System

Abstract

A system is provided.SOLUTION: A system comprising: a server means for analyzing news articles or videos received from user terminals; a fake news analysis means for analyzing the news articles; a fake video analysis means for analyzing the videos; a multimodal integration means for integrating analysis results of the news articles and the videos; and a user interface means for displaying the integration results on the user terminals.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] In recent years, the rapid spread of false information, i.e., fake news and fake videos, on the Internet and social networking sites has become a prominent problem. Such false information has a serious impact on society as a whole, often causing misunderstandings and confusion. Therefore, there is a need for technology that can quickly and accurately detect such false information and prevent users from being misled by it. However, current technology has difficulty fully adapting to the highly evolving fake technology, and improvements in reliability and accuracy are needed. To address these issues, the present invention provides a system that can accurately identify fake news and fake videos and provide users with accurate information. [Means for solving the problem]

[0005] The present invention provides a system including a server that analyzes news articles or videos received from a user terminal, a fake news analysis unit that analyzes news articles, a fake video analysis unit that analyzes videos, a multimodal data integration unit that integrates the analysis results of the news articles and videos, and a user interface unit that displays the integrated results on the user terminal. Specifically, the fake news analysis unit uses a generative AI model that analyzes the language patterns, context, and writing style of news articles. The fake video analysis unit analyzes video footage frame by frame and uses technology to detect iris movements and changes in facial expressions. The multimodal data integration unit then integrates these analysis results to provide the user with an overall reliability score. This system enables users to easily identify false information and make decisions based on accurate information.

[0006] A "user terminal" is a device used by a user to input or receive information.

[0007] The "server means" is a device or system that has the function of receiving data sent from a user terminal and passing that data to an appropriate analysis module.

[0008] A "fake news analysis tool" is a device or system that uses generative AI models to analyze the language patterns, context, and writing style of news articles and distinguish between false and real information.

[0009] A "fake video analysis means" is a device or system that analyzes the video and audio of a video and detects any unnatural or abnormal points therein.

[0010] A "multimodal data integration means" is a device or system for integrating the analysis results of news articles and videos and calculating an overall reliability score.

[0011] The "user interface means" is a device or system that displays the analysis results on a user terminal in a visually easy-to-understand format, allowing the user to easily understand the information.

[0012] A "generative AI model" is a model that uses machine learning based on large datasets to analyze the authenticity of given text or video.

[0013] A "trustworthiness score" is a numerical evaluation index that indicates how trustworthy the information is based on the analysis of news articles and videos.

[0014] "Language patterns" refer to the characteristic vocabulary and grammatical structures found in text, and are used to determine the truth of information.

[0015] "Context" refers to the relevance of surrounding sentences and the entire text to ensure that the content of a news article or text is consistent, and is a factor that is taken into account during analysis.

[0016] "Writing style" refers to the way a piece of writing is expressed and the characteristics of the writing style, and is one factor used to determine whether the information is fake.

[0017] The iris is a detail inside the eye, and analyzing the movement and changes in this part of the video is an important factor in determining the authenticity of the video.

[0018] "Changes in facial expression" is an item used to detect whether facial expressions in a video are natural, and is used to identify fake videos. [Brief explanation of the drawings]

[0019] [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

[0020] 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.

[0021] First, the terms used in the following description will be explained.

[0022] 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).

[0023] 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.

[0024] 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.

[0025] 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.

[0026] 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."

[0027] [First embodiment]

[0028] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0029] 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.

[0030] 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).

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0036] 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.

[0037] 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.

[0038] 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.

[0039] 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."

[0040] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, and a user interface means.

[0041] Program processing

[0042] The user sends information suspected to be fake news or fake video from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0043] Fake News Analysis Module

[0044] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0045] Examples:

[0046] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0047] Fake video analysis module

[0048] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0049] Examples:

[0050] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0051] Multimodal Data Integration Module

[0052] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0053] Examples:

[0054] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0055] User Interface Module

[0056] The final confidence score and analysis results are sent to the user interface module, which receives them and displays them on the user's terminal.

[0057] Examples:

[0058] When the user checks the results on their device, the analysis result will show "This news article is likely to be unreliable." At the same time, the analysis result for the video will show "This video is likely fake." This will help users make decisions based on accurate information, without being misled by false information.

[0059] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0060] The processing flow will be explained below.

[0061] Step 1:

[0062] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0063] Step 2:

[0064] The server distributes the data received from users to either the fake news analysis module or the fake video analysis module, verifying the data format and content and performing appropriate preprocessing (e.g., text cleaning and video frame segmentation).

[0065] Step 3:

[0066] The server sends news articles to a fake news analysis module, which uses a generative AI model to analyze the article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0067] Step 4:

[0068] The server sends the video to a fake video analysis module, which analyzes the video and audio data separately. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis analyzes the tone and patterns of the voice. A generative AI model combines these data to determine whether the video is authentic or fake.

[0069] Step 5:

[0070] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0071] Step 6:

[0072] The server sends the final reliability score and analysis results to a user interface module, which displays the results on a user interface and provides the results to the user terminal in a visually easy-to-understand format.

[0073] Step 7:

[0074] Users can check the analysis results displayed on their device and make decisions based on them. The analysis results are presented with a specific reliability score and the basis for the analysis, allowing users to refer to this information and obtain accurate information without being misled by incorrect information.

[0075] Example 1

[0076] 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."

[0077] Today, there are many news articles and videos on the Internet, some of which contain false information such as fake news and videos. This makes it difficult for users to access accurate information, which can have a negative impact on decision-making. In particular, in the case of political and economic news, which have a great social impact, the spread of false information increases the risk of causing confusion and anxiety. Therefore, a system is needed that allows users to easily determine the authenticity of information on the Internet.

[0078] 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.

[0079] In this invention, the server includes an information processing device that analyzes news articles or videos received from a user terminal, an information analysis device that analyzes the news articles, a video analysis device that analyzes the videos, a data integration device that integrates the analysis results of the news articles and videos, and a display device that displays the integration results on the user terminal, thereby enabling users to determine the authenticity of the news articles and videos with high accuracy and make decisions based on accurate information.

[0080] A "user terminal" is a device through which a user enters information and communicates with the system.

[0081] An "information processing device" is a device that has the function of distributing news articles and videos received from user terminals to analysis modules.

[0082] An "information analysis device" is a device that analyzes received news articles using language patterns, context, and writing style.

[0083] A "video analysis device" is a device that analyzes the video and audio of received video on a frame-by-frame basis to determine its authenticity.

[0084] The "data integration device" is a device that integrates the analysis results of news articles and videos and calculates an overall reliability score.

[0085] The "display device" is a device that displays the analysis results and integration results on a user terminal.

[0086] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can accurately determine the authenticity of news articles and videos.

[0087] A "prompt statement" is an instruction statement that causes the generative AI model to perform analysis.

[0088] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of an information processing device, an information analysis device, a video analysis device, a data integration device, and a display device.

[0089] Information processing device

[0090] The server receives news articles and videos sent by users from their devices. Users have the ability to upload news article text files and video files. The received data is sorted into the appropriate analysis devices depending on whether it is a news article or a video.

[0091] Information analysis device

[0092] The information analysis device analyzes the text of news articles. This process utilizes a generative AI model. The generative AI model is trained based on past datasets and analyzes the language patterns, context, and writing style of news articles. This allows it to determine the veracity of news articles with high accuracy and returns the results to the server as a reliability score.

[0093] For example, if a user uploads an article about a politician's remarks from their device, the server sends the article to an information analysis device, where the generative AI model analyzes the language patterns and context. As a result, it determines that the news article has low reliability and generates a reliability score.

[0094] Video analysis equipment

[0095] Video analysis devices analyze videos. In this process, the video is divided into frames, and the video and audio data of each frame is analyzed. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis uses technology to analyze voice tone and patterns. A generative AI model combines these data to determine the authenticity of the video. The results are sent back to the server as a reliability score.

[0096] For example, if a user uploads a video introducing a new product from their device, the server sends the video to a video analysis device, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the server determines that the video is likely to be fake and generates a reliability score.

[0097] Data integration device

[0098] The server receives the analysis results from the information analysis device and the video analysis device and sends them to the data integration device. This device integrates the analysis results of the news article and the video and calculates an overall reliability score. This comprehensively evaluates the authenticity of the news article and the video.

[0099] display device

[0100] The final reliability score and analysis results are sent to a display device, and the server receives them and displays them on the user's device. The user can check the results through their device, and the analysis results will be displayed as "This news article is likely to be unreliable" or "This video is likely to be fake." This will allow users to make decisions based on accurate information, without being misled by false information.

[0101] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0102] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0103] Step 1:

[0104] Users upload suspected fake news or fake videos from their own devices. They select a text file or video file of the news article and click the send button. The input is the text file or video file of the news article, and the output is that this data is sent to the server.

[0105] Specific behavior:

[0106] 1. The user opens a browser on their device and accesses the system's web page.

[0107] 2. Drag and drop your news article or video file or click the "Choose File" button to upload it.

[0108] 3. Once the upload is complete, the user clicks the "Submit" button.

[0109] Step 2:

[0110] The server receives the data sent by the user and distributes the news articles and videos to the corresponding analysis modules. The input is the text file or video file of the news article sent by the user, and the output is the distributed data.

[0111] Specific behavior:

[0112] 1. The server receives the transmitted data.

[0113] 2. Determine the type of data (text or video).

[0114] 3. Text data is sent to the information analysis device, and video data is sent to the video analysis device.

[0115] Step 3:

[0116] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The input is the news article text, and the output is a credibility score.

[0117] Specific behavior:

[0118] 1. The text of a news article is sent to an information analysis device.

[0119] 2. The generative AI model receives a prompt to analyze the news article and begins the analysis.

[0120] 3. Language patterns, context, and writing style are analyzed.

[0121] 4. A confidence score is calculated and sent back to the server.

[0122] Step 4:

[0123] The server sends the video to the fake video analysis module, which splits the video into frames and analyzes the video and audio data. The input is the video file, and the output is a confidence score.

[0124] Specific behavior:

[0125] 1. The video file is sent to the video analyzer.

[0126] 2. The video is divided into frames.

[0127] 3. The video data is analyzed using technology that detects iris movement and changes in facial expression.

[0128] 4. Audio data is analyzed for tone and patterns of voice.

[0129] 5. A generative AI model combines the video and audio data to calculate a reliability score.

[0130] 6. The confidence score is sent back to the server.

[0131] Step 5:

[0132] The server receives the analysis results from the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module. The input is the analyzed credibility score, and the output is the integrated overall credibility score.

[0133] Specific behavior:

[0134] 1. The credibility scores of the news article and the video arrive at the server.

[0135] 2. These scores are sent to a data aggregator.

[0136] 3. The data aggregator combines the credibility scores of the news article and the video to calculate an overall credibility score.

