264] The server analyzes the received post content using a natural language processing (NLP) model. In this step, the context and meaning of the post are analyzed. The input is the stored text data, and the output is the analysis result data. Specifically, the server analyzes the text using NLP tools such as spaCy or TextBlob to extract semantic information.

[0265] Step 4:

[0266] Based on the analysis results, the server determines whether the post content is malicious. The emotion engine analyzes keywords and emotion scores to determine whether the post is malicious. The input is the data obtained by NLP analysis, and the output is the judgment result. Specifically, the emotion analysis algorithm is used to calculate the emotion score of the text.

[0267] Step 5:

[0268] If a post is determined to be malicious, the server uses generative artificial intelligence (AI) to generate background information and additional information. The input is the judgment result and the original text data, and the output is the generated background information and additional information. Specifically, it uses OpenAI (registered trademark) GPT-3 or similar to generate prompt sentences like the following:

[0269] User Review: The quality of this product is terrible!

[0270] Sentiment score: -0.8

[0271] Generate background information.

[0272] The generative AI generates background and additional information based on this prompt.

[0273] Step 6:

[0274] The server creates explanatory notes based on the generated background information and additional information. The input is the generated background information and additional information, and the output is explanatory notes. Specifically, the server arranges the background information and additional information into a consistent format and compiles them into a note to explain to the user.

[0275] Step 7:

[0276] Finally, the server adds the explanatory note to the original post and displays it to the user. The input is the original post and the explanatory note, and the output is the corrected review displayed on the user's device. The user checks the explanatory note through their device and receives the correct information. Specifically, the corrected text data is displayed through the user interface.

[0277] 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.

[0278] 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.

[0279] 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.

[0280] [Second embodiment]

[0281] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0282] 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.

[0283] 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).

[0284] 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.

[0285] 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.

[0286] 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).

[0287] 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.

[0288] 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.

[0289] 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.

[0290] 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.

[0291] 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.

[0292] 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."

[0293] The present invention provides a system that analyzes posts sent by users, detects malicious content, and adds background information and correction information. The overall configuration of the system and each processing step are described in detail below.

[0294] First, the user inputs the content of the post through their terminal and sends it to the server. The server receives the content and temporarily stores it in preparation for the next step. At this stage, the user directly interacts with the system.

[0295] The server then sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context, word choice, and emotional expression of the post to determine whether the post is malicious. The results of this analysis are returned to the server.

[0296] If the server determines that a post is malicious based on the analysis results, it invokes a generative artificial intelligence (AI) to generate background information and corrections. The generative AI gathers relevant information from the internet and databases and automatically creates appropriate corrections and additional information for the post. This information may include specific fact-checking and statistical data.

[0297] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model will determine that the post is malicious. The generative AI will generate background information, such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0298] The server creates an "explanatory note" based on this generated background information. The explanatory note contains specific information obtained from the generative AI. The explanatory note is provided to the user as an attachment to the original post.

[0299] Finally, the corrected post and explanatory notes are sent to the device and displayed to the user, who can then review the displayed explanatory notes to obtain accurate information. Through this process, users can correct malicious posts and spread more accurate information.

[0300] The system of the present invention contributes to the prevention of inappropriate posts by users and the provision of accurate information. As a result, it is possible to improve the quality of online communication and protect the reputations of companies and individuals. This system gives users the opportunity to reconsider their posts and encourages them to learn how to avoid misleading information.

[0301] The processing flow will be explained below.

[0302] Step 1:

[0303] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0304] Step 2:

[0305] The server receives the posted content sent by the user and temporarily stores it in a database.

[0306] Step 3:

[0307] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the context and sentiment of the posts.

[0308] Step 4:

[0309] The analysis results from the natural language processing model are returned to the server, which then receives the analysis results and determines whether the post is malicious.

[0310] Step 5:

[0311] If a post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and corrections related to the post.

[0312] Step 6:

[0313] The background information and corrections generated by the generative AI are returned to the server, which then creates an "explanatory note" based on this information.

[0314] Step 7:

[0315] The server appends the generated explanatory note to the original post to create a modified post.

[0316] Step 8:

[0317] The server sends the modified post and explanatory notes to the device.

[0318] Step 9:

[0319] The terminal displays the received corrected post and explanatory notes to the user, who can then check the displayed explanatory notes to obtain accurate information.

[0320] This process allows the entire system to cooperate, curb malicious postings, and contribute to the provision of accurate information.

[0321] Example 1

[0322] 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."

[0323] In recent years, the amount of communication via user posts on the Internet has increased dramatically, but inappropriate posts and malicious information have become more prominent. Such posts not only damage the reputations of companies and individuals, but also risk causing misunderstandings and confusion. Therefore, there is a growing need for a system that can automatically detect malicious posts and add appropriate background information and corrections.

[0324] 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.

[0325] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and additional information using a generative artificial intelligence if the posted content is determined to be malicious, means for creating an explanatory note based on the generated background information and additional information, means for adding the explanatory note to the original post and displaying it to the user, and means for sending a prompt message to the generative artificial intelligence to generate appropriate background information and additional information. This makes it possible to automatically detect malicious posts and add appropriate information to present to the user.

[0326] A "user" is an operator of a terminal who uses the system to input and send posting content.

[0327] "Posted content" is text data that a user inputs via a terminal and sends to the server.

[0328] "Server" refers to a central computer system that receives, stores, analyzes, and provides generated information to users.

[0329] A "natural language processing model" is an algorithm or machine learning model for analyzing text data and understanding sentiment and context. Examples include high-performance models such as BERT and GPT-3.

[0330] The "analysis results" are the evaluation results of the posted content generated by the natural language processing model, and include information on whether the post is malicious or not.

[0331] "Generative AI" refers to algorithms or machine learning models that gather information from the internet or databases and generate necessary background information or corrections.

[0332] "Background information" is supplementary information related to the content of a post, and is information that complements the context and situation of the post.

[0333] "Additional information" is information used to correct or supplement the posted content, including accurate data and statistical information.

[0334] An "explanatory note" is a document created based on background information and additional information generated by generative artificial intelligence, and is intended to provide explanations and corrections to the user.

[0335] A "prompt sentence" is an instruction sentence that requests the generative artificial intelligence to generate background information or additional information.

[0336] The present invention provides a system for analyzing posts sent by users, detecting malicious content, and adding background information and correction information. Specific implementation methods of the system will be described in detail below.

[0337] First, the user enters the content of the post through their device and sends it to the server. The device can be a PC, smartphone, tablet, etc. The user enters the content of the post into the text box using the device's browser or a dedicated app, and presses the send button. For example, the user might enter, "The service is so bad! I can't connect at all!"

[0338] Next, the server receives the post sent by the user and temporarily stores it. The server is a high-performance computer, and the database is MySQL or PostgreSQL. The text data included in the body of the post request, "The service is terrible! I can't connect at all!", is stored.

[0339] The server then sends the received post content to a natural language processing (NLP) model for analysis. NLP models such as BERT and GPT-3 are used. The server calls the model's API and analyzes the post content. For example, a post such as "The service is so bad! I can't connect at all!" is judged to be "malicious."

[0340] Based on the analysis results, the server uses generative artificial intelligence (AI) to generate background and additional information. An example of a generative AI model is GPT-3. By sending specific prompts to this model, it automatically generates relevant information. An example of a prompt is, "When a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service."

[0341] The server then creates an "explanatory note" based on the generated background information and additional information. The explanatory note contains specific details obtained from the generative AI and is formatted as a text file or HTML. For example, the note might include information such as, "We have recently installed many new base stations and are working to improve the quality of our service."

[0342] Finally, the server sends the created explanatory note and the corrections to the original post to the user's device, and displays them in the browser or dedicated app. For example, the original post "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0343] This system allows users to check the background information and corrections of posted content and obtain accurate information. This will help prevent users from posting inappropriate content and provide accurate information, improving the quality of online communication. It will also help protect the reputations of companies and individuals and promote learning to avoid misunderstandings.

[0344] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0345] Step 1:

[0346] The user enters the content of the post and sends it. Using a device (PC or smartphone), the user enters the content of the post into a text box on a browser or dedicated app. For example, the user might enter "The service is so bad! I can't connect at all!" and click the "Send" button. The data entered is in text format, and the device sends this input data to the server as a post request.

[0347] Step 2:

[0348] The server receives the posted content and temporarily stores it. The server catches the incoming request and extracts the posted content. The received data is text data saying "The service is so bad! I can't connect at all!" The server temporarily stores this data in a database (e.g. MySQL, PostgreSQL). Here, data processing is performed, converting the format so that it can be stored appropriately in the database. The output is the temporarily stored posted data.

[0349] Step 3:

[0350] The server analyzes the post content using a natural language processing (NLP) model. The server retrieves the saved post content and sends it to an NLP model (e.g., BERT, GPT-3). The input includes the text "The service is so bad! It doesn't connect at all!" The model analyzes this text and infers whether the post is malicious based on sentiment and keywords. The output is an analysis result such as "Malicious." Specifically, the server calls the model's API and sends an analysis request in JSON format.

[0351] Step 4:

[0352] Based on the analysis results, the server requests the generative AI model to generate background information and correction information. The server receives the analysis result of "malicious" obtained from the NLP model and sends a prompt message to the generative AI model based on that result. The input sent is the prompt message, "If a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service." The AI ​​model generates background information in response, and the output is the information, "We have recently installed many new base stations and are working to improve the quality of our service."

[0353] Step 5:

[0354] The server creates an "explanatory note" based on the generated information. The server creates an "explanatory note" based on background information and additional information obtained from the generative AI model. The input includes the generated information (e.g., "We have recently installed many new base stations and are working to improve the quality of our service") and formats it in a format that is easy for the user to understand. The output is an explanatory note formatted as a text file or HTML, and specific operations include format conversion and the use of document generation tools.

[0355] Step 6:

[0356] The server sends the corrected post and explanatory note to the user's device. The server combines the created explanatory note with the corrections to the original post and sends it to the user's device. The input includes the corrected post and explanatory note. Specifically, the server generates a response and sends it to the browser or dedicated app. The output is the corrected post and explanatory note displayed on the user's device. For example, a post saying "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0357] Through these processing steps, users can see how their posts have been analyzed and what contextual information has been added, which is expected to improve the quality of online communication.

[0358] (Application example 1)

[0359] 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."

[0360] Malicious and inappropriate posts are frequently found on social media and message boards on the Internet. Such posts not only damage the reputations of individuals and companies, but also encourage the spread of misinformation and degrade the quality of online communication. A system is needed to detect malicious posts in real time and provide appropriate background information and corrections to clear up users' misunderstandings and prevent the spread of misinformation.

[0361] 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.

[0362] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and correction information using generative artificial intelligence if the posted content is determined to be malicious, means for generating a reminder based on the generated background information and correction information, means for adding the reminder to the original post and displaying it to the user, and means for analyzing the posted content in real time, thereby enabling the detection and correction of inappropriate posted content in real time.

[0363] "User" means an individual or organization that posts using the system.

[0364] "Postings" means text or messages that users send through the system.

[0365] A "natural language processing model" is a collection of algorithms and software that analyzes text data and understands its meaning and sentiment.

[0366] "Analysis results" refers to the information and data obtained after the natural language processing model analyzes the content of a post.

[0367] "Hateful posts" are inappropriate posts that contain negative emotions or intentions.

[0368] "Generative artificial intelligence" refers to an AI system that has the ability to collect appropriate background information and correction information from the internet or databases and generate new information.

[0369] "Background information" refers to factual and evidence-based explanations or information related to the content of a post.

[0370] "Correction information" means accurate information to correct a misunderstanding or error.

[0371] A "reminder" is a notice or statement provided to encourage the user to review or correct the generated background information or correction information.

[0372] "Analyzing in real time" means that the content is analyzed immediately after the user inputs and sends the content to be posted, and the corresponding processing is carried out.

[0373] An "explanatory note" is a document that contains a collection of generated background information and corrections.

[0374] "System" means the combination of equipment and software that performs a series of processes to receive, analyze, determine, and provide correction information for user posts.

[0375] To implement this invention, a series of processes involving a server, a terminal, and a user are required. Each step proceeds as follows:

[0376] First, when a user posts on social media or a message board, the user device sends the post to a server, which then temporarily stores the received post.