[0137] Step 6:

[0138] The final reliability score and analysis results are sent from the server to the user interface module, which receives them and displays the results on the user's terminal. The input is the overall reliability score, and the output is the display of the analysis results.

[0139] Specific behavior:

[0140] 1. The overall reliability score and analysis results are sent to the user interface module.

[0141] 2. The user interface module processes the results to display them on the user terminal.

[0142] 3. The user checks the results on their device and sees messages such as "This news article is likely to be unreliable" or "This video may be fake."

[0143] In this way, the system can achieve high accuracy in identifying fake news and fake videos and provide accurate information to users.

[0144] (Application example 1)

[0145] 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."

[0146] In recent years, a large number of news articles and videos have been circulating on the Internet, and many of them contain fake news and videos. Such misinformation can lead to incorrect perceptions and judgments among users, potentially causing social unrest. Therefore, there is a need for a method to accurately determine the authenticity of news articles and videos and provide users with accurate information. In particular, for content distribution services, it is important to have a function that can determine the reliability of the content viewed by users in real time and immediately display a reliability score.

[0147] 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.

[0148] In this invention, the server includes server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing news articles, fake video analysis means for analyzing videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and means for determining the authenticity of the news articles and videos viewed by the user in real time and displaying a reliability score. This allows users to check the reliability of the content they view in real time, enabling them to make decisions based on accurate information without being misled by misinformation.

[0149] A "user terminal" is a device operated by a user, and includes a smartphone, tablet, personal computer, and the like.

[0150] The "server means" is a central processing unit that performs analysis and data processing, and has the function of communicating with user terminals via a network.

[0151] The "fake news analysis tool" is an analytical module that determines the authenticity of news articles and has the ability to analyze language patterns, context, and writing style.

[0152] The "fake video analysis means" is an analysis module that determines the authenticity of a video, and includes means for analyzing the video on a frame-by-frame basis and detecting iris movement and changes in facial expression.

[0153] The "multimodal data integration means" has the function of integrating the analysis results obtained from the fake news analysis means and the fake video analysis means and calculating an overall reliability score.

[0154] The "user interface means" is an interface that displays the analysis results to the user, and refers to a screen or application that the user operates.

[0155] "Real-time judgment" is a process that instantly analyzes the authenticity of news articles and videos viewed by users and immediately provides the results to the users.

[0156] A "trust score" is a number generated based on the analysis results that indicates the reliability of a news article or video.

[0157] MODE FOR CARRYING OUT THE INVENTION

[0158] The present invention provides a system for analyzing news articles or videos sent from a user terminal and evaluating their reliability, which includes a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, a user interface unit, and a unit for displaying a reliability score in real time.

[0159] System Configuration

[0160] Server means:

[0161] The server distributes news articles or videos received from user terminals to the analysis module. The server functions as a central processing unit with high-performance processing capabilities, and receives and transmits data in real time via the network.

[0162] Fake news analysis methods:

[0163] The fake news analysis tool analyzes the language patterns, context, and writing style of news articles using a specific generative AI model. The generative AI model is trained on a large number of past news articles and can determine the veracity of news articles with high accuracy. The server receives the analysis results as a credibility score.

[0164] Fake video analysis methods:

[0165] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. Video analysis detects iris movement and changes in facial expressions, while audio analysis analyzes voice tone and patterns. A generative AI model integrates these data to determine the authenticity of the video. The server receives the analysis results as a reliability score.

[0166] Multimodal data integration methods:

[0167] The multimodal data integration method integrates the results obtained from fake news analysis and fake video analysis to calculate a comprehensive credibility score, which enables accurate credibility assessment that takes into account not only the judgment results of a single news article or video, but also the relationship between the two.

[0168] User Interface Methods:

[0169] The user interface means plays a role in displaying the analysis results to the user. The analysis results are displayed on the user terminal in real time, allowing the user to immediately confirm the reliability of the content.

[0170] Example of operation

[0171] When a user uploads a news article (path / to / content / news.txt) from their smartphone, the server routes the news article to the fake news analysis tool, which analyzes language patterns and context using a generative AI model. The resulting trust score is displayed in real time on the user's device, such as "News Trust Score: 85." Similarly, when a user uploads a video, the fake video analysis tool analyzes the video and audio, and displays a trust score.

[0172] Prompt Sentence Examples

[0173] An example of a prompt sentence to input to a generative AI model is, "Please rate the credibility of the following news article." By providing the text of the news article along with this prompt sentence to the generative AI model, the model analyzes the article's language patterns and context and returns a credibility score.

[0174] As described above, the system of the present invention can accurately determine the authenticity of news articles and videos viewed by users and display a reliability score in real time, allowing users to make decisions based on accurate information without being misled by misinformation.

[0175] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0176] Step 1:

[0177] The user selects a news article or video from a user device such as a smartphone and sends it to the server. The input is a text file or video file of the news article, which the server receives. Specifically, the user presses the upload button in the application, selects a file from the file selection screen, and sends it.

[0178] Step 2:

[0179] The server distributes the received data to the analysis modules. It receives news articles or video files as input, and distributes them to the fake news analysis means if they are news articles, and to the fake video analysis means if they are videos. Specifically, the server determines the type of file it receives and transfers it to the corresponding analysis module.

[0180] Step 3:

[0181] The fake news analysis method analyzes the content of news articles. It receives the text of a news article as input and uses a generative AI model to analyze the language patterns, context, and writing style. It generates a credibility score as output. Specifically, the text of a news article is input into the generative AI model, and a credibility score is returned as the analysis result.

[0182] Step 4:

[0183] The fake video analysis method analyzes videos frame by frame and simultaneously analyzes audio data. It receives a video file as input and analyzes the video and audio data for each frame. It uses a generative AI model to determine iris movement, facial expression changes, and vocal tone and patterns, and generates a credibility score as output. Specifically, the video is divided into frames, and the data for each frame is analyzed.

[0184] Step 5:

[0185] The multimodal data integration means integrates the results from the fake news analysis means and the fake video analysis means. It receives the credibility scores of news articles and videos as input and calculates an overall credibility score. It produces this overall score as output. Specifically, multiple credibility scores are input into the integration algorithm, and a final credibility score is calculated.

[0186] Step 6:

[0187] The user interface means displays the calculated trustworthiness score on the user's terminal. It receives the overall trustworthiness score from the server as input and converts it into a format to be displayed on the user's terminal. As an output, the trustworthiness score is displayed on the user's screen. Specifically, the trustworthiness score is displayed on the application screen in a format such as "News Trust Score: 85."

[0188] 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.

[0189] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, and a user interface unit. Furthermore, the present invention is equipped with an emotion engine that recognizes the user's emotions and incorporates a function to optimize the presentation of the analysis results.

[0190] Program processing

[0191] The user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0192] Fake News Analysis Module

[0193] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0194] Examples:

[0195] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0196] Fake video analysis module

[0197] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0198] Examples:

[0199] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0200] Emotion Engine

[0201] Furthermore, an emotion engine is incorporated into the user interface means to recognize the user's emotional state in real time. The emotion engine analyzes the user's facial expressions, voice tone, and input patterns to understand the user's emotional state when receiving information. This allows the method of presenting the analysis results to be optimized according to the user's emotions.

[0202] Examples:

[0203] When the emotion engine detects anxiety or discomfort from the user's facial expression or voice while the user is viewing the analysis results, the user interface means provides the analysis results in a more easily understandable format or displays additional explanations to help the user understand.

[0204] Multimodal Data Integration Module

[0205] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0206] Examples:

[0207] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0208] User Interface Module

[0209] The final confidence score and analysis results are sent to the user interface module. The server receives them and displays the results on the user's device. The emotional engine detects the user's emotional state, so the display format of the results is adjusted according to the user's emotional state.

[0210] Examples:

[0211] When the user checks the results on their device, the analysis result will display, "This news article is unreliable and is likely fake news." At the same time, the analysis result for the video will display, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0212] In this way, the system of the present invention achieves highly accurate identification of fake news and fake videos, and furthermore, by optimizing the way in which the analysis results are presented according to the user's emotional state, it is possible to provide accurate information to users and prevent them from being misled by false information.

[0213] The processing flow will be explained below.

[0214] Step 1:

[0215] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0216] Step 2:

[0217] The server receives the data sent by the user, determines whether it is a news article or a video, and then assigns it to either the fake news analysis module or the fake video analysis module depending on the determined data format.

[0218] Step 3:

[0219] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The result of this analysis is a credibility score, which is returned to the server.

[0220] Step 4:

[0221] The server sends the video to a fake video analysis module, which divides the video into frames and detects changes in iris movement and facial expressions in each frame. It also analyzes the audio data, analyzing the tone and patterns of the voice. Based on this data, the generative AI model determines whether the video is authentic, and a reliability score is generated and sent back to the server.

[0222] Step 5:

[0223] The server receives the analysis results provided by the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module, which integrates the analysis results of the news article and the video to calculate an overall credibility score.

[0224] Step 6:

[0225] The server recognizes the user's emotional state in real time through the emotion engine, which grasps the user's emotional state by analyzing the user's facial expressions, voice tone, and input patterns through the user interface means.

[0226] Step 7:

[0227] The server sends the final confidence score and analysis results to the user interface module, which presents the analysis results to the user in the most appropriate manner based on feedback from the emotion engine.

[0228] Step 8:

[0229] The user checks the analysis results on their device, which display a credibility score and supporting reasons. If the emotion engine detects a negative emotion from the user, the analysis results will provide further detailed explanations and supplementary information. For example, it may say, "This news article has low credibility and is likely fake news," providing details of the language patterns and writing style that support this.

[0230] This allows users to identify unreliable information and make decisions based on accurate information. The introduction of an emotion engine improves user understanding and makes analysis results more easily accepted.

[0231] Example 2

[0232] 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."

[0233] Conventional systems for determining the authenticity of news articles and videos are biased toward analyzing a single data mode (only articles or videos), making it difficult to integrate and analyze multimodal data. Furthermore, when users receive analysis results, they are presented in a uniform display format without taking into account their level of understanding or emotional state, which can result in insufficient information being provided to the user. Furthermore, technologies for improving the reliability of analysis results are limited. There is a need for a system that can resolve these issues, achieve high-accuracy identification of fake news and fake videos, and optimize the presentation of analysis results according to the user's emotional state.

[0234] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0235] In this invention, the server includes a processing device that analyzes text data or video data received from a user device, a false information analysis device that analyzes the text data, a video analysis device that analyzes the video data, a display device that displays the analysis results and integration results on the user device, and an emotion recognition device incorporated in the display device that recognizes the user's emotional state in real time and optimizes the presentation method of the analysis results. This enables multimodal data integration that integrates the analysis results of news articles and videos, realizing optimal information presentation according to the user's emotional state. Furthermore, high-precision analysis using a generative AI model enables reliable identification of fake news and fake videos.

[0236] A "user terminal" is a computing device operated by a user, and is a device for transmitting news articles and video data to a server.

[0237] "Text data" refers to information in text format such as news articles.

[0238] "Video data" refers to frame-by-frame video information such as news videos and other video files.

[0239] "Processor means" refers to a computing device for analyzing data received from a user terminal and is a means for identifying audio and video data.

[0240] A "disinformation analysis tool" is a device for determining the authenticity of received text data, and uses a generative AI model to analyze language patterns, context, and writing style.

[0241] The "video analysis means" is a device for determining the authenticity of received video data, analyzing each frame and detecting iris movements and changes in facial expression.

[0242] The "data integration means" is a device that integrates the analysis results obtained by the false information analysis means and the video analysis means and calculates an overall reliability score.

[0243] The "display means" is a device that displays the analysis results and the integration results on a user terminal.

[0244] The "emotion recognition means" is a device that recognizes the user's emotional state in real time and optimizes the method of presenting the analysis results.

[0245] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can determine truth or falsehood with high accuracy.

[0246] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is mainly composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion recognition means. Each means will be described in detail below.

[0247] Server means:

[0248] The server has a processing device for analyzing text data (news articles) or video data (videos) received from user terminals. The server collects data from users and distributes it to the appropriate analysis module. Specifically, text data is sent to the fake news analysis means, and video data is sent to the fake video analysis means.

[0249] Fake news analysis methods:

[0250] The server analyzes the received news articles using a disinformation analysis methodology. This analysis methodology incorporates a generative AI model that performs detailed analysis of the news article's language patterns, context, and writing style. The generative AI model is trained on a large amount of previously collected data sets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0251] Examples:

[0252] When a user uploads a news article about a politician's remarks from their device, the server sends the article to a fake news analysis tool. The generative AI model analyzes language patterns and context, determining that the news article is low in credibility and generating a credibility score.

[0253] Example prompt sentence:

[0254] Please determine whether this news article is true or false.

[0255] Fake video analysis methods:

[0256] The server analyzes the received video using a video analysis method. The video analysis method analyzes the video frame by frame, analyzing iris movement, changes in facial expressions, and voice tone and patterns. The generative AI model integrates this data to determine the authenticity of the video. The analysis result is also returned to the server as a reliability score.

[0257] Examples:

[0258] When a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis means. The video analysis means analyzes the iris movement and changes in facial expressions, and the audio analysis means analyzes the tone and patterns of the voice. As a result, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0259] Example prompt sentence:

[0260] Please tell me if this video is real or not

[0261] Multimodal data integration methods:

[0262] The server sends the analysis results from the false information analysis means and the video analysis means to the data integration means, which integrates the results of both analyses and calculates an overall reliability score, thereby matching the analysis results of the news article and the video, enabling a more comprehensive determination of whether the news article is true or false.

[0263] Examples:

[0264] When a user uploads a news article and its associated video at the same time, the server combines the analysis results of these data to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0265] Emotion recognition means:

[0266] The server is equipped with an emotion recognition system that optimizes the presentation of analysis results according to the user's emotional state. This system analyzes the user's facial expressions, tone of voice, and input patterns in real time to understand the user's emotional state when receiving information.

[0267] Examples:

[0268] If the emotion recognition means detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the analysis results will be presented in a more understandable format or additional explanations will be displayed to help the user understand.

[0269] User Interface Methods:

[0270] The server displays the final reliability score and analysis results on the user's device via a user interface. The user can check the analysis results and receive detailed explanations through the device. The emotion recognition device monitors the user's emotional state in real time and adjusts the display format accordingly.

[0271] Examples:

[0272] When the user checks the results on their device, the analysis results will display information such as "This news article is unreliable and is likely fake news." At the same time, the analysis results for the video will also display "This video is likely fake." If the emotion recognition method detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0273] Advantages of this embodiment:

[0274] The system of the present invention not only achieves high-accuracy identification of fake news and fake videos, but also provides an optimal display method that matches the user's emotional state, thereby providing users with more accurate and reliable information and preventing them from being misled by false information.

[0275] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0276] Step 1:

[0277] Users select suspected fake news articles or videos from their own devices and send them to the server. The input is the text data or video file of the news article, and the output is the data sent to the server. Users can either drag and drop files into a dedicated upload form or click the select button to select a file.

[0278] Step 2:

[0279] The server distributes data received from users to the analysis modules. The input is text data or video data sent by the user, and the output is data sent to each analysis module. The server first checks the data format and then sends text data to the fake news analysis module and video data to the fake video analysis module.

[0280] Step 3:

[0281] The server sends the received news article to the fake news analysis module. The input is the text data of the news article, and the output is a credibility score. The generative AI model analyzes the news article's language patterns, context, and writing style. Specifically, it checks for specific keywords, sentence structure, and contextual consistency within the text, and returns the analysis result to the server as a credibility score.

[0282] Step 4:

[0283] The server sends the received video to the fake video analysis module. The input is the video data, and the output is a credibility score. The analysis module divides the video into frames and analyzes the video and audio data of each frame individually. Specifically, video analysis detects iris movement and changes in facial expression, and audio analysis meticulously analyzes the tone and patterns of the voice. The generative AI model integrates this data, determines whether the video is authentic, and sends a credibility score back to the server.

[0284] Step 5:

[0285] The server sends the analysis results from the fake news analysis module and fake video analysis module to the data integration means. The input is the analysis results of the news article and video, and the output is an overall credibility score. The data integration means integrates the credibility score of the news article and the credibility score of the video to calculate an overall credibility score. The integration includes weighting each score and analyzing correlations.

[0286] Step 6:

[0287] The server uses emotion recognition means to analyze the user's emotional state in real time and optimize the display method of the analysis results. The input is the user's facial expression data and voice tone, and the output is optimized display content. The emotion recognition means analyzes the user's facial expression, voice tone, and input patterns to understand the emotional state the user is in when receiving information. The user's emotional state is monitored through a camera and microphone, and the display content is adjusted based on the analysis results.

[0288] Step 7:

[0289] The server displays the final reliability score and analysis results on the user's device via the user interface means. The input is the overall reliability score and analysis results, and the output is the results displayed on the user's device. The user can check the analysis results through their device and receive a detailed explanation. Specifically, the results display screen will display information such as "This news article is unreliable and therefore likely to be fake news," and if the emotion recognition means detects the user's anxiety, additional explanations and answers to questions will be displayed.

[0290] (Application example 2)

[0291] 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."

[0292] In modern society, fake news and fake videos have become a serious problem because many news articles and videos spread instantly via the Internet. In particular, in the advertising industry, there is a high risk that companies' trust will be damaged if advertisements containing unreliable information are delivered to consumers. The present invention aims to provide a system that can accurately identify such fake news and fake videos and check the reliability of advertisements in real time.

[0293] The identification processing 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 server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing the news articles, fake video analysis means for analyzing the videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and an emotion engine for recognizing the user's emotional state in real time and optimizing the presentation method of the analysis results. This increases the reliability of information in advertisements, reduces corporate risks due to incorrect information, and enables the creation of an environment in which users can view advertisements with peace of mind.

[0294] The "server means" refers to a device that distributes data received from user terminals to processing and analysis modules.

[0295] A "fake news analysis tool" is a device that analyzes the language patterns, context, and writing style of news articles to identify fake news.

[0296] A "generative AI model" is an artificial intelligence model that is trained on historical datasets and analyzes news articles and videos.

[0297] A "fake video analysis means" is a device that analyzes video on a frame-by-frame basis and detects iris movements and changes in facial expression.

[0298] A "multimodal data integration means" is a device that integrates the results of fake news analysis and fake video analysis and calculates an overall reliability score.

[0299] The "user interface means" refers to a device that displays the analysis results on a user terminal, allowing the user to confirm the results.

[0300] An "emotion engine" is a device that analyzes a user's facial expressions and tone of voice in real time to recognize the user's emotional state.

[0301] A "credibility score" is a numerical representation of the reliability of a news article or video based on its veracity.

[0302] This invention is a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. The system includes a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion engine.

[0303] First, the user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0304] The server means has the function of receiving news articles or videos from user terminals and distributing them to the fake news analysis means and fake video analysis means.

[0305] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of received news articles. This generative AI model is trained based on past datasets and can determine the veracity of news with high accuracy. For example, if a user uploads an article about "a politician's statement" from their device, the server sends the article to the fake news analysis method, where the generative AI model analyzes the language patterns and context. As a result, it determines that "this news article has low credibility" and generates a credibility score.

[0306] The fake video analysis method divides the received video into frames and analyzes the video and audio data of each frame. Specifically, the video analysis uses OpenCV to detect iris movements and changes in facial expressions, and the audio analysis uses LibROSA to analyze voice tone and patterns. The generative AI model integrates this data and determines the authenticity of the video. For example, if a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis method, which analyzes the video and audio. If the iris movements and changes in facial expressions are unnatural, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0307] The emotion engine recognizes the user's emotional state in real time while the user is checking the analysis results. The emotion engine understands the emotional state of the user when receiving information by analyzing the user's facial expressions and tone of voice. For example, if the emotion engine detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the user interface means will provide the analysis results in a more understandable format or display additional explanations to help the user understand.

[0308] The multimodal data integration means receives the analysis results of the fake news analysis means and the fake video analysis means from the server, and integrates these results to calculate an overall credibility score. For example, if a user simultaneously uploads a news article and its related video, the server integrates these analysis results to generate an overall credibility score. This score is a comprehensive assessment of the authenticity of the news article and the video.

[0309] The user interface means has the function of displaying the final reliability score and analysis results on the user's device. The server receives this and displays the results on the user's device. Because the emotion engine detects the user's emotional state, the display format of the results is adjusted according to the user's emotional state. For example, when the user checks the results on their device, the analysis result may say, "This news article is unreliable, so it is likely to be fake news." At the same time, the analysis result for the video may also say, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to questions may be displayed to help the user understand.

[0310] Example prompt sentence:

[0311] Please check whether the contents of the advertised article below are trustworthy.

[0312] "Politician A announced a new policy that he said would dramatically improve the economy."

[0313] Result: This advert has low credibility. A similar pattern was found in many fake news stories. Credibility score: 30%

[0314] In this way, the system of the present invention achieves high-accuracy identification of fake news and fake videos, and further optimizes the presentation method of the analysis results according to the user's emotional state, thereby providing users with accurate and safe information and increasing the reliability of information in advertisements.

[0315] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0316] Step 1:

[0317] A user uploads a news article or video from a terminal. The input data includes a news article text file or video file. The terminal uploads this data and sends it to the server. The server receives this data.

[0318] Step 2:

[0319] The server distributes the received data to the analysis modules. If a news article is received, the data is sent to the fake news analysis means, and if a video is received, the data is sent to the fake video analysis means. The input data is a news article or a video file. The output is the data to be processed by each analysis module.

[0320] Step 3:

[0321] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of news articles. The text of the news article is used as input, and the generative AI model performs data analysis. The output is a credibility score based on the analyzed language patterns, context, and writing style.