[0377] The server then sends the received content to a natural language processing (NLP) model to analyze it. The NLP model is designed to analyze the context, keywords, and sentiment of the text data to determine whether the content is malicious. This analysis is performed using libraries such as the spacy and transformers libraries.

[0378] Once the analysis results are returned, the server uses them to determine whether the post is malicious. If it is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and corrections for the post. The generative AI collects relevant information from external databases and the Internet to provide accurate and appropriate information. For example, if a user posts, "The service is so bad! I can't connect at all!", the AI ​​generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0379] Based on the generated background and correction information, the server generates a reminder and displays it to the user along with the original post, allowing the user to review their post and make corrections if necessary.

[0380] To achieve real-time post analysis, the server must process data quickly and efficiently. This requires a high-performance CPU, large memory capacity, and a high-speed network connection. In addition to the aforementioned natural language processing model and generative artificial intelligence, the server also uses a database management system.

[0381] For example, here's a prompt that might be used when a user posts, "The service is so bad! I can't get through at all!":

[0382] User Post: "The service is terrible! I can't connect at all!"

[0383] Generative AI prompt: "Please provide information about recent improvements in the quality of your telecommunications services."

[0384] This allows the generative AI to provide appropriate context and give users the opportunity to correct misinformation, improving the quality of online communication and clearing up misunderstandings.

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

[0386] Step 1: The user enters the content to post on a social media platform or bulletin board and sends it from the device to the server. The input is the user's post (text data), and the output is request data including the post content.

[0387] Step 2: The server temporarily saves the received post. The input is the request data containing the post, and the output is the save operation result in the session database. Specifically, the server stores the post in the database.

[0388] Step 3: The server sends the saved post content to a natural language processing model (NLP model) for analysis. The input is the saved post content, and the output is the analysis result. Specifically, the server calls the NLP model and performs text analysis on the post content.

[0389] Step 4: The NLP model analyzes the post's context, keywords, and sentiment and returns the results. The input is the text data of the post, and the output is the context analysis, keyword matching, and sentiment analysis results. Specifically, the natural language processing algorithm analyzes the text data and scores it for sentiment and keywords.

[0390] Step 5: Based on the analysis results, the server determines whether the post is malicious. The input is the analysis results, and the output is the judgment result (whether malicious or not). Specifically, the server performs a threshold judgment based on the sentiment analysis results and keyword scores.

[0391] Step 6: If the post is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and correction information. The input is the malicious post content and a prompt to the generative AI, and the output is background information and correction information. Specifically, the server sends the prompt to the AI ​​model and obtains relevant information.

[0392] Step 7: The generative AI generates background information and correction information and returns it to the server. The input is the prompt text, and the output is the generated information (background information and correction information). Specifically, the generative AI collects data from the internet and databases and generates appropriate information.

[0393] Step 8: The server generates a reminder based on the generated background information and correction information. The input is the information received from the generation AI, and the output is the reminder (explanatory note). Specifically, the server constructs an explanatory note from the generated information.

[0394] Step 9: Attach the reminder to the original post and display it to the user. The input is the reminder and the original post, and the output is the information displayed on the device. Specifically, the server integrates the reminder into the post and sends it to the user's device.

[0395] Step 10: The user checks the displayed reminder and reconsiders or modifies the post. The input is the displayed reminder and the original post, and the output is the new post by the user. Specifically, the user decides whether to repost or modify the post based on the displayed information. This process allows the user to reconsider the content of their post and, if necessary, modify it to the appropriate information.

[0396] 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.

[0397] This invention is a system that detects malicious posts by analyzing the content of posts entered by users and recognizing emotions. This system combines a natural language processing model with an emotion engine and also utilizes generative artificial intelligence to provide background information and corrections, allowing users to receive accurate information.

[0398] First, the user enters the content of the post through their device and presses the send button. The device then sends this content to the server. The server receives the content sent by the user and temporarily stores it in a database. Next, the server sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context and word choice of the post and passes the content to the system's analysis engine.

[0399] Furthermore, the system includes an emotion engine that recognizes users' emotions from the content of their posts. The emotion engine analyzes keywords and expressions contained in the posts to determine emotions such as anger, dissatisfaction, and sadness. This information is fed back to the analysis engine, which generates comprehensive analysis results.

[0400] The server receives the analysis results from the analysis engine and determines whether the post content is malicious. If the post content is determined to be malicious based on the results of the analysis, it calls on generative artificial intelligence (AI) to generate background information and correction information. The generative AI generates background information and correction information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. This allows information to be provided in a format that is more acceptable to the user.

[0401] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model analyzes the post and the emotion engine recognizes the emotion of anger. Based on this information, the generative AI generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service," and writes an explanatory note in a calm tone.

[0402] Next, the server creates an "explanatory note" based on the generated background information. The explanatory note contains specific information obtained from the generative AI and is provided to the user as an attachment to the original post. The corrected post and explanatory note are sent to the device, which displays them to the user. The user can then check the explanatory note and receive accurate information.

[0403] This invention provides users with an opportunity to reconsider their posts and encourages them to learn how to avoid misleading others. This system contributes to the suppression of inappropriate posts and the provision of accurate information, thereby improving the quality of online communication and protecting the reputations of companies and individuals.

[0404] The processing flow will be explained below.

[0405] Step 1:

[0406] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0407] Step 2:

[0408] The server receives the posted content sent by the user and temporarily stores it in a database.

[0409] Step 3:

[0410] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the post's context and word choice.

[0411] Step 4:

[0412] The server sends the analyzed post content to the emotion engine, which analyzes the keywords and expressions contained in the post content to determine the user's emotion (anger, dissatisfaction, sadness, etc.).

[0413] Step 5:

[0414] The server then integrates the analysis results again based on the user's emotional information recognized by the emotion engine, and receives the integrated analysis results to determine whether the posted content is malicious.

[0415] Step 6:

[0416] If the post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and additional information based on the post's content and emotional information.

[0417] Step 7:

[0418] The generative AI generates background and additional information, which is then returned to the server. The server then uses this information to create an "explanatory note," which is written in a tone that takes into account the user's emotions.

[0419] Step 8:

[0420] The server appends the generated explanatory note to the original post and creates the revised post, preparing it for delivery to the user.

[0421] Step 9:

[0422] The server transmits the corrected post and explanatory notes to the terminal, which displays the received corrected post and explanatory notes to the user.

[0423] Step 10:

[0424] The user checks the explanatory notes displayed on the terminal, and can receive and understand accurate information and appropriate background explanations from the explanatory notes.

[0425] This process allows users to review their posts and spread accurate information, and also helps curb inappropriate posts to protect the reputations of businesses and individuals.

[0426] Example 2

[0427] 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."

[0428] In today's online communication environment, content posted by users often contains negative and malicious language. Such posts not only damage the reputation of companies and individuals, but also risk misleading other users. Furthermore, simply deleting negative posts may result in ignoring users' opinions and lowering their satisfaction. To solve these issues, it is necessary to properly analyze the content of posts and provide users with accurate and precise information.

[0429] 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 a means for receiving content posted by a user, a means for analyzing the received posted content using a natural language processing model, a means for determining whether the posted content is malicious based on the analysis result, a means for generating background information and supplemental information using generative artificial intelligence if the posted content is determined to be malicious, a means for creating an explanatory text based on the generated background information and supplemental information, and a means for adding the explanatory text to the original post and displaying it to the user. This allows the posted content to be appropriately analyzed, enabling the user to receive accurate and precise information.

[0430] A "User" is a person who accesses the Service or System and creates and inputs Posted Content.

[0431] "Posted content" refers to information such as messages, opinions, and feedback sent by users through the system.

[0432] The "receiving means" is a function or process by which the server acquires the posted content sent by the user.

[0433] A "natural language processing model" is an algorithm or software that enables a computer to understand and analyze human language.

[0434] The "analysis means" is a function for analyzing received posted content using a natural language processing model and understanding its content and sentiment.

[0435] The "determination means" is a system function for determining whether posted content is malicious based on the analysis results.

[0436] "Generative artificial intelligence" refers to algorithms or processes that use generative AI models to create background or supplemental information.

[0437] "Generation means" is a function that uses generative artificial intelligence to create background information and supplementary information in a form that is easy for users to understand.

[0438] A "description" is information that is created based on the generated background information and supplementary information and added to the original post.

[0439] The "display means" is a function for adding the generated explanatory text to the original post and displaying it to the user.

[0440] "Emotion recognition" is the process by which a natural language processing model or emotion engine analyzes and identifies emotions in posted content.

[0441] The present invention is a system that analyzes content posted by users, determines whether the content is malicious, and uses generative artificial intelligence to provide background and supplemental information. This system uses a combination of a natural language processing model and an emotion engine to deeply understand the posted content and generate appropriate feedback. Specific techniques for implementing the present invention are described in detail below.

[0442] Hardware and software used

[0443] This system uses the following hardware and software:

[0444] 1. Server: Use a high-performance server computer. You can use MySQL or PostgreSQL as the database server and Apache or Nginx as the web server.

[0445] 2. Terminal: The device through which the user accesses the system, including PCs, smartphones, tablets, etc.

[0446] 3. Natural Language Processing Models (NLP): Use advanced NLP models such as BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer).

[0447] 4. Generative AI: Use GPT-3 and other generative AI models to generate relevant context and supporting information.

[0448] Data processing and calculation

[0449] This system processes and calculates data in the following procedure.

[0450] 1. Receiving and storing your submissions:

[0451] The user uses the terminal to input the content to be posted and presses the send button.

[0452] The terminal sends the posted content to the server via an HTTP request.

[0453] The server stores the received post content in a database such as MySQL.

[0454] 2. Analysis using Natural Language Processing (NLP) models:

[0455] The server sends the received posted content to an NLP model for linguistic analysis.

[0456] The NLP model analyzes the context of the post and word choice and returns the analysis results.

[0457] 3. Emotion Recognition with Emotion Engine:

[0458] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions.

[0459] The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness.

[0460] 4. Malicious intent detection and AI generation:

[0461] The server determines whether the posted content is malicious based on the analysis results from the analysis engine.

[0462] If the content of a post is determined to be malicious, the server calls on generative artificial intelligence to generate background and supplementary information.

[0463] The generative AI generates background and supplementary information with appropriate tone and content based on the user's emotions recognized by the emotion engine.

[0464] 5. Creating and providing explanatory notes:

[0465] The server creates a description based on the generated background information.

[0466] The explanation is added to the original post and provided to the user.

[0467] The revised post and description are sent to the device, which displays it to the user.

[0468] Examples and prompts

[0469] For example, if a user posts, "The service is terrible! I can't connect at all!", the following processing will occur:

[0470] User comments: "The service is terrible! I can't connect at all!"

[0471] Prompt for generative AI model: "An angry user has posted a complaint about your service. Please respond in a calm tone with context."

[0472] The NLP model analyzes the post, and the emotion engine recognizes the emotion of "anger." The server invokes the generative AI to generate background information such as, "We have recently installed many new base stations and are working to improve the quality of our service." The server then creates a description based on the generated background information, and the revised post and description are displayed on the user's device.

[0473] In this way, the present invention allows users to receive information in a manner that is free from misunderstanding, thereby improving the quality of online communication.

[0474] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0475] Step 1:

[0476] The user uses the device to input posted content that represents their own opinion or feedback. The input posted content is the input in step 1. Specifically, the user uses the keyboard or touch screen to input "The service is so bad! I can't connect at all!" This posted content is stored in the text input field on the device.

[0477] Step 2:

[0478] When the user presses the send button, the entered post content is sent from the terminal to the server. The click event of the send button generates an HTTP request and the posted content is sent to the server. The input is the posted content entered by the user, and the output is the posted content received by the server.

[0479] Step 3:

[0480] The server receives the posted content sent by the user and temporarily stores it in a database. Specifically, it receives an HTTP request and inserts the content into a database such as MySQL or PostgreSQL. The input is the posted content sent, and the output is the posted content stored in the database.

[0481] Step 4:

[0482] The server sends the received post content to a natural language processing (NLP) model and begins analysis. Specifically, the server-side program sends an HTTP request to the NLP model via a REST API to obtain the analysis results. The input is the post content stored in the database, and the output is the analysis results returned by the NLP model.