[0322] Step 4:

[0323] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. The video file is used as input. Specifically, OpenCV is used to detect iris movement and changes in facial expressions, and LibROSA is used to analyze voice tone and patterns. The output is a reliability score based on the video and audio analysis results.

[0324] Step 5:

[0325] The server transmits the analysis results from the fake news analysis means and the fake video analysis means to the multimodal data integration means. The input data are reliability scores. The multimodal data integration means calculates an overall reliability score based on these results. The output is an integrated overall reliability score.

[0326] Step 6:

[0327] The user interface means displays the received overall reliability score on the user terminal. The input data is the reliability score. When the user checks the results, the emotion engine analyzes the user's emotional state in real time and optimizes the display format. Specifically, it analyzes the user's facial expressions and tone of voice and selects a method for presenting the results according to their emotional state. The output is a display of the optimized reliability score.

[0328] Step 7:

[0329] The user checks the results on their device. The input data is the analyzed confidence score and its detailed explanation. If the emotion engine detects anxiety or discomfort while the user is viewing the analysis results, it provides additional explanations and details to help the user understand. The output is a display of the detailed analysis results according to the user's emotional state.

[0330] 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.

[0331] 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.

[0332] 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.

[0333] [Second embodiment]

[0334] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0335] 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.

[0336] 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).

[0337] 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.

[0338] 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.

[0339] 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).

[0340] 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.

[0341] 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.

[0342] 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.

[0343] 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.

[0344] 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.

[0345] 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."

[0346] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, and a user interface means.

[0347] Program processing

[0348] The user sends information suspected to be fake news or fake video from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0349] Fake News Analysis Module

[0350] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0351] Examples:

[0352] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0353] Fake video analysis module

[0354] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0355] Examples:

[0356] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0357] Multimodal Data Integration Module

[0358] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0359] Examples:

[0360] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0361] User Interface Module

[0362] The final confidence score and analysis results are sent to the user interface module, which receives them and displays them on the user's terminal.

[0363] Examples:

[0364] When the user checks the results on their device, the analysis result will show "This news article is likely to be unreliable." At the same time, the analysis result for the video will show "This video is likely fake." This will help users make decisions based on accurate information, without being misled by false information.

[0365] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0366] The processing flow will be explained below.

[0367] Step 1:

[0368] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0369] Step 2:

[0370] The server distributes the data received from users to either the fake news analysis module or the fake video analysis module, verifying the data format and content and performing appropriate preprocessing (e.g., text cleaning and video frame segmentation).

[0371] Step 3:

[0372] The server sends news articles to a fake news analysis module, which uses a generative AI model to analyze the article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0373] Step 4:

[0374] The server sends the video to a fake video analysis module, which analyzes the video and audio data separately. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis analyzes the tone and patterns of the voice. A generative AI model combines these data to determine whether the video is authentic or fake.

[0375] Step 5:

[0376] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0377] Step 6:

[0378] The server sends the final reliability score and analysis results to a user interface module, which displays the results on a user interface and provides the results to the user terminal in a visually easy-to-understand format.

[0379] Step 7:

[0380] Users can check the analysis results displayed on their device and make decisions based on them. The analysis results are presented with a specific reliability score and the basis for the analysis, allowing users to refer to this information and obtain accurate information without being misled by incorrect information.

[0381] Example 1

[0382] 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."

[0383] Today, there are many news articles and videos on the Internet, some of which contain false information such as fake news and videos. This makes it difficult for users to access accurate information, which can have a negative impact on decision-making. In particular, in the case of political and economic news, which have a great social impact, the spread of false information increases the risk of causing confusion and anxiety. Therefore, a system is needed that allows users to easily determine the authenticity of information on the Internet.

[0384] 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.

[0385] In this invention, the server includes an information processing device that analyzes news articles or videos received from a user terminal, an information analysis device that analyzes the news articles, a video analysis device that analyzes the videos, a data integration device that integrates the analysis results of the news articles and videos, and a display device that displays the integration results on the user terminal, thereby enabling users to determine the authenticity of the news articles and videos with high accuracy and make decisions based on accurate information.

[0386] A "user terminal" is a device through which a user enters information and communicates with the system.

[0387] An "information processing device" is a device that has the function of distributing news articles and videos received from user terminals to analysis modules.

[0388] An "information analysis device" is a device that analyzes received news articles using language patterns, context, and writing style.

[0389] A "video analysis device" is a device that analyzes the video and audio of received video on a frame-by-frame basis to determine its authenticity.

[0390] The "data integration device" is a device that integrates the analysis results of news articles and videos and calculates an overall reliability score.

[0391] The "display device" is a device that displays the analysis results and integration results on a user terminal.

[0392] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can accurately determine the authenticity of news articles and videos.

[0393] A "prompt statement" is an instruction statement that causes the generative AI model to perform analysis.

[0394] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of an information processing device, an information analysis device, a video analysis device, a data integration device, and a display device.

[0395] Information processing device

[0396] The server receives news articles and videos sent by users from their devices. Users have the ability to upload news article text files and video files. The received data is sorted into the appropriate analysis devices depending on whether it is a news article or a video.

[0397] Information analysis device

[0398] The information analysis device analyzes the text of news articles. This process utilizes a generative AI model. The generative AI model is trained based on past datasets and analyzes the language patterns, context, and writing style of news articles. This allows it to determine the veracity of news articles with high accuracy and returns the results to the server as a reliability score.

[0399] For example, if a user uploads an article about a politician's remarks from their device, the server sends the article to an information analysis device, where the generative AI model analyzes the language patterns and context. As a result, it determines that the news article has low reliability and generates a reliability score.

[0400] Video analysis equipment

[0401] Video analysis devices analyze videos. In this process, the video is divided into frames, and the video and audio data of each frame is analyzed. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis uses technology to analyze voice tone and patterns. A generative AI model combines these data to determine the authenticity of the video. The results are sent back to the server as a reliability score.

[0402] For example, if a user uploads a video introducing a new product from their device, the server sends the video to a video analysis device, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the server determines that the video is likely to be fake and generates a reliability score.

[0403] Data integration device

[0404] The server receives the analysis results from the information analysis device and the video analysis device and sends them to the data integration device. This device integrates the analysis results of the news article and the video and calculates an overall reliability score. This comprehensively evaluates the authenticity of the news article and the video.

[0405] display device

[0406] The final reliability score and analysis results are sent to a display device, and the server receives them and displays them on the user's device. The user can check the results through their device, and the analysis results will be displayed as "This news article is likely to be unreliable" or "This video is likely to be fake." This will allow users to make decisions based on accurate information, without being misled by false information.

[0407] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0408] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0409] Step 1:

[0410] Users upload suspected fake news or fake videos from their own devices. They select a text file or video file of the news article and click the send button. The input is the text file or video file of the news article, and the output is that this data is sent to the server.

[0411] Specific behavior:

[0412] 1. The user opens a browser on their device and accesses the system's web page.

[0413] 2. Drag and drop your news article or video file or click the "Choose File" button to upload it.

[0414] 3. Once the upload is complete, the user clicks the "Submit" button.

[0415] Step 2:

[0416] The server receives the data sent by the user and distributes the news articles and videos to the corresponding analysis modules. The input is the text file or video file of the news article sent by the user, and the output is the distributed data.

[0417] Specific behavior:

[0418] 1. The server receives the transmitted data.

[0419] 2. Determine the type of data (text or video).

[0420] 3. Text data is sent to the information analysis device, and video data is sent to the video analysis device.

[0421] Step 3:

[0422] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The input is the news article text, and the output is a credibility score.

[0423] Specific behavior:

[0424] 1. The text of a news article is sent to an information analysis device.

[0425] 2. The generative AI model receives a prompt to analyze the news article and begins the analysis.

[0426] 3. Language patterns, context, and writing style are analyzed.

[0427] 4. A confidence score is calculated and sent back to the server.

[0428] Step 4:

[0429] The server sends the video to the fake video analysis module, which splits the video into frames and analyzes the video and audio data. The input is the video file, and the output is a confidence score.

[0430] Specific behavior:

[0431] 1. The video file is sent to the video analyzer.

[0432] 2. The video is divided into frames.

[0433] 3. The video data is analyzed using technology that detects iris movement and changes in facial expression.

[0434] 4. Audio data is analyzed for tone and patterns of voice.

[0435] 5. A generative AI model combines the video and audio data to calculate a reliability score.

[0436] 6. The confidence score is sent back to the server.

[0437] Step 5:

[0438] The server receives the analysis results from the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module. The input is the analyzed credibility score, and the output is the integrated overall credibility score.

[0439] Specific behavior:

[0440] 1. The credibility scores of the news article and the video arrive at the server.

[0441] 2. These scores are sent to a data aggregator.

[0442] 3. The data aggregator combines the credibility scores of the news article and the video to calculate an overall credibility score.

[0443] Step 6:

[0444] The final reliability score and analysis results are sent from the server to the user interface module, which receives them and displays the results on the user's terminal. The input is the overall reliability score, and the output is the display of the analysis results.

[0445] Specific behavior:

[0446] 1. The overall reliability score and analysis results are sent to the user interface module.

[0447] 2. The user interface module processes the results to display them on the user terminal.

[0448] 3. The user checks the results on their device and sees messages such as "This news article is likely to be unreliable" or "This video may be fake."

[0449] In this way, the system can achieve high accuracy in identifying fake news and fake videos and provide accurate information to users.

[0450] (Application example 1)

[0451] 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."

[0452] In recent years, a large number of news articles and videos have been circulating on the Internet, and many of them contain fake news and videos. Such misinformation can lead to incorrect perceptions and judgments among users, potentially causing social unrest. Therefore, there is a need for a method to accurately determine the authenticity of news articles and videos and provide users with accurate information. In particular, for content distribution services, it is important to have a function that can determine the reliability of the content viewed by users in real time and immediately display a reliability score.

[0453] 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.

[0454] In this invention, the server includes server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing news articles, fake video analysis means for analyzing videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and means for determining the authenticity of the news articles and videos viewed by the user in real time and displaying a reliability score. This allows users to check the reliability of the content they view in real time, enabling them to make decisions based on accurate information without being misled by misinformation.

[0455] A "user terminal" is a device operated by a user, and includes a smartphone, tablet, personal computer, and the like.

[0456] The "server means" is a central processing unit that performs analysis and data processing, and has the function of communicating with user terminals via a network.

[0457] The "fake news analysis tool" is an analytical module that determines the authenticity of news articles and has the ability to analyze language patterns, context, and writing style.

[0458] The "fake video analysis means" is an analysis module that determines the authenticity of a video, and includes means for analyzing the video on a frame-by-frame basis and detecting iris movement and changes in facial expression.

[0459] The "multimodal data integration means" has the function of integrating the analysis results obtained from the fake news analysis means and the fake video analysis means and calculating an overall reliability score.