[0483] Step 5:

[0484] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions. The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness. Specifically, the emotion engine receives the analysis results from the NLP model and performs emotion analysis. The input is the analysis results of the NLP model, and the output is the emotion recognition results from the emotion engine.

[0485] Step 6:

[0486] The server determines whether the posted content is malicious based on the analysis results from the analysis engine. If the posted content is determined to be malicious, it calls on a generative artificial intelligence (generative AI model) to generate background information and supplementary information. The generative AI generates background information and supplementary information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. The input is the emotion recognition result, and the output is the generated result from the generative AI model.

[0487] Step 7:

[0488] The server creates a description based on the generated background information. For example, using the background information obtained from the generative AI model, it creates a description such as, "We have recently installed many new base stations and are working to improve the quality of our service." The input is the result generated by the generative AI model, and the output is the created description.

[0489] Step 8:

[0490] The modified post and description are sent from the server to the user's device, which then displays them to the user. The server sends the modified post and description as an HTTP response. The input is the created description and the modified post content, and the output is what is displayed on the user's device. The user can check the displayed description and receive accurate information.

[0491] (Application example 2)

[0492] 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."

[0493] On modern online shopping sites, customer reviews are an important source of information for other customers. However, some customer reviews are emotional, overly negative, and contain factually incorrect information. This can lead to the spread of inaccurate information, misrepresenting product ratings, and damaging a company's reputation. To solve this problem, a system is needed that can appropriately address inaccurate or emotional posts and provide background information or explanatory notes.

[0494] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving content posted by a user; means for analyzing the received content using a natural language processing model; means for determining whether the content of the post is malicious based on the analysis result; means for generating background information and additional information using generative artificial intelligence if the content is determined to be malicious; means for creating an explanatory note based on the generated background information and additional information; means for adding the explanatory note to the original post and displaying it to the user; means for monitoring reviews on the online shopping site in real time and determining whether the content of reviews posted by users is appropriate; and means for generating background information and correction information for reviews determined to be inappropriate using generative artificial intelligence and presenting the explanatory note to the user. This makes it possible to prevent the spread of inaccurate information on the online shopping site and provide customers with accurate and reliable information.

[0495] The "means for receiving content posted by users" refers to a mechanism for transmitting text data entered by users via their electronic devices to a server and receiving it.

[0496] "Means for analyzing received posts using a natural language processing model" refers to a mechanism in which the server uses a computer program to analyze the text data it receives and understand the meaning and context of the text.

[0497] "Means for determining whether the content of a post is malicious based on the analysis results" refers to a mechanism for evaluating the analysis data obtained from a natural language processing model to determine whether the content of a post is offensive, insulting, or misleading.

[0498] "Means of using generative AI to generate background and additional information when a post is determined to be malicious" refers to a mechanism that uses generative AI to automatically create information to supplement the situation for posts that are detected to be malicious.

[0499] The "means for creating explanatory notes based on the generated background information and additional information" is a mechanism for organizing the generated supplementary information and creating a document to present it to the user in an easy-to-understand format.

[0500] The "means for adding an explanatory note to the original post and displaying it to the user" is a mechanism for associating the created explanatory note with the content of the original post and displaying it in a format that is easy for the user to view.

[0501] "Means for monitoring reviews on online shopping sites in real time and determining whether the content of reviews posted by users is appropriate" is a system for constantly observing customer reviews on online shopping sites and instantly evaluating the appropriateness of the posts.

[0502] "Means of using generative AI to generate background information and correction information for reviews that are judged to be inappropriate and presenting it to the user as an explanatory note" refers to a mechanism that, when an inappropriate review is detected, uses generative AI to create information and background explanations for correction and presents them to the user as an explanatory note.

[0503] The present invention is a system that analyzes the content of reviews posted by users through an integrated system via electronic devices in real time, generates appropriate background information and correction information as needed, and provides it to the user as explanatory notes.

[0504] First, the review content posted by the user is sent to the server. The server receives the text data using a means for receiving content posted by the user. Next, the received post content is analyzed using a natural language processing (NLP) model. Examples of models used here include spaCy and TextBlob. This analysis analyzes the context and sentiment of the review.

[0505] The server uses an emotion engine to determine whether the post is malicious based on the analysis results. The emotion engine analyzes keywords and expressions contained in the post to determine emotions such as anger, dissatisfaction, and sadness. This determines whether the post is appropriate, and only if it is not is it used to generate background information or corrections using generative artificial intelligence (AI).

[0506] The generative AI used is OpenAI's GPT-3. This generative AI uses user reviews and analysis results as prompts to generate appropriate background information and corrections. An example of a generative prompt is as follows:

[0507] User Review: The quality of this product is terrible!

[0508] Sentiment score: -0.8

[0509] Generate background information.

[0510] The server then uses the generated background and additional information to create explanatory notes. These notes contain information obtained from the generative AI and are presented in a format that is easy for the user to understand. For example, the notes might state, "To improve product quality, we are strengthening our inspection system and introducing a new quality control process."

[0511] Finally, the server adds explanatory notes to the original post and displays them to the user. This system prevents the spread of inaccurate information on online shopping sites and allows customers to obtain reliable information. The entire system operates in real time and aims to improve the quality of reviews on online shopping sites.

[0512] The above is a specific embodiment of the system according to the present invention.

[0513] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0514] Step 1:

[0515] Users post reviews using their electronic devices. The posted content is sent from the user's device to the server. The input is text data entered by the user, and the output is text data sent to the server.

[0516] Step 2:

[0517] The server receives the content posted by the user. The input is text data sent from the terminal, and the received data is temporarily stored in a database. The output is the stored text data.

[0518] Step 3:

[0519] The server analyzes the received post content using a natural language processing (NLP) model. In this step, the context and meaning of the post are analyzed. The input is the stored text data, and the output is the analysis result data. Specifically, the server analyzes the text using NLP tools such as spaCy or TextBlob to extract semantic information.

[0520] Step 4:

[0521] Based on the analysis results, the server determines whether the post content is malicious. The emotion engine analyzes keywords and emotion scores to determine whether the post is malicious. The input is the data obtained by NLP analysis, and the output is the judgment result. Specifically, the emotion analysis algorithm is used to calculate the emotion score of the text.

[0522] Step 5:

[0523] If a post is determined to be malicious, the server uses generative artificial intelligence (AI) to generate background information and additional information. The input is the judgment result and the original text data, and the output is the generated background information and additional information. Specifically, it uses OpenAI GPT-3 or similar to generate a prompt sentence like the following:

[0524] User Review: The quality of this product is terrible!

[0525] Sentiment score: -0.8

[0526] Generate background information.

[0527] The generative AI generates background and additional information based on this prompt.

[0528] Step 6:

[0529] The server creates explanatory notes based on the generated background information and additional information. The input is the generated background information and additional information, and the output is explanatory notes. Specifically, the server arranges the background information and additional information into a consistent format and compiles them into a note to explain to the user.

[0530] Step 7:

[0531] Finally, the server adds the explanatory note to the original post and displays it to the user. The input is the original post and the explanatory note, and the output is the corrected review displayed on the user's device. The user checks the explanatory note through their device and receives the correct information. Specifically, the corrected text data is displayed through the user interface.

[0532] 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.

[0533] 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.

[0534] 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.

[0535] [Third embodiment]

[0536] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0537] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0538] 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).

[0539] 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.

[0540] 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.

[0541] 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).

[0542] 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.

[0543] 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.

[0544] 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.

[0545] 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.

[0546] 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.

[0547] 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."

[0548] The present invention provides a system that analyzes posts sent by users, detects malicious content, and adds background information and correction information. The overall configuration of the system and each processing step are described in detail below.

[0549] First, the user inputs the content of the post through their terminal and sends it to the server. The server receives the content and temporarily stores it in preparation for the next step. At this stage, the user directly interacts with the system.

[0550] The server then sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context, word choice, and emotional expression of the post to determine whether the post is malicious. The results of this analysis are returned to the server.

[0551] If the server determines that a post is malicious based on the analysis results, it invokes a generative artificial intelligence (AI) to generate background information and corrections. The generative AI gathers relevant information from the internet and databases and automatically creates appropriate corrections and additional information for the post. This information may include specific fact-checking and statistical data.

[0552] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model will determine that the post is malicious. The generative AI will generate background information, such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0553] The server creates an "explanatory note" based on this generated background information. The explanatory note contains specific information obtained from the generative AI. The explanatory note is provided to the user as an attachment to the original post.

[0554] Finally, the corrected post and explanatory notes are sent to the device and displayed to the user, who can then review the displayed explanatory notes to obtain accurate information. Through this process, users can correct malicious posts and spread more accurate information.

[0555] The system of the present invention contributes to the prevention of inappropriate posts by users and the provision of accurate information. As a result, it is possible to improve the quality of online communication and protect the reputations of companies and individuals. This system gives users the opportunity to reconsider their posts and encourages them to learn how to avoid misleading information.

[0556] The processing flow will be explained below.

[0557] Step 1:

[0558] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0559] Step 2:

[0560] The server receives the posted content sent by the user and temporarily stores it in a database.

[0561] Step 3:

[0562] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the context and sentiment of the posts.

[0563] Step 4:

[0564] The analysis results from the natural language processing model are returned to the server, which then receives the analysis results and determines whether the post is malicious.

[0565] Step 5:

[0566] If a post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and corrections related to the post.

[0567] Step 6:

[0568] The background information and corrections generated by the generative AI are returned to the server, which then creates an "explanatory note" based on this information.

[0569] Step 7:

[0570] The server appends the generated explanatory note to the original post to create a modified post.

[0571] Step 8:

[0572] The server sends the modified post and explanatory notes to the device.

[0573] Step 9:

[0574] The terminal displays the received corrected post and explanatory notes to the user, who can then check the displayed explanatory notes to obtain accurate information.

[0575] This process allows the entire system to cooperate, curb malicious postings, and contribute to the provision of accurate information.

[0576] Example 1

[0577] 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."

[0578] In recent years, the amount of communication via user posts on the Internet has increased dramatically, but inappropriate posts and malicious information have become more prominent. Such posts not only damage the reputations of companies and individuals, but also risk causing misunderstandings and confusion. Therefore, there is a growing need for a system that can automatically detect malicious posts and add appropriate background information and corrections.

[0579] 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.

[0580] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and additional information using a generative artificial intelligence if the posted content is determined to be malicious, means for creating an explanatory note based on the generated background information and additional information, means for adding the explanatory note to the original post and displaying it to the user, and means for sending a prompt message to the generative artificial intelligence to generate appropriate background information and additional information. This makes it possible to automatically detect malicious posts and add appropriate information to present to the user.

[0581] A "user" is an operator of a terminal who uses the system to input and send posting content.

[0582] "Posted content" is text data that a user inputs via a terminal and sends to the server.

[0583] "Server" refers to a central computer system that receives, stores, analyzes, and provides generated information to users.

[0584] A "natural language processing model" is an algorithm or machine learning model for analyzing text data and understanding sentiment and context. Examples include high-performance models such as BERT and GPT-3.

[0585] The "analysis results" are the evaluation results of the posted content generated by the natural language processing model, and include information on whether the post is malicious or not.

[0586] "Generative AI" refers to algorithms or machine learning models that gather information from the internet or databases and generate necessary background information or corrections.

[0587] "Background information" is supplementary information related to the content of a post, and is information that complements the context and situation of the post.

[0588] "Additional information" is information used to correct or supplement the posted content, including accurate data and statistical information.

[0589] An "explanatory note" is a document created based on background information and additional information generated by generative artificial intelligence, and is intended to provide explanations and corrections to the user.

[0590] A "prompt sentence" is an instruction sentence that requests the generative artificial intelligence to generate background information or additional information.

[0591] The present invention provides a system for analyzing posts sent by users, detecting malicious content, and adding background information and correction information. Specific implementation methods of the system will be described in detail below.

[0592] First, the user enters the content of the post through their device and sends it to the server. The device can be a PC, smartphone, tablet, etc. The user enters the content of the post into the text box using the device's browser or a dedicated app, and presses the send button. For example, the user might enter, "The service is so bad! I can't connect at all!"

[0593] Next, the server receives the post sent by the user and temporarily stores it. The server is a high-performance computer, and the database is MySQL or PostgreSQL. The text data included in the body of the post request, "The service is terrible! I can't connect at all!", is stored.

[0594] The server then sends the received post content to a natural language processing (NLP) model for analysis. NLP models such as BERT and GPT-3 are used. The server calls the model's API and analyzes the post content. For example, a post such as "The service is so bad! I can't connect at all!" is judged to be "malicious."

[0595] Based on the analysis results, the server uses generative artificial intelligence (AI) to generate background and additional information. An example of a generative AI model is GPT-3. By sending specific prompts to this model, it automatically generates relevant information. An example of a prompt is, "When a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service."

[0596] The server then creates an "explanatory note" based on the generated background information and additional information. The explanatory note contains specific details obtained from the generative AI and is formatted as a text file or HTML. For example, the note might include information such as, "We have recently installed many new base stations and are working to improve the quality of our service."

[0597] Finally, the server sends the created explanatory note and the corrections to the original post to the user's device, and displays them in the browser or dedicated app. For example, the original post "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0598] This system allows users to check the background information and corrections of posted content and obtain accurate information. This will help prevent users from posting inappropriate content and provide accurate information, improving the quality of online communication. It will also help protect the reputations of companies and individuals and promote learning to avoid misunderstandings.

[0599] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0600] Step 1:

[0601] The user enters the content of the post and sends it. Using a device (PC or smartphone), the user enters the content of the post into a text box on a browser or dedicated app. For example, the user might enter "The service is so bad! I can't connect at all!" and click the "Send" button. The data entered is in text format, and the device sends this input data to the server as a post request.

[0602] Step 2:

[0603] The server receives the posted content and temporarily stores it. The server catches the incoming request and extracts the posted content. The received data is text data saying "The service is so bad! I can't connect at all!" The server temporarily stores this data in a database (e.g. MySQL, PostgreSQL). Here, data processing is performed, converting the format so that it can be stored appropriately in the database. The output is the temporarily stored posted data.

[0604] Step 3:

[0605] The server analyzes the post content using a natural language processing (NLP) model. The server retrieves the saved post content and sends it to an NLP model (e.g., BERT, GPT-3). The input includes the text "The service is so bad! It doesn't connect at all!" The model analyzes this text and infers whether the post is malicious based on sentiment and keywords. The output is an analysis result such as "Malicious." Specifically, the server calls the model's API and sends an analysis request in JSON format.

[0606] Step 4:

[0607] Based on the analysis results, the server requests the generative AI model to generate background information and correction information. The server receives the analysis result of "malicious" obtained from the NLP model and sends a prompt message to the generative AI model based on that result. The input sent is the prompt message, "If a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service." The AI ​​model generates background information in response, and the output is the information, "We have recently installed many new base stations and are working to improve the quality of our service."

[0608] Step 5:

[0609] The server creates an "explanatory note" based on the generated information. The server creates an "explanatory note" based on background information and additional information obtained from the generative AI model. The input includes the generated information (e.g., "We have recently installed many new base stations and are working to improve the quality of our service") and formats it in a format that is easy for the user to understand. The output is an explanatory note formatted as a text file or HTML, and specific operations include format conversion and the use of document generation tools.

[0610] Step 6:

[0611] The server sends the corrected post and explanatory note to the user's device. The server combines the created explanatory note with the corrections to the original post and sends it to the user's device. The input includes the corrected post and explanatory note. Specifically, the server generates a response and sends it to the browser or dedicated app. The output is the corrected post and explanatory note displayed on the user's device. For example, a post saying "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0612] Through these processing steps, users can see how their posts have been analyzed and what contextual information has been added, which is expected to improve the quality of online communication.

[0613] (Application example 1)

[0614] 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."

[0615] Malicious and inappropriate posts are frequently found on social media and message boards on the Internet. Such posts not only damage the reputations of individuals and companies, but also encourage the spread of misinformation and degrade the quality of online communication. A system is needed to detect malicious posts in real time and provide appropriate background information and corrections to clear up users' misunderstandings and prevent the spread of misinformation.

[0616] 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.

[0617] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and correction information using generative artificial intelligence if the posted content is determined to be malicious, means for generating a reminder based on the generated background information and correction information, means for adding the reminder to the original post and displaying it to the user, and means for analyzing the posted content in real time, thereby enabling the detection and correction of inappropriate posted content in real time.

[0618] "User" means an individual or organization that posts using the system.

[0619] "Postings" means text or messages that users send through the system.

[0620] A "natural language processing model" is a collection of algorithms and software that analyzes text data and understands its meaning and sentiment.

[0621] "Analysis results" refers to the information and data obtained after the natural language processing model analyzes the content of a post.

[0622] "Hateful posts" are inappropriate posts that contain negative emotions or intentions.

[0623] "Generative artificial intelligence" refers to an AI system that has the ability to collect appropriate background information and correction information from the internet or databases and generate new information.

[0624] "Background information" refers to factual and evidence-based explanations or information related to the content of a post.

[0625] "Correction information" means accurate information to correct a misunderstanding or error.

[0626] A "reminder" is a notice or statement provided to encourage the user to review or correct the generated background information or correction information.

[0627] "Analyzing in real time" means that the content is analyzed immediately after the user inputs and sends the content to be posted, and the corresponding processing is carried out.

[0628] An "explanatory note" is a document that contains a collection of generated background information and corrections.

[0629] "System" means the combination of equipment and software that performs a series of processes to receive, analyze, determine, and provide correction information for user posts.

[0630] To implement this invention, a series of processes involving a server, a terminal, and a user are required. Each step proceeds as follows:

[0631] First, when a user posts on social media or a message board, the user device sends the post to a server, which then temporarily stores the received post.

[0632] The server then sends the received content to a natural language processing (NLP) model to analyze it. The NLP model is designed to analyze the context, keywords, and sentiment of the text data to determine whether the content is malicious. This analysis is performed using libraries such as the spacy and transformers libraries.

[0633] Once the analysis results are returned, the server uses them to determine whether the post is malicious. If it is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and corrections for the post. The generative AI collects relevant information from external databases and the Internet to provide accurate and appropriate information. For example, if a user posts, "The service is so bad! I can't connect at all!", the AI ​​generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0634] Based on the generated background and correction information, the server generates a reminder and displays it to the user along with the original post, allowing the user to review their post and make corrections if necessary.

[0635] To achieve real-time post analysis, the server must process data quickly and efficiently. This requires a high-performance CPU, large memory capacity, and a high-speed network connection. In addition to the aforementioned natural language processing model and generative artificial intelligence, the server also uses a database management system.

[0636] For example, here's a prompt that might be used when a user posts, "The service is so bad! I can't get through at all!":

[0637] User Post: "The service is terrible! I can't connect at all!"

[0638] Generative AI prompt: "Please provide information about recent improvements in the quality of your telecommunications services."

[0639] This allows the generative AI to provide appropriate context and give users the opportunity to correct misinformation, improving the quality of online communication and clearing up misunderstandings.

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

[0641] Step 1: The user enters the content to post on a social media platform or bulletin board and sends it from the device to the server. The input is the user's post (text data), and the output is request data including the post content.

[0642] Step 2: The server temporarily saves the received post. The input is the request data containing the post, and the output is the save operation result in the session database. Specifically, the server stores the post in the database.

[0643] Step 3: The server sends the saved post content to a natural language processing model (NLP model) for analysis. The input is the saved post content, and the output is the analysis result. Specifically, the server calls the NLP model and performs text analysis on the post content.

[0644] Step 4: The NLP model analyzes the post's context, keywords, and sentiment and returns the results. The input is the text data of the post, and the output is the context analysis, keyword matching, and sentiment analysis results. Specifically, the natural language processing algorithm analyzes the text data and scores it for sentiment and keywords.

[0645] Step 5: Based on the analysis results, the server determines whether the post is malicious. The input is the analysis results, and the output is the judgment result (whether malicious or not). Specifically, the server performs a threshold judgment based on the sentiment analysis results and keyword scores.

[0646] Step 6: If the post is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and correction information. The input is the malicious post content and a prompt to the generative AI, and the output is background information and correction information. Specifically, the server sends the prompt to the AI ​​model and obtains relevant information.

[0647] Step 7: The generative AI generates background information and correction information and returns it to the server. The input is the prompt text, and the output is the generated information (background information and correction information). Specifically, the generative AI collects data from the internet and databases and generates appropriate information.

[0648] Step 8: The server generates a reminder based on the generated background information and correction information. The input is the information received from the generation AI, and the output is the reminder (explanatory note). Specifically, the server constructs an explanatory note from the generated information.

[0649] Step 9: Attach the reminder to the original post and display it to the user. The input is the reminder and the original post, and the output is the information displayed on the device. Specifically, the server integrates the reminder into the post and sends it to the user's device.

[0650] Step 10: The user checks the displayed reminder and reconsiders or modifies the post. The input is the displayed reminder and the original post, and the output is the new post by the user. Specifically, the user decides whether to repost or modify the post based on the displayed information. This process allows the user to reconsider the content of their post and, if necessary, modify it to the appropriate information.

[0651] 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.

[0652] This invention is a system that detects malicious posts by analyzing the content of posts entered by users and recognizing emotions. This system combines a natural language processing model with an emotion engine and also utilizes generative artificial intelligence to provide background information and corrections, allowing users to receive accurate information.

[0653] First, the user enters the content of the post through their device and presses the send button. The device then sends this content to the server. The server receives the content sent by the user and temporarily stores it in a database. Next, the server sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context and word choice of the post and passes the content to the system's analysis engine.

[0654] Furthermore, the system includes an emotion engine that recognizes users' emotions from the content of their posts. The emotion engine analyzes keywords and expressions contained in the posts to determine emotions such as anger, dissatisfaction, and sadness. This information is fed back to the analysis engine, which generates comprehensive analysis results.

[0655] The server receives the analysis results from the analysis engine and determines whether the post content is malicious. If the post content is determined to be malicious based on the results of the analysis, it calls on generative artificial intelligence (AI) to generate background information and correction information. The generative AI generates background information and correction information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. This allows information to be provided in a format that is more acceptable to the user.

[0656] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model analyzes the post and the emotion engine recognizes the emotion of anger. Based on this information, the generative AI generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service," and writes an explanatory note in a calm tone.

[0657] Next, the server creates an "explanatory note" based on the generated background information. The explanatory note contains specific information obtained from the generative AI and is provided to the user as an attachment to the original post. The corrected post and explanatory note are sent to the device, which displays them to the user. The user can then check the explanatory note and receive accurate information.

[0658] This invention provides users with an opportunity to reconsider their posts and encourages them to learn how to avoid misleading others. This system contributes to the suppression of inappropriate posts and the provision of accurate information, thereby improving the quality of online communication and protecting the reputations of companies and individuals.

[0659] The processing flow will be explained below.

[0660] Step 1:

[0661] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0662] Step 2:

[0663] The server receives the posted content sent by the user and temporarily stores it in a database.

[0664] Step 3:

[0665] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the post's context and word choice.

[0666] Step 4:

[0667] The server sends the analyzed post content to the emotion engine, which analyzes the keywords and expressions contained in the post content to determine the user's emotion (anger, dissatisfaction, sadness, etc.).

[0668] Step 5:

[0669] The server then integrates the analysis results again based on the user's emotional information recognized by the emotion engine, and receives the integrated analysis results to determine whether the posted content is malicious.

[0670] Step 6:

[0671] If the post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and additional information based on the post's content and emotional information.

[0672] Step 7:

[0673] The generative AI generates background and additional information, which is then returned to the server. The server then uses this information to create an "explanatory note," which is written in a tone that takes into account the user's emotions.

[0674] Step 8:

[0675] The server appends the generated explanatory note to the original post and creates the revised post, preparing it for delivery to the user.

[0676] Step 9:

[0677] The server transmits the corrected post and explanatory notes to the terminal, which displays the received corrected post and explanatory notes to the user.

[0678] Step 10:

[0679] The user checks the explanatory notes displayed on the terminal, and can receive and understand accurate information and appropriate background explanations from the explanatory notes.

[0680] This process allows users to review their posts and spread accurate information, and also helps curb inappropriate posts to protect the reputations of businesses and individuals.

[0681] Example 2

[0682] 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."

[0683] In today's online communication environment, content posted by users often contains negative and malicious language. Such posts not only damage the reputation of companies and individuals, but also risk misleading other users. Furthermore, simply deleting negative posts may result in ignoring users' opinions and lowering their satisfaction. To solve these issues, it is necessary to properly analyze the content of posts and provide users with accurate and precise information.