[0460] The "user interface means" is an interface that displays the analysis results to the user, and refers to a screen or application that the user operates.

[0461] "Real-time judgment" is a process that instantly analyzes the authenticity of news articles and videos viewed by users and immediately provides the results to the users.

[0462] A "trust score" is a number generated based on the analysis results that indicates the reliability of a news article or video.

[0463] MODE FOR CARRYING OUT THE INVENTION

[0464] The present invention provides a system for analyzing news articles or videos sent from a user terminal and evaluating their reliability, which includes a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, a user interface unit, and a unit for displaying a reliability score in real time.

[0465] System Configuration

[0466] Server means:

[0467] The server distributes news articles or videos received from user terminals to the analysis module. The server functions as a central processing unit with high-performance processing capabilities, and receives and transmits data in real time via the network.

[0468] Fake news analysis methods:

[0469] The fake news analysis tool analyzes the language patterns, context, and writing style of news articles using a specific generative AI model. The generative AI model is trained on a large number of past news articles and can determine the veracity of news articles with high accuracy. The server receives the analysis results as a credibility score.

[0470] Fake video analysis methods:

[0471] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. Video analysis detects iris movement and changes in facial expressions, while audio analysis analyzes voice tone and patterns. A generative AI model integrates these data to determine the authenticity of the video. The server receives the analysis results as a reliability score.

[0472] Multimodal data integration methods:

[0473] The multimodal data integration method integrates the results obtained from fake news analysis and fake video analysis to calculate a comprehensive credibility score, which enables accurate credibility assessment that takes into account not only the judgment results of a single news article or video, but also the relationship between the two.

[0474] User Interface Methods:

[0475] The user interface means plays a role in displaying the analysis results to the user. The analysis results are displayed on the user terminal in real time, allowing the user to immediately confirm the reliability of the content.

[0476] Example of operation

[0477] When a user uploads a news article (path / to / content / news.txt) from their smartphone, the server routes the news article to the fake news analysis tool, which analyzes language patterns and context using a generative AI model. The resulting trust score is displayed in real time on the user's device, such as "News Trust Score: 85." Similarly, when a user uploads a video, the fake video analysis tool analyzes the video and audio, and displays a trust score.

[0478] Prompt Sentence Examples

[0479] An example of a prompt sentence to input to a generative AI model is, "Please rate the credibility of the following news article." By providing the text of the news article along with this prompt sentence to the generative AI model, the model analyzes the article's language patterns and context and returns a credibility score.

[0480] As described above, the system of the present invention can accurately determine the authenticity of news articles and videos viewed by users and display a reliability score in real time, allowing users to make decisions based on accurate information without being misled by misinformation.

[0481] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0482] Step 1:

[0483] The user selects a news article or video from a user device such as a smartphone and sends it to the server. The input is a text file or video file of the news article, which the server receives. Specifically, the user presses the upload button in the application, selects a file from the file selection screen, and sends it.

[0484] Step 2:

[0485] The server distributes the received data to the analysis modules. It receives news articles or video files as input, and distributes them to the fake news analysis means if they are news articles, and to the fake video analysis means if they are videos. Specifically, the server determines the type of file it receives and transfers it to the corresponding analysis module.

[0486] Step 3:

[0487] The fake news analysis method analyzes the content of news articles. It receives the text of a news article as input and uses a generative AI model to analyze the language patterns, context, and writing style. It generates a credibility score as output. Specifically, the text of a news article is input into the generative AI model, and a credibility score is returned as the analysis result.

[0488] Step 4:

[0489] The fake video analysis method analyzes videos frame by frame and simultaneously analyzes audio data. It receives a video file as input and analyzes the video and audio data for each frame. It uses a generative AI model to determine iris movement, facial expression changes, and vocal tone and patterns, and generates a credibility score as output. Specifically, the video is divided into frames, and the data for each frame is analyzed.

[0490] Step 5:

[0491] The multimodal data integration means integrates the results from the fake news analysis means and the fake video analysis means. It receives the credibility scores of news articles and videos as input and calculates an overall credibility score. It produces this overall score as output. Specifically, multiple credibility scores are input into the integration algorithm, and a final credibility score is calculated.

[0492] Step 6:

[0493] The user interface means displays the calculated trustworthiness score on the user's terminal. It receives the overall trustworthiness score from the server as input and converts it into a format to be displayed on the user's terminal. As an output, the trustworthiness score is displayed on the user's screen. Specifically, the trustworthiness score is displayed on the application screen in a format such as "News Trust Score: 85."

[0494] 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.

[0495] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, and a user interface unit. Furthermore, the present invention is equipped with an emotion engine that recognizes the user's emotions and incorporates a function to optimize the presentation of the analysis results.

[0496] Program processing

[0497] The user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0498] Fake News Analysis Module

[0499] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0500] Examples:

[0501] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0502] Fake video analysis module

[0503] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0504] Examples:

[0505] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0506] Emotion Engine

[0507] Furthermore, an emotion engine is incorporated into the user interface means to recognize the user's emotional state in real time. The emotion engine analyzes the user's facial expressions, voice tone, and input patterns to understand the user's emotional state when receiving information. This allows the method of presenting the analysis results to be optimized according to the user's emotions.

[0508] Examples:

[0509] When the emotion engine detects anxiety or discomfort from the user's facial expression or voice while the user is viewing the analysis results, the user interface means provides the analysis results in a more easily understandable format or displays additional explanations to help the user understand.

[0510] Multimodal Data Integration Module

[0511] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0512] Examples:

[0513] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0514] User Interface Module

[0515] The final confidence score and analysis results are sent to the user interface module. The server receives them and displays the results on the user's device. The emotional engine detects the user's emotional state, so the display format of the results is adjusted according to the user's emotional state.

[0516] Examples:

[0517] When the user checks the results on their device, the analysis result will display, "This news article is unreliable and is likely fake news." At the same time, the analysis result for the video will display, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0518] In this way, the system of the present invention achieves highly accurate identification of fake news and fake videos, and furthermore, by optimizing the way in which the analysis results are presented according to the user's emotional state, it is possible to provide accurate information to users and prevent them from being misled by false information.

[0519] The processing flow will be explained below.

[0520] Step 1:

[0521] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0522] Step 2:

[0523] The server receives the data sent by the user, determines whether it is a news article or a video, and then assigns it to either the fake news analysis module or the fake video analysis module depending on the determined data format.

[0524] Step 3:

[0525] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The result of this analysis is a credibility score, which is returned to the server.

[0526] Step 4:

[0527] The server sends the video to a fake video analysis module, which divides the video into frames and detects changes in iris movement and facial expressions in each frame. It also analyzes the audio data, analyzing the tone and patterns of the voice. Based on this data, the generative AI model determines whether the video is authentic, and a reliability score is generated and sent back to the server.

[0528] Step 5:

[0529] The server receives the analysis results provided by the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module, which integrates the analysis results of the news article and the video to calculate an overall credibility score.

[0530] Step 6:

[0531] The server recognizes the user's emotional state in real time through the emotion engine, which grasps the user's emotional state by analyzing the user's facial expressions, voice tone, and input patterns through the user interface means.

[0532] Step 7:

[0533] The server sends the final confidence score and analysis results to the user interface module, which presents the analysis results to the user in the most appropriate manner based on feedback from the emotion engine.

[0534] Step 8:

[0535] The user checks the analysis results on their device, which display a credibility score and supporting reasons. If the emotion engine detects a negative emotion from the user, the analysis results will provide further detailed explanations and supplementary information. For example, it may say, "This news article has low credibility and is likely fake news," providing details of the language patterns and writing style that support this.

[0536] This allows users to identify unreliable information and make decisions based on accurate information. The introduction of an emotion engine improves user understanding and makes analysis results more easily accepted.

[0537] Example 2

[0538] 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."

[0539] Conventional systems for determining the authenticity of news articles and videos are biased toward analyzing a single data mode (only articles or videos), making it difficult to integrate and analyze multimodal data. Furthermore, when users receive analysis results, they are presented in a uniform display format without taking into account their level of understanding or emotional state, which can result in insufficient information being provided to the user. Furthermore, technologies for improving the reliability of analysis results are limited. There is a need for a system that can resolve these issues, achieve high-accuracy identification of fake news and fake videos, and optimize the presentation of analysis results according to the user's emotional state.

[0540] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0541] In this invention, the server includes a processing device that analyzes text data or video data received from a user device, a false information analysis device that analyzes the text data, a video analysis device that analyzes the video data, a display device that displays the analysis results and integration results on the user device, and an emotion recognition device incorporated in the display device that recognizes the user's emotional state in real time and optimizes the presentation method of the analysis results. This enables multimodal data integration that integrates the analysis results of news articles and videos, realizing optimal information presentation according to the user's emotional state. Furthermore, high-precision analysis using a generative AI model enables reliable identification of fake news and fake videos.

[0542] A "user terminal" is a computing device operated by a user, and is a device for transmitting news articles and video data to a server.

[0543] "Text data" refers to information in text format such as news articles.

[0544] "Video data" refers to frame-by-frame video information such as news videos and other video files.

[0545] "Processor means" refers to a computing device for analyzing data received from a user terminal and is a means for identifying audio and video data.

[0546] A "disinformation analysis tool" is a device for determining the authenticity of received text data, and uses a generative AI model to analyze language patterns, context, and writing style.

[0547] The "video analysis means" is a device for determining the authenticity of received video data, analyzing each frame and detecting iris movements and changes in facial expression.

[0548] The "data integration means" is a device that integrates the analysis results obtained by the false information analysis means and the video analysis means and calculates an overall reliability score.

[0549] The "display means" is a device that displays the analysis results and the integration results on a user terminal.

[0550] The "emotion recognition means" is a device that recognizes the user's emotional state in real time and optimizes the method of presenting the analysis results.

[0551] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can determine truth or falsehood with high accuracy.

[0552] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is mainly composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion recognition means. Each means will be described in detail below.

[0553] Server means:

[0554] The server has a processing device for analyzing text data (news articles) or video data (videos) received from user terminals. The server collects data from users and distributes it to the appropriate analysis module. Specifically, text data is sent to the fake news analysis means, and video data is sent to the fake video analysis means.

[0555] Fake news analysis methods:

[0556] The server analyzes the received news articles using a disinformation analysis methodology. This analysis methodology incorporates a generative AI model that performs detailed analysis of the news article's language patterns, context, and writing style. The generative AI model is trained on a large amount of previously collected data sets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0557] Examples:

[0558] When a user uploads a news article about a politician's remarks from their device, the server sends the article to a fake news analysis tool. The generative AI model analyzes language patterns and context, determining that the news article is low in credibility and generating a credibility score.

[0559] Example prompt sentence:

[0560] Please determine whether this news article is true or false.