[0684] 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 a means for receiving content posted by a user, a means for analyzing the received posted content using a natural language processing model, a means for determining whether the posted content is malicious based on the analysis result, a means for generating background information and supplemental information using generative artificial intelligence if the posted content is determined to be malicious, a means for creating an explanatory text based on the generated background information and supplemental information, and a means for adding the explanatory text to the original post and displaying it to the user. This allows the posted content to be appropriately analyzed, enabling the user to receive accurate and precise information.

[0685] A "User" is a person who accesses the Service or System and creates and inputs Posted Content.

[0686] "Posted content" refers to information such as messages, opinions, and feedback sent by users through the system.

[0687] The "receiving means" is a function or process by which the server acquires the posted content sent by the user.

[0688] A "natural language processing model" is an algorithm or software that enables a computer to understand and analyze human language.

[0689] The "analysis means" is a function for analyzing received posted content using a natural language processing model and understanding its content and sentiment.

[0690] The "determination means" is a system function for determining whether posted content is malicious based on the analysis results.

[0691] "Generative artificial intelligence" refers to algorithms or processes that use generative AI models to create background or supplemental information.

[0692] "Generation means" is a function that uses generative artificial intelligence to create background information and supplementary information in a form that is easy for users to understand.

[0693] A "description" is information that is created based on the generated background information and supplementary information and added to the original post.

[0694] The "display means" is a function for adding the generated explanatory text to the original post and displaying it to the user.

[0695] "Emotion recognition" is the process by which a natural language processing model or emotion engine analyzes and identifies emotions in posted content.

[0696] The present invention is a system that analyzes content posted by users, determines whether the content is malicious, and uses generative artificial intelligence to provide background and supplemental information. This system uses a combination of a natural language processing model and an emotion engine to deeply understand the posted content and generate appropriate feedback. Specific techniques for implementing the present invention are described in detail below.

[0697] Hardware and software used

[0698] This system uses the following hardware and software:

[0699] 1. Server: Use a high-performance server computer. You can use MySQL or PostgreSQL as the database server and Apache or Nginx as the web server.

[0700] 2. Terminal: The device through which the user accesses the system, including PCs, smartphones, tablets, etc.

[0701] 3. Natural Language Processing Models (NLP): Use advanced NLP models such as BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer).

[0702] 4. Generative AI: Use GPT-3 and other generative AI models to generate relevant context and supporting information.

[0703] Data processing and calculation

[0704] This system processes and calculates data in the following procedure.

[0705] 1. Receiving and storing your submissions:

[0706] The user uses the terminal to input the content to be posted and presses the send button.

[0707] The terminal sends the posted content to the server via an HTTP request.

[0708] The server stores the received post content in a database such as MySQL.

[0709] 2. Analysis using Natural Language Processing (NLP) models:

[0710] The server sends the received posted content to an NLP model for linguistic analysis.

[0711] The NLP model analyzes the context of the post and word choice and returns the analysis results.

[0712] 3. Emotion Recognition with Emotion Engine:

[0713] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions.

[0714] The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness.

[0715] 4. Malicious intent detection and AI generation:

[0716] The server determines whether the posted content is malicious based on the analysis results from the analysis engine.

[0717] If the content of a post is determined to be malicious, the server calls on generative artificial intelligence to generate background and supplementary information.

[0718] The generative AI generates background and supplementary information with appropriate tone and content based on the user's emotions recognized by the emotion engine.

[0719] 5. Creating and providing explanatory notes:

[0720] The server creates a description based on the generated background information.

[0721] The explanation is added to the original post and provided to the user.

[0722] The revised post and description are sent to the device, which displays it to the user.

[0723] Examples and prompts

[0724] For example, if a user posts, "The service is terrible! I can't connect at all!", the following processing will occur:

[0725] User comments: "The service is terrible! I can't connect at all!"

[0726] Prompt for generative AI model: "An angry user has posted a complaint about your service. Please respond in a calm tone with context."

[0727] The NLP model analyzes the post, and the emotion engine recognizes the emotion of "anger." The server invokes the generative AI to generate background information such as, "We have recently installed many new base stations and are working to improve the quality of our service." The server then creates a description based on the generated background information, and the revised post and description are displayed on the user's device.

[0728] In this way, the present invention allows users to receive information in a manner that is free from misunderstanding, thereby improving the quality of online communication.

[0729] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0730] Step 1:

[0731] The user uses the device to input posted content that represents their own opinion or feedback. The input posted content is the input in step 1. Specifically, the user uses the keyboard or touch screen to input "The service is so bad! I can't connect at all!" This posted content is stored in the text input field on the device.

[0732] Step 2:

[0733] When the user presses the send button, the entered post content is sent from the terminal to the server. The click event of the send button generates an HTTP request and the posted content is sent to the server. The input is the posted content entered by the user, and the output is the posted content received by the server.

[0734] Step 3:

[0735] The server receives the posted content sent by the user and temporarily stores it in a database. Specifically, it receives an HTTP request and inserts the content into a database such as MySQL or PostgreSQL. The input is the posted content sent, and the output is the posted content stored in the database.

[0736] Step 4:

[0737] The server sends the received post content to a natural language processing (NLP) model and begins analysis. Specifically, the server-side program sends an HTTP request to the NLP model via a REST API to obtain the analysis results. The input is the post content stored in the database, and the output is the analysis results returned by the NLP model.

[0738] Step 5:

[0739] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions. The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness. Specifically, the emotion engine receives the analysis results from the NLP model and performs emotion analysis. The input is the analysis results of the NLP model, and the output is the emotion recognition results from the emotion engine.

[0740] Step 6:

[0741] The server determines whether the posted content is malicious based on the analysis results from the analysis engine. If the posted content is determined to be malicious, it calls on a generative artificial intelligence (generative AI model) to generate background information and supplementary information. The generative AI generates background information and supplementary information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. The input is the emotion recognition result, and the output is the generated result from the generative AI model.

[0742] Step 7:

[0743] The server creates a description based on the generated background information. For example, using the background information obtained from the generative AI model, it creates a description such as, "We have recently installed many new base stations and are working to improve the quality of our service." The input is the result generated by the generative AI model, and the output is the created description.

[0744] Step 8:

[0745] The modified post and description are sent from the server to the user's device, which then displays them to the user. The server sends the modified post and description as an HTTP response. The input is the created description and the modified post content, and the output is what is displayed on the user's device. The user can check the displayed description and receive accurate information.

[0746] (Application example 2)

[0747] 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."

[0748] On modern online shopping sites, customer reviews are an important source of information for other customers. However, some customer reviews are emotional, overly negative, and contain factually incorrect information. This can lead to the spread of inaccurate information, misrepresenting product ratings, and damaging a company's reputation. To solve this problem, a system is needed that can appropriately address inaccurate or emotional posts and provide background information or explanatory notes.

[0749] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving content posted by a user; means for analyzing the received content using a natural language processing model; means for determining whether the content of the post is malicious based on the analysis result; means for generating background information and additional information using generative artificial intelligence if the content is determined to be malicious; means for creating an explanatory note based on the generated background information and additional information; means for adding the explanatory note to the original post and displaying it to the user; means for monitoring reviews on the online shopping site in real time and determining whether the content of reviews posted by users is appropriate; and means for generating background information and correction information for reviews determined to be inappropriate using generative artificial intelligence and presenting the explanatory note to the user. This makes it possible to prevent the spread of inaccurate information on the online shopping site and provide customers with accurate and reliable information.

[0750] The "means for receiving content posted by users" refers to a mechanism for transmitting text data entered by users via their electronic devices to a server and receiving it.

[0751] "Means for analyzing received posts using a natural language processing model" refers to a mechanism in which the server uses a computer program to analyze the text data it receives and understand the meaning and context of the text.

[0752] "Means for determining whether the content of a post is malicious based on the analysis results" refers to a mechanism for evaluating the analysis data obtained from a natural language processing model to determine whether the content of a post is offensive, insulting, or misleading.

[0753] "Means of using generative AI to generate background and additional information when a post is determined to be malicious" refers to a mechanism that uses generative AI to automatically create information to supplement the situation for posts that are detected to be malicious.

[0754] The "means for creating explanatory notes based on the generated background information and additional information" is a mechanism for organizing the generated supplementary information and creating a document to present it to the user in an easy-to-understand format.

[0755] The "means for adding an explanatory note to the original post and displaying it to the user" is a mechanism for associating the created explanatory note with the content of the original post and displaying it in a format that is easy for the user to view.

[0756] "Means for monitoring reviews on online shopping sites in real time and determining whether the content of reviews posted by users is appropriate" is a system for constantly observing customer reviews on online shopping sites and instantly evaluating the appropriateness of the posts.

[0757] "Means of using generative AI to generate background information and correction information for reviews that are judged to be inappropriate and presenting it to the user as an explanatory note" refers to a mechanism that, when an inappropriate review is detected, uses generative AI to create information and background explanations for correction and presents them to the user as an explanatory note.

[0758] The present invention is a system that analyzes the content of reviews posted by users through an integrated system via electronic devices in real time, generates appropriate background information and correction information as needed, and provides it to the user as explanatory notes.

[0759] First, the review content posted by the user is sent to the server. The server receives the text data using a means for receiving content posted by the user. Next, the received post content is analyzed using a natural language processing (NLP) model. Examples of models used here include spaCy and TextBlob. This analysis analyzes the context and sentiment of the review.

[0760] The server uses an emotion engine to determine whether the post is malicious based on the analysis results. The emotion engine analyzes keywords and expressions contained in the post to determine emotions such as anger, dissatisfaction, and sadness. This determines whether the post is appropriate, and only if it is not is it used to generate background information or corrections using generative artificial intelligence (AI).

[0761] The generative AI used is OpenAI's GPT-3. This generative AI uses user reviews and analysis results as prompts to generate appropriate background information and corrections. An example of a generative prompt is as follows:

[0762] User Review: The quality of this product is terrible!

[0763] Sentiment score: -0.8

[0764] Generate background information.

[0765] The server then uses the generated background and additional information to create explanatory notes. These notes contain information obtained from the generative AI and are presented in a format that is easy for the user to understand. For example, the notes might state, "To improve product quality, we are strengthening our inspection system and introducing a new quality control process."

[0766] Finally, the server adds explanatory notes to the original post and displays them to the user. This system prevents the spread of inaccurate information on online shopping sites and allows customers to obtain reliable information. The entire system operates in real time and aims to improve the quality of reviews on online shopping sites.

[0767] The above is a specific embodiment of the system according to the present invention.

[0768] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0769] Step 1:

[0770] Users post reviews using their electronic devices. The posted content is sent from the user's device to the server. The input is text data entered by the user, and the output is text data sent to the server.

[0771] Step 2:

[0772] The server receives the content posted by the user. The input is text data sent from the terminal, and the received data is temporarily stored in a database. The output is the stored text data.

[0773] Step 3:

[0774] The server analyzes the received post content using a natural language processing (NLP) model. In this step, the context and meaning of the post are analyzed. The input is the stored text data, and the output is the analysis result data. Specifically, the server analyzes the text using NLP tools such as spaCy or TextBlob to extract semantic information.

[0775] Step 4:

[0776] Based on the analysis results, the server determines whether the post content is malicious. The emotion engine analyzes keywords and emotion scores to determine whether the post is malicious. The input is the data obtained by NLP analysis, and the output is the judgment result. Specifically, the emotion analysis algorithm is used to calculate the emotion score of the text.

[0777] Step 5:

[0778] If a post is determined to be malicious, the server uses generative artificial intelligence (AI) to generate background information and additional information. The input is the judgment result and the original text data, and the output is the generated background information and additional information. Specifically, it uses OpenAI GPT-3 or similar to generate a prompt sentence like the following:

[0779] User Review: The quality of this product is terrible!

[0780] Sentiment score: -0.8

[0781] Generate background information.

[0782] The generative AI generates background and additional information based on this prompt.

[0783] Step 6:

[0784] The server creates explanatory notes based on the generated background information and additional information. The input is the generated background information and additional information, and the output is explanatory notes. Specifically, the server arranges the background information and additional information into a consistent format and compiles them into a note to explain to the user.

[0785] Step 7:

[0786] Finally, the server adds the explanatory note to the original post and displays it to the user. The input is the original post and the explanatory note, and the output is the corrected review displayed on the user's device. The user checks the explanatory note through their device and receives the correct information. Specifically, the corrected text data is displayed through the user interface.

[0787] 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.

[0788] 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.

[0789] 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.

[0790] [Fourth embodiment]

[0791] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0792] 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.

[0793] 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).

[0794] 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.

[0795] 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.

[0796] 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).

[0797] 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.

[0798] 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.

[0799] 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.

[0800] 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.

[0801] 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.

[0802] 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.

[0803] 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."

[0804] The present invention provides a system that analyzes posts sent by users, detects malicious content, and adds background information and correction information. The overall configuration of the system and each processing step are described in detail below.

[0805] First, the user inputs the content of the post through their terminal and sends it to the server. The server receives the content and temporarily stores it in preparation for the next step. At this stage, the user directly interacts with the system.

[0806] The server then sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context, word choice, and emotional expression of the post to determine whether the post is malicious. The results of this analysis are returned to the server.

[0807] If the server determines that a post is malicious based on the analysis results, it invokes a generative artificial intelligence (AI) to generate background information and corrections. The generative AI gathers relevant information from the internet and databases and automatically creates appropriate corrections and additional information for the post. This information may include specific fact-checking and statistical data.

[0808] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model will determine that the post is malicious. The generative AI will generate background information, such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0809] The server creates an "explanatory note" based on this generated background information. The explanatory note contains specific information obtained from the generative AI. The explanatory note is provided to the user as an attachment to the original post.

[0810] Finally, the corrected post and explanatory notes are sent to the device and displayed to the user, who can then review the displayed explanatory notes to obtain accurate information. Through this process, users can correct malicious posts and spread more accurate information.

[0811] The system of the present invention contributes to the prevention of inappropriate posts by users and the provision of accurate information. As a result, it is possible to improve the quality of online communication and protect the reputations of companies and individuals. This system gives users the opportunity to reconsider their posts and encourages them to learn how to avoid misleading information.

[0812] The processing flow will be explained below.

[0813] Step 1:

[0814] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0815] Step 2:

[0816] The server receives the posted content sent by the user and temporarily stores it in a database.

[0817] Step 3:

[0818] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the context and sentiment of the posts.

[0819] Step 4:

[0820] The analysis results from the natural language processing model are returned to the server, which then receives the analysis results and determines whether the post is malicious.

[0821] Step 5:

[0822] If a post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and corrections related to the post.

[0823] Step 6:

[0824] The background information and corrections generated by the generative AI are returned to the server, which then creates an "explanatory note" based on this information.

[0825] Step 7:

[0826] The server appends the generated explanatory note to the original post to create a modified post.

[0827] Step 8:

[0828] The server sends the modified post and explanatory notes to the device.

[0829] Step 9:

[0830] The terminal displays the received corrected post and explanatory notes to the user, who can then check the displayed explanatory notes to obtain accurate information.

[0831] This process allows the entire system to cooperate, curb malicious postings, and contribute to the provision of accurate information.

[0832] Example 1

[0833] 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."

[0834] In recent years, the amount of communication via user posts on the Internet has increased dramatically, but inappropriate posts and malicious information have become more prominent. Such posts not only damage the reputations of companies and individuals, but also risk causing misunderstandings and confusion. Therefore, there is a growing need for a system that can automatically detect malicious posts and add appropriate background information and corrections.

[0835] 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.

[0836] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and additional information using a generative artificial intelligence if the posted content is determined to be malicious, means for creating an explanatory note based on the generated background information and additional information, means for adding the explanatory note to the original post and displaying it to the user, and means for sending a prompt message to the generative artificial intelligence to generate appropriate background information and additional information. This makes it possible to automatically detect malicious posts and add appropriate information to present to the user.

[0837] A "user" is an operator of a terminal who uses the system to input and send posting content.

[0838] "Posted content" is text data that a user inputs via a terminal and sends to the server.

[0839] "Server" refers to a central computer system that receives, stores, analyzes, and provides generated information to users.

[0840] A "natural language processing model" is an algorithm or machine learning model for analyzing text data and understanding sentiment and context. Examples include high-performance models such as BERT and GPT-3.

[0841] The "analysis results" are the evaluation results of the posted content generated by the natural language processing model, and include information on whether the post is malicious or not.

[0842] "Generative AI" refers to algorithms or machine learning models that gather information from the internet or databases and generate necessary background information or corrections.

[0843] "Background information" is supplementary information related to the content of a post, and is information that complements the context and situation of the post.

[0844] "Additional information" is information used to correct or supplement the posted content, including accurate data and statistical information.

[0845] An "explanatory note" is a document created based on background information and additional information generated by generative artificial intelligence, and is intended to provide explanations and corrections to the user.

[0846] A "prompt sentence" is an instruction sentence that requests the generative artificial intelligence to generate background information or additional information.

[0847] The present invention provides a system for analyzing posts sent by users, detecting malicious content, and adding background information and correction information. Specific implementation methods of the system will be described in detail below.

[0848] First, the user enters the content of the post through their device and sends it to the server. The device can be a PC, smartphone, tablet, etc. The user enters the content of the post into the text box using the device's browser or a dedicated app, and presses the send button. For example, the user might enter, "The service is so bad! I can't connect at all!"

[0849] Next, the server receives the post sent by the user and temporarily stores it. The server is a high-performance computer, and the database is MySQL or PostgreSQL. The text data included in the body of the post request, "The service is terrible! I can't connect at all!", is stored.

[0850] The server then sends the received post content to a natural language processing (NLP) model for analysis. NLP models such as BERT and GPT-3 are used. The server calls the model's API and analyzes the post content. For example, a post such as "The service is so bad! I can't connect at all!" is judged to be "malicious."

[0851] Based on the analysis results, the server uses generative artificial intelligence (AI) to generate background and additional information. An example of a generative AI model is GPT-3. By sending specific prompts to this model, it automatically generates relevant information. An example of a prompt is, "When a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service."

[0852] The server then creates an "explanatory note" based on the generated background information and additional information. The explanatory note contains specific details obtained from the generative AI and is formatted as a text file or HTML. For example, the note might include information such as, "We have recently installed many new base stations and are working to improve the quality of our service."

[0853] Finally, the server sends the created explanatory note and the corrections to the original post to the user's device, and displays them in the browser or dedicated app. For example, the original post "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0854] This system allows users to check the background information and corrections of posted content and obtain accurate information. This will help prevent users from posting inappropriate content and provide accurate information, improving the quality of online communication. It will also help protect the reputations of companies and individuals and promote learning to avoid misunderstandings.

[0855] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0856] Step 1:

[0857] The user enters the content of the post and sends it. Using a device (PC or smartphone), the user enters the content of the post into a text box on a browser or dedicated app. For example, the user might enter "The service is so bad! I can't connect at all!" and click the "Send" button. The data entered is in text format, and the device sends this input data to the server as a post request.

[0858] Step 2:

[0859] The server receives the posted content and temporarily stores it. The server catches the incoming request and extracts the posted content. The received data is text data saying "The service is so bad! I can't connect at all!" The server temporarily stores this data in a database (e.g. MySQL, PostgreSQL). Here, data processing is performed, converting the format so that it can be stored appropriately in the database. The output is the temporarily stored posted data.

[0860] Step 3:

[0861] The server analyzes the post content using a natural language processing (NLP) model. The server retrieves the saved post content and sends it to an NLP model (e.g., BERT, GPT-3). The input includes the text "The service is so bad! It doesn't connect at all!" The model analyzes this text and infers whether the post is malicious based on sentiment and keywords. The output is an analysis result such as "Malicious." Specifically, the server calls the model's API and sends an analysis request in JSON format.

[0862] Step 4:

[0863] Based on the analysis results, the server requests the generative AI model to generate background information and correction information. The server receives the analysis result of "malicious" obtained from the NLP model and sends a prompt message to the generative AI model based on that result. The input sent is the prompt message, "If a user posts, 'The service is so bad! I can't connect at all!' please generate specific information about our efforts to improve our service." The AI ​​model generates background information in response, and the output is the information, "We have recently installed many new base stations and are working to improve the quality of our service."

[0864] Step 5:

[0865] The server creates an "explanatory note" based on the generated information. The server creates an "explanatory note" based on background information and additional information obtained from the generative AI model. The input includes the generated information (e.g., "We have recently installed many new base stations and are working to improve the quality of our service") and formats it in a format that is easy for the user to understand. The output is an explanatory note formatted as a text file or HTML, and specific operations include format conversion and the use of document generation tools.

[0866] Step 6:

[0867] The server sends the corrected post and explanatory note to the user's device. The server combines the created explanatory note with the corrections to the original post and sends it to the user's device. The input includes the corrected post and explanatory note. Specifically, the server generates a response and sends it to the browser or dedicated app. The output is the corrected post and explanatory note displayed on the user's device. For example, a post saying "The service is so bad! I can't connect at all!" is displayed together with an explanatory note saying "We've recently installed many new base stations and are working to improve the quality of our service."

[0868] Through these processing steps, users can see how their posts have been analyzed and what contextual information has been added, which is expected to improve the quality of online communication.

[0869] (Application example 1)

[0870] 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."

[0871] Malicious and inappropriate posts are frequently found on social media and message boards on the Internet. Such posts not only damage the reputations of individuals and companies, but also encourage the spread of misinformation and degrade the quality of online communication. A system is needed to detect malicious posts in real time and provide appropriate background information and corrections to clear up users' misunderstandings and prevent the spread of misinformation.

[0872] 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.

[0873] In this invention, the server includes means for receiving content posted by users, means for analyzing the received posted content using a natural language processing model, means for determining whether the posted content is malicious based on the analysis result, means for generating background information and correction information using generative artificial intelligence if the posted content is determined to be malicious, means for generating a reminder based on the generated background information and correction information, means for adding the reminder to the original post and displaying it to the user, and means for analyzing the posted content in real time, thereby enabling the detection and correction of inappropriate posted content in real time.

[0874] "User" means an individual or organization that posts using the system.

[0875] "Postings" means text or messages that users send through the system.

[0876] A "natural language processing model" is a collection of algorithms and software that analyzes text data and understands its meaning and sentiment.

[0877] "Analysis results" refers to the information and data obtained after the natural language processing model analyzes the content of a post.

[0878] "Hateful posts" are inappropriate posts that contain negative emotions or intentions.

[0879] "Generative artificial intelligence" refers to an AI system that has the ability to collect appropriate background information and correction information from the internet or databases and generate new information.

[0880] "Background information" refers to factual and evidence-based explanations or information related to the content of a post.

[0881] "Correction information" means accurate information to correct a misunderstanding or error.

[0882] A "reminder" is a notice or statement provided to encourage the user to review or correct the generated background information or correction information.

[0883] "Analyzing in real time" means that the content is analyzed immediately after the user inputs and sends the content to be posted, and the corresponding processing is carried out.

[0884] An "explanatory note" is a document that contains a collection of generated background information and corrections.

[0885] "System" means the combination of equipment and software that performs a series of processes to receive, analyze, determine, and provide correction information for user posts.

[0886] To implement this invention, a series of processes involving a server, a terminal, and a user are required. Each step proceeds as follows:

[0887] First, when a user posts on social media or a message board, the user device sends the post to a server, which then temporarily stores the received post.

[0888] The server then sends the received content to a natural language processing (NLP) model to analyze it. The NLP model is designed to analyze the context, keywords, and sentiment of the text data to determine whether the content is malicious. This analysis is performed using libraries such as the spacy and transformers libraries.

[0889] Once the analysis results are returned, the server uses them to determine whether the post is malicious. If it is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and corrections for the post. The generative AI collects relevant information from external databases and the Internet to provide accurate and appropriate information. For example, if a user posts, "The service is so bad! I can't connect at all!", the AI ​​generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service."

[0890] Based on the generated background and correction information, the server generates a reminder and displays it to the user along with the original post, allowing the user to review their post and make corrections if necessary.

[0891] To achieve real-time post analysis, the server must process data quickly and efficiently. This requires a high-performance CPU, large memory capacity, and a high-speed network connection. In addition to the aforementioned natural language processing model and generative artificial intelligence, the server also uses a database management system.