[0561] Fake video analysis methods:

[0562] The server analyzes the received video using a video analysis method. The video analysis method analyzes the video frame by frame, analyzing iris movement, changes in facial expressions, and voice tone and patterns. The generative AI model integrates this data to determine the authenticity of the video. The analysis result is also returned to the server as a reliability score.

[0563] Examples:

[0564] When a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis means. The video analysis means analyzes the iris movement and changes in facial expressions, and the audio analysis means analyzes the tone and patterns of the voice. As a result, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0565] Example prompt sentence:

[0566] Please tell me if this video is real or not

[0567] Multimodal data integration methods:

[0568] The server sends the analysis results from the false information analysis means and the video analysis means to the data integration means, which integrates the results of both analyses and calculates an overall reliability score, thereby matching the analysis results of the news article and the video, enabling a more comprehensive determination of whether the news article is true or false.

[0569] Examples:

[0570] When a user uploads a news article and its associated video at the same time, the server combines the analysis results of these data to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0571] Emotion recognition means:

[0572] The server is equipped with an emotion recognition system that optimizes the presentation of analysis results according to the user's emotional state. This system analyzes the user's facial expressions, tone of voice, and input patterns in real time to understand the user's emotional state when receiving information.

[0573] Examples:

[0574] If the emotion recognition means detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the analysis results will be presented in a more understandable format or additional explanations will be displayed to help the user understand.

[0575] User Interface Methods:

[0576] The server displays the final reliability score and analysis results on the user's device via a user interface. The user can check the analysis results and receive detailed explanations through the device. The emotion recognition device monitors the user's emotional state in real time and adjusts the display format accordingly.

[0577] Examples:

[0578] When the user checks the results on their device, the analysis results will display information such as "This news article is unreliable and is likely fake news." At the same time, the analysis results for the video will also display "This video is likely fake." If the emotion recognition method detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0579] Advantages of this embodiment:

[0580] The system of the present invention not only achieves high-accuracy identification of fake news and fake videos, but also provides an optimal display method that matches the user's emotional state, thereby providing users with more accurate and reliable information and preventing them from being misled by false information.

[0581] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0582] Step 1:

[0583] Users select suspected fake news articles or videos from their own devices and send them to the server. The input is the text data or video file of the news article, and the output is the data sent to the server. Users can either drag and drop files into a dedicated upload form or click the select button to select a file.

[0584] Step 2:

[0585] The server distributes data received from users to the analysis modules. The input is text data or video data sent by the user, and the output is data sent to each analysis module. The server first checks the data format and then sends text data to the fake news analysis module and video data to the fake video analysis module.

[0586] Step 3:

[0587] The server sends the received news article to the fake news analysis module. The input is the text data of the news article, and the output is a credibility score. The generative AI model analyzes the news article's language patterns, context, and writing style. Specifically, it checks for specific keywords, sentence structure, and contextual consistency within the text, and returns the analysis result to the server as a credibility score.

[0588] Step 4:

[0589] The server sends the received video to the fake video analysis module. The input is the video data, and the output is a credibility score. The analysis module divides the video into frames and analyzes the video and audio data of each frame individually. Specifically, video analysis detects iris movement and changes in facial expression, and audio analysis meticulously analyzes the tone and patterns of the voice. The generative AI model integrates this data, determines whether the video is authentic, and sends a credibility score back to the server.

[0590] Step 5:

[0591] The server sends the analysis results from the fake news analysis module and fake video analysis module to the data integration means. The input is the analysis results of the news article and video, and the output is an overall credibility score. The data integration means integrates the credibility score of the news article and the credibility score of the video to calculate an overall credibility score. The integration includes weighting each score and analyzing correlations.

[0592] Step 6:

[0593] The server uses emotion recognition means to analyze the user's emotional state in real time and optimize the display method of the analysis results. The input is the user's facial expression data and voice tone, and the output is optimized display content. The emotion recognition means analyzes the user's facial expression, voice tone, and input patterns to understand the emotional state the user is in when receiving information. The user's emotional state is monitored through a camera and microphone, and the display content is adjusted based on the analysis results.

[0594] Step 7:

[0595] The server displays the final reliability score and analysis results on the user's device via the user interface means. The input is the overall reliability score and analysis results, and the output is the results displayed on the user's device. The user can check the analysis results through their device and receive a detailed explanation. Specifically, the results display screen will display information such as "This news article is unreliable and therefore likely to be fake news," and if the emotion recognition means detects the user's anxiety, additional explanations and answers to questions will be displayed.

[0596] (Application example 2)

[0597] 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."

[0598] In modern society, fake news and fake videos have become a serious problem because many news articles and videos spread instantly via the Internet. In particular, in the advertising industry, there is a high risk that companies' trust will be damaged if advertisements containing unreliable information are delivered to consumers. The present invention aims to provide a system that can accurately identify such fake news and fake videos and check the reliability of advertisements in real time.

[0599] The identification processing 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 server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing the news articles, fake video analysis means for analyzing the videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and an emotion engine for recognizing the user's emotional state in real time and optimizing the presentation method of the analysis results. This increases the reliability of information in advertisements, reduces corporate risks due to incorrect information, and enables the creation of an environment in which users can view advertisements with peace of mind.

[0600] The "server means" refers to a device that distributes data received from user terminals to processing and analysis modules.

[0601] A "fake news analysis tool" is a device that analyzes the language patterns, context, and writing style of news articles to identify fake news.

[0602] A "generative AI model" is an artificial intelligence model that is trained on historical datasets and analyzes news articles and videos.

[0603] A "fake video analysis means" is a device that analyzes video on a frame-by-frame basis and detects iris movements and changes in facial expression.

[0604] A "multimodal data integration means" is a device that integrates the results of fake news analysis and fake video analysis and calculates an overall reliability score.

[0605] The "user interface means" refers to a device that displays the analysis results on a user terminal, allowing the user to confirm the results.

[0606] An "emotion engine" is a device that analyzes a user's facial expressions and tone of voice in real time to recognize the user's emotional state.

[0607] A "credibility score" is a numerical representation of the reliability of a news article or video based on its veracity.

[0608] This invention is a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. The system includes a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion engine.

[0609] First, the user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0610] The server means has the function of receiving news articles or videos from user terminals and distributing them to the fake news analysis means and fake video analysis means.

[0611] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of received news articles. This generative AI model is trained based on past datasets and can determine the veracity of news with high accuracy. For example, if a user uploads an article about "a politician's statement" from their device, the server sends the article to the fake news analysis method, where the generative AI model analyzes the language patterns and context. As a result, it determines that "this news article has low credibility" and generates a credibility score.

[0612] The fake video analysis method divides the received video into frames and analyzes the video and audio data of each frame. Specifically, the video analysis uses OpenCV to detect iris movements and changes in facial expressions, and the audio analysis uses LibROSA to analyze voice tone and patterns. The generative AI model integrates this data and determines the authenticity of the video. For example, if a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis method, which analyzes the video and audio. If the iris movements and changes in facial expressions are unnatural, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0613] The emotion engine recognizes the user's emotional state in real time while the user is checking the analysis results. The emotion engine understands the emotional state of the user when receiving information by analyzing the user's facial expressions and tone of voice. For example, if the emotion engine detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the user interface means will provide the analysis results in a more understandable format or display additional explanations to help the user understand.

[0614] The multimodal data integration means receives the analysis results of the fake news analysis means and the fake video analysis means from the server, and integrates these results to calculate an overall credibility score. For example, if a user simultaneously uploads a news article and its related video, the server integrates these analysis results to generate an overall credibility score. This score is a comprehensive assessment of the authenticity of the news article and the video.

[0615] The user interface means has the function of displaying the final reliability score and analysis results on the user's device. The server receives this and displays the results on the user's device. Because the emotion engine detects the user's emotional state, the display format of the results is adjusted according to the user's emotional state. For example, when the user checks the results on their device, the analysis result may say, "This news article is unreliable, so it is likely to be fake news." At the same time, the analysis result for the video may also say, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to questions may be displayed to help the user understand.

[0616] Example prompt sentence:

[0617] Please check whether the contents of the advertised article below are trustworthy.

[0618] "Politician A announced a new policy that he said would dramatically improve the economy."

[0619] Result: This advert has low credibility. A similar pattern was found in many fake news stories. Credibility score: 30%

[0620] In this way, the system of the present invention achieves high-accuracy identification of fake news and fake videos, and further optimizes the presentation method of the analysis results according to the user's emotional state, thereby providing users with accurate and safe information and increasing the reliability of information in advertisements.

[0621] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0622] Step 1:

[0623] A user uploads a news article or video from a terminal. The input data includes a news article text file or video file. The terminal uploads this data and sends it to the server. The server receives this data.

[0624] Step 2:

[0625] The server distributes the received data to the analysis modules. If a news article is received, the data is sent to the fake news analysis means, and if a video is received, the data is sent to the fake video analysis means. The input data is a news article or a video file. The output is the data to be processed by each analysis module.

[0626] Step 3:

[0627] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of news articles. The text of the news article is used as input, and the generative AI model performs data analysis. The output is a credibility score based on the analyzed language patterns, context, and writing style.

[0628] Step 4:

[0629] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. The video file is used as input. Specifically, OpenCV is used to detect iris movement and changes in facial expressions, and LibROSA is used to analyze voice tone and patterns. The output is a reliability score based on the video and audio analysis results.

[0630] Step 5:

[0631] The server transmits the analysis results from the fake news analysis means and the fake video analysis means to the multimodal data integration means. The input data are reliability scores. The multimodal data integration means calculates an overall reliability score based on these results. The output is an integrated overall reliability score.

[0632] Step 6:

[0633] The user interface means displays the received overall reliability score on the user terminal. The input data is the reliability score. When the user checks the results, the emotion engine analyzes the user's emotional state in real time and optimizes the display format. Specifically, it analyzes the user's facial expressions and tone of voice and selects a method for presenting the results according to their emotional state. The output is a display of the optimized reliability score.

[0634] Step 7:

[0635] The user checks the results on their device. The input data is the analyzed confidence score and its detailed explanation. If the emotion engine detects anxiety or discomfort while the user is viewing the analysis results, it provides additional explanations and details to help the user understand. The output is a display of the detailed analysis results according to the user's emotional state.

[0636] 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.

[0637] 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.

[0638] 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.

[0639] [Third embodiment]

[0640] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0641] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0642] 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).

[0643] 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.

[0644] 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.

[0645] 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).

[0646] 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.

[0647] 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.

[0648] 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.

[0649] 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.

[0650] 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.

[0651] 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."

[0652] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, and a user interface means.

[0653] Program processing

[0654] The user sends information suspected to be fake news or fake video from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0655] Fake News Analysis Module

[0656] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0657] Examples:

[0658] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0659] Fake video analysis module

[0660] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0661] Examples:

[0662] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0663] Multimodal Data Integration Module

[0664] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0665] Examples:

[0666] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0667] User Interface Module

[0668] The final confidence score and analysis results are sent to the user interface module, which receives them and displays them on the user's terminal.

[0669] Examples:

[0670] When the user checks the results on their device, the analysis result will show "This news article is likely to be unreliable." At the same time, the analysis result for the video will show "This video is likely fake." This will help users make decisions based on accurate information, without being misled by false information.

[0671] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0672] The processing flow will be explained below.

[0673] Step 1:

[0674] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0675] Step 2:

[0676] The server distributes the data received from users to either the fake news analysis module or the fake video analysis module, verifying the data format and content and performing appropriate preprocessing (e.g., text cleaning and video frame segmentation).

[0677] Step 3:

[0678] The server sends news articles to a fake news analysis module, which uses a generative AI model to analyze the article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0679] Step 4:

[0680] The server sends the video to a fake video analysis module, which analyzes the video and audio data separately. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis analyzes the tone and patterns of the voice. A generative AI model combines these data to determine whether the video is authentic or fake.

[0681] Step 5:

[0682] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0683] Step 6:

[0684] The server sends the final reliability score and analysis results to a user interface module, which displays the results on a user interface and provides the results to the user terminal in a visually easy-to-understand format.

[0685] Step 7:

[0686] Users can check the analysis results displayed on their device and make decisions based on them. The analysis results are presented with a specific reliability score and the basis for the analysis, allowing users to refer to this information and obtain accurate information without being misled by incorrect information.

[0687] Example 1

[0688] 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."

[0689] Today, there are many news articles and videos on the Internet, some of which contain false information such as fake news and videos. This makes it difficult for users to access accurate information, which can have a negative impact on decision-making. In particular, in the case of political and economic news, which have a great social impact, the spread of false information increases the risk of causing confusion and anxiety. Therefore, a system is needed that allows users to easily determine the authenticity of information on the Internet.

[0690] 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.

[0691] In this invention, the server includes an information processing device that analyzes news articles or videos received from a user terminal, an information analysis device that analyzes the news articles, a video analysis device that analyzes the videos, a data integration device that integrates the analysis results of the news articles and videos, and a display device that displays the integration results on the user terminal, thereby enabling users to determine the authenticity of the news articles and videos with high accuracy and make decisions based on accurate information.

[0692] A "user terminal" is a device through which a user enters information and communicates with the system.

[0693] An "information processing device" is a device that has the function of distributing news articles and videos received from user terminals to analysis modules.

[0694] An "information analysis device" is a device that analyzes received news articles using language patterns, context, and writing style.

[0695] A "video analysis device" is a device that analyzes the video and audio of received video on a frame-by-frame basis to determine its authenticity.

[0696] The "data integration device" is a device that integrates the analysis results of news articles and videos and calculates an overall reliability score.

[0697] The "display device" is a device that displays the analysis results and integration results on a user terminal.

[0698] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can accurately determine the authenticity of news articles and videos.

[0699] A "prompt statement" is an instruction statement that causes the generative AI model to perform analysis.

[0700] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of an information processing device, an information analysis device, a video analysis device, a data integration device, and a display device.

[0701] Information processing device

[0702] The server receives news articles and videos sent by users from their devices. Users have the ability to upload news article text files and video files. The received data is sorted into the appropriate analysis devices depending on whether it is a news article or a video.

[0703] Information analysis device

[0704] The information analysis device analyzes the text of news articles. This process utilizes a generative AI model. The generative AI model is trained based on past datasets and analyzes the language patterns, context, and writing style of news articles. This allows it to determine the veracity of news articles with high accuracy and returns the results to the server as a reliability score.

[0705] For example, if a user uploads an article about a politician's remarks from their device, the server sends the article to an information analysis device, where the generative AI model analyzes the language patterns and context. As a result, it determines that the news article has low reliability and generates a reliability score.

[0706] Video analysis equipment

[0707] Video analysis devices analyze videos. In this process, the video is divided into frames, and the video and audio data of each frame is analyzed. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis uses technology to analyze voice tone and patterns. A generative AI model combines these data to determine the authenticity of the video. The results are sent back to the server as a reliability score.

[0708] For example, if a user uploads a video introducing a new product from their device, the server sends the video to a video analysis device, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the server determines that the video is likely to be fake and generates a reliability score.

[0709] Data integration device

[0710] The server receives the analysis results from the information analysis device and the video analysis device and sends them to the data integration device. This device integrates the analysis results of the news article and the video and calculates an overall reliability score. This comprehensively evaluates the authenticity of the news article and the video.

[0711] display device

[0712] The final reliability score and analysis results are sent to a display device, and the server receives them and displays them on the user's device. The user can check the results through their device, and the analysis results will be displayed as "This news article is likely to be unreliable" or "This video is likely to be fake." This will allow users to make decisions based on accurate information, without being misled by false information.

[0713] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0714] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0715] Step 1:

[0716] Users upload suspected fake news or fake videos from their own devices. They select a text file or video file of the news article and click the send button. The input is the text file or video file of the news article, and the output is that this data is sent to the server.

[0717] Specific behavior:

[0718] 1. The user opens a browser on their device and accesses the system's web page.

[0719] 2. Drag and drop your news article or video file or click the "Choose File" button to upload it.

[0720] 3. Once the upload is complete, the user clicks the "Submit" button.

[0721] Step 2:

[0722] The server receives the data sent by the user and distributes the news articles and videos to the corresponding analysis modules. The input is the text file or video file of the news article sent by the user, and the output is the distributed data.

[0723] Specific behavior:

[0724] 1. The server receives the transmitted data.

[0725] 2. Determine the type of data (text or video).

[0726] 3. Text data is sent to the information analysis device, and video data is sent to the video analysis device.

[0727] Step 3:

[0728] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The input is the news article text, and the output is a credibility score.

[0729] Specific behavior:

[0730] 1. The text of a news article is sent to an information analysis device.

[0731] 2. The generative AI model receives a prompt to analyze the news article and begins the analysis.

[0732] 3. Language patterns, context, and writing style are analyzed.

[0733] 4. A confidence score is calculated and sent back to the server.

[0734] Step 4:

[0735] The server sends the video to the fake video analysis module, which splits the video into frames and analyzes the video and audio data. The input is the video file, and the output is a confidence score.

[0736] Specific behavior:

[0737] 1. The video file is sent to the video analyzer.

[0738] 2. The video is divided into frames.

[0739] 3. The video data is analyzed using technology that detects iris movement and changes in facial expression.

[0740] 4. Audio data is analyzed for tone and patterns of voice.

[0741] 5. A generative AI model combines the video and audio data to calculate a reliability score.

[0742] 6. The confidence score is sent back to the server.

[0743] Step 5:

[0744] The server receives the analysis results from the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module. The input is the analyzed credibility score, and the output is the integrated overall credibility score.

[0745] Specific behavior:

[0746] 1. The credibility scores of the news article and the video arrive at the server.

[0747] 2. These scores are sent to a data aggregator.

[0748] 3. The data aggregator combines the credibility scores of the news article and the video to calculate an overall credibility score.

[0749] Step 6:

[0750] The final reliability score and analysis results are sent from the server to the user interface module, which receives them and displays the results on the user's terminal. The input is the overall reliability score, and the output is the display of the analysis results.

[0751] Specific behavior:

[0752] 1. The overall reliability score and analysis results are sent to the user interface module.

[0753] 2. The user interface module processes the results to display them on the user terminal.

[0754] 3. The user checks the results on their device and sees messages such as "This news article is likely to be unreliable" or "This video may be fake."

[0755] In this way, the system can achieve high accuracy in identifying fake news and fake videos and provide accurate information to users.

[0756] (Application example 1)

[0757] 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."

[0758] In recent years, a large number of news articles and videos have been circulating on the Internet, and many of them contain fake news and videos. Such misinformation can lead to incorrect perceptions and judgments among users, potentially causing social unrest. Therefore, there is a need for a method to accurately determine the authenticity of news articles and videos and provide users with accurate information. In particular, for content distribution services, it is important to have a function that can determine the reliability of the content viewed by users in real time and immediately display a reliability score.

[0759] 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.

[0760] In this invention, the server includes server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing news articles, fake video analysis means for analyzing videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and means for determining the authenticity of the news articles and videos viewed by the user in real time and displaying a reliability score. This allows users to check the reliability of the content they view in real time, enabling them to make decisions based on accurate information without being misled by misinformation.

[0761] A "user terminal" is a device operated by a user, and includes a smartphone, tablet, personal computer, and the like.

[0762] The "server means" is a central processing unit that performs analysis and data processing, and has the function of communicating with user terminals via a network.

[0763] The "fake news analysis tool" is an analytical module that determines the authenticity of news articles and has the ability to analyze language patterns, context, and writing style.

[0764] The "fake video analysis means" is an analysis module that determines the authenticity of a video, and includes means for analyzing the video on a frame-by-frame basis and detecting iris movement and changes in facial expression.

[0765] The "multimodal data integration means" has the function of integrating the analysis results obtained from the fake news analysis means and the fake video analysis means and calculating an overall reliability score.

[0766] The "user interface means" is an interface that displays the analysis results to the user, and refers to a screen or application that the user operates.

[0767] "Real-time judgment" is a process that instantly analyzes the authenticity of news articles and videos viewed by users and immediately provides the results to the users.

[0768] A "trust score" is a number generated based on the analysis results that indicates the reliability of a news article or video.

[0769] MODE FOR CARRYING OUT THE INVENTION

[0770] The present invention provides a system for analyzing news articles or videos sent from a user terminal and evaluating their reliability, which includes a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, a user interface unit, and a unit for displaying a reliability score in real time.

[0771] System Configuration

[0772] Server means:

[0773] The server distributes news articles or videos received from user terminals to the analysis module. The server functions as a central processing unit with high-performance processing capabilities, and receives and transmits data in real time via the network.

[0774] Fake news analysis methods:

[0775] The fake news analysis tool analyzes the language patterns, context, and writing style of news articles using a specific generative AI model. The generative AI model is trained on a large number of past news articles and can determine the veracity of news articles with high accuracy. The server receives the analysis results as a credibility score.

[0776] Fake video analysis methods:

[0777] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. Video analysis detects iris movement and changes in facial expressions, while audio analysis analyzes voice tone and patterns. A generative AI model integrates these data to determine the authenticity of the video. The server receives the analysis results as a reliability score.

[0778] Multimodal data integration methods:

[0779] The multimodal data integration method integrates the results obtained from fake news analysis and fake video analysis to calculate a comprehensive credibility score, which enables accurate credibility assessment that takes into account not only the judgment results of a single news article or video, but also the relationship between the two.

[0780] User Interface Methods:

[0781] The user interface means plays a role in displaying the analysis results to the user. The analysis results are displayed on the user terminal in real time, allowing the user to immediately confirm the reliability of the content.