[0892] For example, here's a prompt that might be used when a user posts, "The service is so bad! I can't get through at all!":

[0893] User Post: "The service is terrible! I can't connect at all!"

[0894] Generative AI prompt: "Please provide information about recent improvements in the quality of your telecommunications services."

[0895] This allows the generative AI to provide appropriate context and give users the opportunity to correct misinformation, improving the quality of online communication and clearing up misunderstandings.

[0896] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0897] Step 1: The user enters the content to post on a social media platform or bulletin board and sends it from the device to the server. The input is the user's post (text data), and the output is request data including the post content.

[0898] Step 2: The server temporarily saves the received post. The input is the request data containing the post, and the output is the save operation result in the session database. Specifically, the server stores the post in the database.

[0899] Step 3: The server sends the saved post content to a natural language processing model (NLP model) for analysis. The input is the saved post content, and the output is the analysis result. Specifically, the server calls the NLP model and performs text analysis on the post content.

[0900] Step 4: The NLP model analyzes the post's context, keywords, and sentiment and returns the results. The input is the text data of the post, and the output is the context analysis, keyword matching, and sentiment analysis results. Specifically, the natural language processing algorithm analyzes the text data and scores it for sentiment and keywords.

[0901] Step 5: Based on the analysis results, the server determines whether the post is malicious. The input is the analysis results, and the output is the judgment result (whether malicious or not). Specifically, the server performs a threshold judgment based on the sentiment analysis results and keyword scores.

[0902] Step 6: If the post is determined to be malicious, the server sends a request to a generative artificial intelligence (AI) to generate background information and correction information. The input is the malicious post content and a prompt to the generative AI, and the output is background information and correction information. Specifically, the server sends the prompt to the AI ​​model and obtains relevant information.

[0903] Step 7: The generative AI generates background information and correction information and returns it to the server. The input is the prompt text, and the output is the generated information (background information and correction information). Specifically, the generative AI collects data from the internet and databases and generates appropriate information.

[0904] Step 8: The server generates a reminder based on the generated background information and correction information. The input is the information received from the generation AI, and the output is the reminder (explanatory note). Specifically, the server constructs an explanatory note from the generated information.

[0905] Step 9: Attach the reminder to the original post and display it to the user. The input is the reminder and the original post, and the output is the information displayed on the device. Specifically, the server integrates the reminder into the post and sends it to the user's device.

[0906] Step 10: The user checks the displayed reminder and reconsiders or modifies the post. The input is the displayed reminder and the original post, and the output is the new post by the user. Specifically, the user decides whether to repost or modify the post based on the displayed information. This process allows the user to reconsider the content of their post and, if necessary, modify it to the appropriate information.

[0907] 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.

[0908] This invention is a system that detects malicious posts by analyzing the content of posts entered by users and recognizing emotions. This system combines a natural language processing model with an emotion engine and also utilizes generative artificial intelligence to provide background information and corrections, allowing users to receive accurate information.

[0909] First, the user enters the content of the post through their device and presses the send button. The device then sends this content to the server. The server receives the content sent by the user and temporarily stores it in a database. Next, the server sends the received content to a natural language processing (NLP) model for analysis. The NLP model analyzes the context and word choice of the post and passes the content to the system's analysis engine.

[0910] Furthermore, the system includes an emotion engine that recognizes users' emotions from the content of their posts. The emotion engine analyzes keywords and expressions contained in the posts to determine emotions such as anger, dissatisfaction, and sadness. This information is fed back to the analysis engine, which generates comprehensive analysis results.

[0911] The server receives the analysis results from the analysis engine and determines whether the post content is malicious. If the post content is determined to be malicious based on the results of the analysis, it calls on generative artificial intelligence (AI) to generate background information and correction information. The generative AI generates background information and correction information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. This allows information to be provided in a format that is more acceptable to the user.

[0912] For example, if a user posts, "The service is so bad! I can't connect at all!", the NLP model analyzes the post and the emotion engine recognizes the emotion of anger. Based on this information, the generative AI generates background information such as, "We've recently installed many new base stations and are working to improve the quality of our service," and writes an explanatory note in a calm tone.

[0913] Next, the server creates an "explanatory note" based on the generated background information. The explanatory note contains specific information obtained from the generative AI and is provided to the user as an attachment to the original post. The corrected post and explanatory note are sent to the device, which displays them to the user. The user can then check the explanatory note and receive accurate information.

[0914] This invention provides users with an opportunity to reconsider their posts and encourages them to learn how to avoid misleading others. This system contributes to the suppression of inappropriate posts and the provision of accurate information, thereby improving the quality of online communication and protecting the reputations of companies and individuals.

[0915] The processing flow will be explained below.

[0916] Step 1:

[0917] The user uses the terminal to input the content of the post and presses the send button, which then sends the content to the server.

[0918] Step 2:

[0919] The server receives the posted content sent by the user and temporarily stores it in a database.

[0920] Step 3:

[0921] The server sends the received posts to a natural language processing (NLP) model, which performs morphological and sentiment analysis to analyze the post's context and word choice.

[0922] Step 4:

[0923] The server sends the analyzed post content to the emotion engine, which analyzes the keywords and expressions contained in the post content to determine the user's emotion (anger, dissatisfaction, sadness, etc.).

[0924] Step 5:

[0925] The server then integrates the analysis results again based on the user's emotional information recognized by the emotion engine, and receives the integrated analysis results to determine whether the posted content is malicious.

[0926] Step 6:

[0927] If the post is deemed malicious, the server invokes a generative artificial intelligence (AI) that generates context and additional information based on the post's content and emotional information.

[0928] Step 7:

[0929] The generative AI generates background and additional information, which is then returned to the server. The server then uses this information to create an "explanatory note," which is written in a tone that takes into account the user's emotions.

[0930] Step 8:

[0931] The server appends the generated explanatory note to the original post and creates the revised post, preparing it for delivery to the user.

[0932] Step 9:

[0933] The server transmits the corrected post and explanatory notes to the terminal, which displays the received corrected post and explanatory notes to the user.

[0934] Step 10:

[0935] The user checks the explanatory notes displayed on the terminal, and can receive and understand accurate information and appropriate background explanations from the explanatory notes.

[0936] This process allows users to review their posts and spread accurate information, and also helps curb inappropriate posts to protect the reputations of businesses and individuals.

[0937] Example 2

[0938] 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."

[0939] In today's online communication environment, content posted by users often contains negative and malicious language. Such posts not only damage the reputation of companies and individuals, but also risk misleading other users. Furthermore, simply deleting negative posts may result in ignoring users' opinions and lowering their satisfaction. To solve these issues, it is necessary to properly analyze the content of posts and provide users with accurate and precise information.

[0940] 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 a means for receiving content posted by a user, a means for analyzing the received posted content using a natural language processing model, a means for determining whether the posted content is malicious based on the analysis result, a means for generating background information and supplemental information using generative artificial intelligence if the posted content is determined to be malicious, a means for creating an explanatory text based on the generated background information and supplemental information, and a means for adding the explanatory text to the original post and displaying it to the user. This allows the posted content to be appropriately analyzed, enabling the user to receive accurate and precise information.

[0941] A "User" is a person who accesses the Service or System and creates and inputs Posted Content.

[0942] "Posted content" refers to information such as messages, opinions, and feedback sent by users through the system.

[0943] The "receiving means" is a function or process by which the server acquires the posted content sent by the user.

[0944] A "natural language processing model" is an algorithm or software that enables a computer to understand and analyze human language.

[0945] The "analysis means" is a function for analyzing received posted content using a natural language processing model and understanding its content and sentiment.

[0946] The "determination means" is a system function for determining whether posted content is malicious based on the analysis results.

[0947] "Generative artificial intelligence" refers to algorithms or processes that use generative AI models to create background or supplemental information.

[0948] "Generation means" is a function that uses generative artificial intelligence to create background information and supplementary information in a form that is easy for users to understand.

[0949] A "description" is information that is created based on the generated background information and supplementary information and added to the original post.

[0950] The "display means" is a function for adding the generated explanatory text to the original post and displaying it to the user.

[0951] "Emotion recognition" is the process by which a natural language processing model or emotion engine analyzes and identifies emotions in posted content.

[0952] The present invention is a system that analyzes content posted by users, determines whether the content is malicious, and uses generative artificial intelligence to provide background and supplemental information. This system uses a combination of a natural language processing model and an emotion engine to deeply understand the posted content and generate appropriate feedback. Specific techniques for implementing the present invention are described in detail below.

[0953] Hardware and software used

[0954] This system uses the following hardware and software:

[0955] 1. Server: Use a high-performance server computer. You can use MySQL or PostgreSQL as the database server and Apache or Nginx as the web server.

[0956] 2. Terminal: The device through which the user accesses the system, including PCs, smartphones, tablets, etc.

[0957] 3. Natural Language Processing Models (NLP): Use advanced NLP models such as BERT (Bidirectional Encoder Representations from Transformers) and GPT (Generative Pre-trained Transformer).

[0958] 4. Generative AI: Use GPT-3 and other generative AI models to generate relevant context and supporting information.

[0959] Data processing and calculation

[0960] This system processes and calculates data in the following procedure.

[0961] 1. Receiving and storing your submissions:

[0962] The user uses the terminal to input the content to be posted and presses the send button.

[0963] The terminal sends the posted content to the server via an HTTP request.

[0964] The server stores the received post content in a database such as MySQL.

[0965] 2. Analysis using Natural Language Processing (NLP) models:

[0966] The server sends the received posted content to an NLP model for linguistic analysis.

[0967] The NLP model analyzes the context of the post and word choice and returns the analysis results.

[0968] 3. Emotion Recognition with Emotion Engine:

[0969] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions.

[0970] The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness.

[0971] 4. Malicious intent detection and AI generation:

[0972] The server determines whether the posted content is malicious based on the analysis results from the analysis engine.

[0973] If the content of a post is determined to be malicious, the server calls on generative artificial intelligence to generate background and supplementary information.

[0974] The generative AI generates background and supplementary information with appropriate tone and content based on the user's emotions recognized by the emotion engine.

[0975] 5. Creating and providing explanatory notes:

[0976] The server creates a description based on the generated background information.

[0977] The explanation is added to the original post and provided to the user.

[0978] The revised post and description are sent to the device, which displays it to the user.

[0979] Examples and prompts

[0980] For example, if a user posts, "The service is terrible! I can't connect at all!", the following processing will occur:

[0981] User comments: "The service is terrible! I can't connect at all!"

[0982] Prompt for generative AI model: "An angry user has posted a complaint about your service. Please respond in a calm tone with context."

[0983] The NLP model analyzes the post, and the emotion engine recognizes the emotion of "anger." The server invokes the generative AI to generate background information such as, "We have recently installed many new base stations and are working to improve the quality of our service." The server then creates a description based on the generated background information, and the revised post and description are displayed on the user's device.

[0984] In this way, the present invention allows users to receive information in a manner that is free from misunderstanding, thereby improving the quality of online communication.

[0985] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0986] Step 1:

[0987] The user uses the device to input posted content that represents their own opinion or feedback. The input posted content is the input in step 1. Specifically, the user uses the keyboard or touch screen to input "The service is so bad! I can't connect at all!" This posted content is stored in the text input field on the device.

[0988] Step 2:

[0989] When the user presses the send button, the entered post content is sent from the terminal to the server. The click event of the send button generates an HTTP request and the posted content is sent to the server. The input is the posted content entered by the user, and the output is the posted content received by the server.

[0990] Step 3:

[0991] The server receives the posted content sent by the user and temporarily stores it in a database. Specifically, it receives an HTTP request and inserts the content into a database such as MySQL or PostgreSQL. The input is the posted content sent, and the output is the posted content stored in the database.

[0992] Step 4:

[0993] The server sends the received post content to a natural language processing (NLP) model and begins analysis. Specifically, the server-side program sends an HTTP request to the NLP model via a REST API to obtain the analysis results. The input is the post content stored in the database, and the output is the analysis results returned by the NLP model.

[0994] Step 5:

[0995] The server sends the analysis results from the NLP model to the emotion engine to recognize the user's emotions. The emotion engine analyzes keywords and expressions contained in the posted content to determine emotions such as anger, dissatisfaction, and sadness. Specifically, the emotion engine receives the analysis results from the NLP model and performs emotion analysis. The input is the analysis results of the NLP model, and the output is the emotion recognition results from the emotion engine.

[0996] Step 6:

[0997] The server determines whether the posted content is malicious based on the analysis results from the analysis engine. If the posted content is determined to be malicious, it calls on a generative artificial intelligence (generative AI model) to generate background information and supplementary information. The generative AI generates background information and supplementary information with an appropriate tone and content based on the user's emotions recognized by the emotion engine. The input is the emotion recognition result, and the output is the generated result from the generative AI model.

[0998] Step 7:

[0999] The server creates a description based on the generated background information. For example, using the background information obtained from the generative AI model, it creates a description such as, "We have recently installed many new base stations and are working to improve the quality of our service." The input is the result generated by the generative AI model, and the output is the created description.

[1000] Step 8:

[1001] The modified post and description are sent from the server to the user's device, which then displays them to the user. The server sends the modified post and description as an HTTP response. The input is the created description and the modified post content, and the output is what is displayed on the user's device. The user can check the displayed description and receive accurate information.

[1002] (Application example 2)

[1003] 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."

[1004] On modern online shopping sites, customer reviews are an important source of information for other customers. However, some customer reviews are emotional, overly negative, and contain factually incorrect information. This can lead to the spread of inaccurate information, misrepresenting product ratings, and damaging a company's reputation. To solve this problem, a system is needed that can appropriately address inaccurate or emotional posts and provide background information or explanatory notes.

[1005] The identification process by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes: means for receiving content posted by a user; means for analyzing the received content using a natural language processing model; means for determining whether the content of the post is malicious based on the analysis result; means for generating background information and additional information using generative artificial intelligence if the content is determined to be malicious; means for creating an explanatory note based on the generated background information and additional information; means for adding the explanatory note to the original post and displaying it to the user; means for monitoring reviews on the online shopping site in real time and determining whether the content of reviews posted by users is appropriate; and means for generating background information and correction information for reviews determined to be inappropriate using generative artificial intelligence and presenting the explanatory note to the user. This makes it possible to prevent the spread of inaccurate information on the online shopping site and provide customers with accurate and reliable information.

[1006] The "means for receiving content posted by users" refers to a mechanism for transmitting text data entered by users via their electronic devices to a server and receiving it.

[1007] "Means for analyzing received posts using a natural language processing model" refers to a mechanism in which the server uses a computer program to analyze the text data it receives and understand the meaning and context of the text.

[1008] "Means for determining whether the content of a post is malicious based on the analysis results" refers to a mechanism for evaluating the analysis data obtained from a natural language processing model to determine whether the content of a post is offensive, insulting, or misleading.

[1009] "Means of using generative AI to generate background and additional information when a post is determined to be malicious" refers to a mechanism that uses generative AI to automatically create information to supplement the situation for posts that are detected to be malicious.

[1010] The "means for creating explanatory notes based on the generated background information and additional information" is a mechanism for organizing the generated supplementary information and creating a document to present it to the user in an easy-to-understand format.

[1011] The "means for adding an explanatory note to the original post and displaying it to the user" is a mechanism for associating the created explanatory note with the content of the original post and displaying it in a format that is easy for the user to view.

[1012] "Means for monitoring reviews on online shopping sites in real time and determining whether the content of reviews posted by users is appropriate" is a system for constantly observing customer reviews on online shopping sites and instantly evaluating the appropriateness of the posts.

[1013] "Means of using generative AI to generate background information and correction information for reviews that are judged to be inappropriate and presenting it to the user as an explanatory note" refers to a mechanism that, when an inappropriate review is detected, uses generative AI to create information and background explanations for correction and presents them to the user as an explanatory note.

[1014] The present invention is a system that analyzes the content of reviews posted by users through an integrated system via electronic devices in real time, generates appropriate background information and correction information as needed, and provides it to the user as explanatory notes.

[1015] First, the review content posted by the user is sent to the server. The server receives the text data using a means for receiving content posted by the user. Next, the received post content is analyzed using a natural language processing (NLP) model. Examples of models used here include spaCy and TextBlob. This analysis analyzes the context and sentiment of the review.

[1016] The server uses an emotion engine to determine whether the post is malicious based on the analysis results. The emotion engine analyzes keywords and expressions contained in the post to determine emotions such as anger, dissatisfaction, and sadness. This determines whether the post is appropriate, and only if it is not is it used to generate background information or corrections using generative artificial intelligence (AI).

[1017] The generative AI used is OpenAI's GPT-3. This generative AI uses user reviews and analysis results as prompts to generate appropriate background information and corrections. An example of a generative prompt is as follows:

[1018] User Review: The quality of this product is terrible!

[1019] Sentiment score: -0.8

[1020] Generate background information.

[1021] The server then uses the generated background and additional information to create explanatory notes. These notes contain information obtained from the generative AI and are presented in a format that is easy for the user to understand. For example, the notes might state, "To improve product quality, we are strengthening our inspection system and introducing a new quality control process."

[1022] Finally, the server adds explanatory notes to the original post and displays them to the user. This system prevents the spread of inaccurate information on online shopping sites and allows customers to obtain reliable information. The entire system operates in real time and aims to improve the quality of reviews on online shopping sites.

[1023] The above is a specific embodiment of the system according to the present invention.

[1024] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1025] Step 1:

[1026] Users post reviews using their electronic devices. The posted content is sent from the user's device to the server. The input is text data entered by the user, and the output is text data sent to the server.

[1027] Step 2:

[1028] The server receives the content posted by the user. The input is text data sent from the terminal, and the received data is temporarily stored in a database. The output is the stored text data.

[1029] Step 3:

[1030] The server analyzes the received post content using a natural language processing (NLP) model. In this step, the context and meaning of the post are analyzed. The input is the stored text data, and the output is the analysis result data. Specifically, the server analyzes the text using NLP tools such as spaCy or TextBlob to extract semantic information.

[1031] Step 4:

[1032] Based on the analysis results, the server determines whether the post content is malicious. The emotion engine analyzes keywords and emotion scores to determine whether the post is malicious. The input is the data obtained by NLP analysis, and the output is the judgment result. Specifically, the emotion analysis algorithm is used to calculate the emotion score of the text.

[1033] Step 5:

[1034] If a post is determined to be malicious, the server uses generative artificial intelligence (AI) to generate background information and additional information. The input is the judgment result and the original text data, and the output is the generated background information and additional information. Specifically, it uses OpenAI GPT-3 or similar to generate a prompt sentence like the following:

[1035] User Review: The quality of this product is terrible!

[1036] Sentiment score: -0.8

[1037] Generate background information.

[1038] The generative AI generates background and additional information based on this prompt.

[1039] Step 6:

[1040] The server creates explanatory notes based on the generated background information and additional information. The input is the generated background information and additional information, and the output is explanatory notes. Specifically, the server arranges the background information and additional information into a consistent format and compiles them into a note to explain to the user.

[1041] Step 7:

[1042] Finally, the server adds the explanatory note to the original post and displays it to the user. The input is the original post and the explanatory note, and the output is the corrected review displayed on the user's device. The user checks the explanatory note through their device and receives the correct information. Specifically, the corrected text data is displayed through the user interface.

[1043] 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.

[1044] 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.

[1045] 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.

[1046] 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.

[1047] 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.

[1048] 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.

[1049] 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).

[1050] 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.

[1051] 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."

[1052] 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.

[1053] 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).

[1054] 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.

[1055] 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.

[1056] 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.

[1057] 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.

[1058] 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.

[1059] 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.

[1060] 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.

[1061] 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.

[1062] 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.

[1063] 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.

[1064] The following is further disclosed regarding the above embodiment.

[1065] (Claim 1)

[1066] means for receiving content posted by users;

[1067] A means for analyzing the received post content using a natural language processing model;

[1068] A means for determining whether the posted content is malicious based on the analysis results;

[1069] A means of generating background and additional information using generative artificial intelligence when a post is determined to be malicious; and

[1070] A means of creating explanatory notes based on the generated background and additional information;

[1071] a means for attaching explanatory notes to the original post for display to the user;

[1072] A system including:

[1073] (Claim 2)

[1074] 10. The system of claim 1, wherein the natural language processing model performs sentiment analysis and keyword matching.

[1075] (Claim 3)

[1076] The system of claim 1, wherein the generative artificial intelligence generates background information and additional information based on fact-checking and relevant statistical data.

[1077] "Example 1"

[1078] (Claim 1)

[1079] means for receiving content posted by users;

[1080] A means for analyzing the received post content using a natural language processing model;

[1081] A means for determining whether the posted content is malicious based on the analysis results;

[1082] A means of generating background and additional information using generative artificial intelligence when a post is determined to be malicious; and

[1083] A means of creating explanatory notes based on the generated background and additional information;

[1084] a means for attaching explanatory notes to the original post for display to the user;

[1085] a means for sending a prompt to a generative artificial intelligence to generate appropriate background information and additional information;

[1086] A system including:

[1087] (Claim 2)

[1088] 10. The system of claim 1, wherein the natural language processing model performs sentiment analysis and keyword matching.

[1089] (Claim 3)

[1090] The system of claim 1, wherein the generative artificial intelligence generates background information and additional information based on fact-checking and relevant statistical data.

[1091] "Application Example 1"

[1092] (Claim 1)

[1093] means for receiving content posted by users;

[1094] A means for analyzing the received post content using a natural language processing model;

[1095] A means for determining whether the posted content is malicious based on the analysis results;

[1096] A means for generating background information and correction information using generative artificial intelligence when a post is determined to be malicious; and

[1097] a means for generating reminders based on the generated background information and correction information;

[1098] a means for attaching the reminder to the original post and displaying it to the user;

[1099] A means of analyzing the content of posts in real time,

[1100] A system including:

[1101] (Claim 2)

[1102] 10. The system of claim 1, wherein the natural language processing model performs sentiment analysis and word matching.

[1103] (Claim 3)

[1104] The system of claim 1, wherein the generative artificial intelligence generates background information and correction information based on fact-checking and relevant statistical data.

[1105] "Example 2: Combining Emotion Engines"

[1106] (Claim 1)

[1107] means for receiving content posted by users;

[1108] A means for analyzing the received posted content using a natural language processing model;

[1109] A means for determining whether the posted content is malicious based on the analysis results;

[1110] A means for generating background and supplemental information using generative artificial intelligence when a post is determined to be malicious; and

[1111] A means for creating explanatory text based on the generated background information and supplementary information;

[1112] a means for adding a description to the original post and displaying it to the user;

[1113] A system including:

[1114] (Claim 2)

[1115] 10. The system of claim 1, wherein the natural language processing model performs emotion recognition.

[1116] (Claim 3)

[1117] The system according to claim 1, wherein the generative artificial intelligence generates text taking into account emotion recognition results.

[1118] "Application example 2 when combining emotion engines"

[1119] (Claim 1)

[1120] means for receiving content posted by users;

[1121] A means for analyzing the received post content using a natural language processing model;

[1122] A means for determining whether the posted content is malicious based on the analysis results;

[1123] A means of generating background and additional information using generative artificial intelligence when a post is determined to be malicious; and

[1124] A means of creating explanatory notes based on the generated background and additional information;

[1125] a means for attaching explanatory notes to the original post for display to the user;

[1126] A means for monitoring reviews on online shopping sites in real time and determining whether the content of reviews posted by users is appropriate;

[1127] A means for generating background information and correction information for reviews that are judged to be inappropriate using generative artificial intelligence and presenting them to the user as explanatory notes;

[1128] A system including:

[1129] (Claim 2)

[1130] 10. The system of claim 1, wherein the natural language processing model performs sentiment analysis and keyword matching.

[1131] (Claim 3)

[1132] The system of claim 1, wherein the generative artificial intelligence generates background information and additional information based on fact-checking and relevant statistical data. [Explanation of symbols]

[1133] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving content posted by users; A means for analyzing the received post content using a natural language processing model; A means for determining whether the posted content is malicious based on the analysis results; A means of generating background and additional information using generative artificial intelligence when a post is determined to be malicious; and A means of creating explanatory notes based on the generated background and additional information; a means for attaching explanatory notes to the original post for display to the user; A system including:

2. The system of claim 1 , wherein the natural language processing model performs sentiment analysis and keyword matching.

3. 2. The system of claim 1, wherein the generative artificial intelligence generates background information and additional information based on fact-checking and relevant statistical data.

Images