[0782] Example of operation

[0783] When a user uploads a news article (path / to / content / news.txt) from their smartphone, the server routes the news article to the fake news analysis tool, which analyzes language patterns and context using a generative AI model. The resulting trust score is displayed in real time on the user's device, such as "News Trust Score: 85." Similarly, when a user uploads a video, the fake video analysis tool analyzes the video and audio, and displays a trust score.

[0784] Prompt Sentence Examples

[0785] An example of a prompt sentence to input to a generative AI model is, "Please rate the credibility of the following news article." By providing the text of the news article along with this prompt sentence to the generative AI model, the model analyzes the article's language patterns and context and returns a credibility score.

[0786] As described above, the system of the present invention can accurately determine the authenticity of news articles and videos viewed by users and display a reliability score in real time, allowing users to make decisions based on accurate information without being misled by misinformation.

[0787] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0788] Step 1:

[0789] The user selects a news article or video from a user device such as a smartphone and sends it to the server. The input is a text file or video file of the news article, which the server receives. Specifically, the user presses the upload button in the application, selects a file from the file selection screen, and sends it.

[0790] Step 2:

[0791] The server distributes the received data to the analysis modules. It receives news articles or video files as input, and distributes them to the fake news analysis means if they are news articles, and to the fake video analysis means if they are videos. Specifically, the server determines the type of file it receives and transfers it to the corresponding analysis module.

[0792] Step 3:

[0793] The fake news analysis method analyzes the content of news articles. It receives the text of a news article as input and uses a generative AI model to analyze the language patterns, context, and writing style. It generates a credibility score as output. Specifically, the text of a news article is input into the generative AI model, and a credibility score is returned as the analysis result.

[0794] Step 4:

[0795] The fake video analysis method analyzes videos frame by frame and simultaneously analyzes audio data. It receives a video file as input and analyzes the video and audio data for each frame. It uses a generative AI model to determine iris movement, facial expression changes, and vocal tone and patterns, and generates a credibility score as output. Specifically, the video is divided into frames, and the data for each frame is analyzed.

[0796] Step 5:

[0797] The multimodal data integration means integrates the results from the fake news analysis means and the fake video analysis means. It receives the credibility scores of news articles and videos as input and calculates an overall credibility score. It produces this overall score as output. Specifically, multiple credibility scores are input into the integration algorithm, and a final credibility score is calculated.

[0798] Step 6:

[0799] The user interface means displays the calculated trustworthiness score on the user's terminal. It receives the overall trustworthiness score from the server as input and converts it into a format to be displayed on the user's terminal. As an output, the trustworthiness score is displayed on the user's screen. Specifically, the trustworthiness score is displayed on the application screen in a format such as "News Trust Score: 85."

[0800] 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.

[0801] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, and a user interface unit. Furthermore, the present invention is equipped with an emotion engine that recognizes the user's emotions and incorporates a function to optimize the presentation of the analysis results.

[0802] Program processing

[0803] The user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0804] Fake News Analysis Module

[0805] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0806] Examples:

[0807] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0808] Fake video analysis module

[0809] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0810] Examples:

[0811] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0812] Emotion Engine

[0813] Furthermore, an emotion engine is incorporated into the user interface means to recognize the user's emotional state in real time. The emotion engine analyzes the user's facial expressions, voice tone, and input patterns to understand the user's emotional state when receiving information. This allows the method of presenting the analysis results to be optimized according to the user's emotions.

[0814] Examples:

[0815] When the emotion engine detects anxiety or discomfort from the user's facial expression or voice while the user is viewing the analysis results, the user interface means provides the analysis results in a more easily understandable format or displays additional explanations to help the user understand.

[0816] Multimodal Data Integration Module

[0817] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0818] Examples:

[0819] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0820] User Interface Module

[0821] The final confidence score and analysis results are sent to the user interface module. The server receives them and displays the results on the user's device. The emotional engine detects the user's emotional state, so the display format of the results is adjusted according to the user's emotional state.

[0822] Examples:

[0823] When the user checks the results on their device, the analysis result will display, "This news article is unreliable and is likely fake news." At the same time, the analysis result for the video will display, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0824] In this way, the system of the present invention achieves highly accurate identification of fake news and fake videos, and furthermore, by optimizing the way in which the analysis results are presented according to the user's emotional state, it is possible to provide accurate information to users and prevent them from being misled by false information.

[0825] The processing flow will be explained below.

[0826] Step 1:

[0827] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0828] Step 2:

[0829] The server receives the data sent by the user, determines whether it is a news article or a video, and then assigns it to either the fake news analysis module or the fake video analysis module depending on the determined data format.

[0830] Step 3:

[0831] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The result of this analysis is a credibility score, which is returned to the server.

[0832] Step 4:

[0833] The server sends the video to a fake video analysis module, which divides the video into frames and detects changes in iris movement and facial expressions in each frame. It also analyzes the audio data, analyzing the tone and patterns of the voice. Based on this data, the generative AI model determines whether the video is authentic, and a reliability score is generated and sent back to the server.

[0834] Step 5:

[0835] The server receives the analysis results provided by the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module, which integrates the analysis results of the news article and the video to calculate an overall credibility score.

[0836] Step 6:

[0837] The server recognizes the user's emotional state in real time through the emotion engine, which grasps the user's emotional state by analyzing the user's facial expressions, voice tone, and input patterns through the user interface means.

[0838] Step 7:

[0839] The server sends the final confidence score and analysis results to the user interface module, which presents the analysis results to the user in the most appropriate manner based on feedback from the emotion engine.

[0840] Step 8:

[0841] The user checks the analysis results on their device, which display a credibility score and supporting reasons. If the emotion engine detects a negative emotion from the user, the analysis results will provide further detailed explanations and supplementary information. For example, it may say, "This news article has low credibility and is likely fake news," providing details of the language patterns and writing style that support this.

[0842] This allows users to identify unreliable information and make decisions based on accurate information. The introduction of an emotion engine improves user understanding and makes analysis results more easily accepted.

[0843] Example 2

[0844] 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."

[0845] Conventional systems for determining the authenticity of news articles and videos are biased toward analyzing a single data mode (only articles or videos), making it difficult to integrate and analyze multimodal data. Furthermore, when users receive analysis results, they are presented in a uniform display format without taking into account their level of understanding or emotional state, which can result in insufficient information being provided to the user. Furthermore, technologies for improving the reliability of analysis results are limited. There is a need for a system that can resolve these issues, achieve high-accuracy identification of fake news and fake videos, and optimize the presentation of analysis results according to the user's emotional state.

[0846] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0847] In this invention, the server includes a processing device that analyzes text data or video data received from a user device, a false information analysis device that analyzes the text data, a video analysis device that analyzes the video data, a display device that displays the analysis results and integration results on the user device, and an emotion recognition device incorporated in the display device that recognizes the user's emotional state in real time and optimizes the presentation method of the analysis results. This enables multimodal data integration that integrates the analysis results of news articles and videos, realizing optimal information presentation according to the user's emotional state. Furthermore, high-precision analysis using a generative AI model enables reliable identification of fake news and fake videos.

[0848] A "user terminal" is a computing device operated by a user, and is a device for transmitting news articles and video data to a server.

[0849] "Text data" refers to information in text format such as news articles.

[0850] "Video data" refers to frame-by-frame video information such as news videos and other video files.

[0851] "Processor means" refers to a computing device for analyzing data received from a user terminal and is a means for identifying audio and video data.

[0852] A "disinformation analysis tool" is a device for determining the authenticity of received text data, and uses a generative AI model to analyze language patterns, context, and writing style.

[0853] The "video analysis means" is a device for determining the authenticity of received video data, analyzing each frame and detecting iris movements and changes in facial expression.

[0854] The "data integration means" is a device that integrates the analysis results obtained by the false information analysis means and the video analysis means and calculates an overall reliability score.

[0855] The "display means" is a device that displays the analysis results and the integration results on a user terminal.

[0856] The "emotion recognition means" is a device that recognizes the user's emotional state in real time and optimizes the method of presenting the analysis results.

[0857] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can determine truth or falsehood with high accuracy.

[0858] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is mainly composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion recognition means. Each means will be described in detail below.

[0859] Server means:

[0860] The server has a processing device for analyzing text data (news articles) or video data (videos) received from user terminals. The server collects data from users and distributes it to the appropriate analysis module. Specifically, text data is sent to the fake news analysis means, and video data is sent to the fake video analysis means.

[0861] Fake news analysis methods:

[0862] The server analyzes the received news articles using a disinformation analysis methodology. This analysis methodology incorporates a generative AI model that performs detailed analysis of the news article's language patterns, context, and writing style. The generative AI model is trained on a large amount of previously collected data sets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0863] Examples:

[0864] When a user uploads a news article about a politician's remarks from their device, the server sends the article to a fake news analysis tool. The generative AI model analyzes language patterns and context, determining that the news article is low in credibility and generating a credibility score.

[0865] Example prompt sentence:

[0866] Please determine whether this news article is true or false.

[0867] Fake video analysis methods:

[0868] The server analyzes the received video using a video analysis method. The video analysis method analyzes the video frame by frame, analyzing iris movement, changes in facial expressions, and voice tone and patterns. The generative AI model integrates this data to determine the authenticity of the video. The analysis result is also returned to the server as a reliability score.

[0869] Examples:

[0870] When a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis means. The video analysis means analyzes the iris movement and changes in facial expressions, and the audio analysis means analyzes the tone and patterns of the voice. As a result, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0871] Example prompt sentence:

[0872] Please tell me if this video is real or not

[0873] Multimodal data integration methods:

[0874] The server sends the analysis results from the false information analysis means and the video analysis means to the data integration means, which integrates the results of both analyses and calculates an overall reliability score, thereby matching the analysis results of the news article and the video, enabling a more comprehensive determination of whether the news article is true or false.

[0875] Examples:

[0876] When a user uploads a news article and its associated video at the same time, the server combines the analysis results of these data to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0877] Emotion recognition means:

[0878] The server is equipped with an emotion recognition system that optimizes the presentation of analysis results according to the user's emotional state. This system analyzes the user's facial expressions, tone of voice, and input patterns in real time to understand the user's emotional state when receiving information.

[0879] Examples:

[0880] If the emotion recognition means detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the analysis results will be presented in a more understandable format or additional explanations will be displayed to help the user understand.

[0881] User Interface Methods:

[0882] The server displays the final reliability score and analysis results on the user's device via a user interface. The user can check the analysis results and receive detailed explanations through the device. The emotion recognition device monitors the user's emotional state in real time and adjusts the display format accordingly.

[0883] Examples:

[0884] When the user checks the results on their device, the analysis results will display information such as "This news article is unreliable and is likely fake news." At the same time, the analysis results for the video will also display "This video is likely fake." If the emotion recognition method detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[0885] Advantages of this embodiment:

[0886] The system of the present invention not only achieves high-accuracy identification of fake news and fake videos, but also provides an optimal display method that matches the user's emotional state, thereby providing users with more accurate and reliable information and preventing them from being misled by false information.

[0887] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0888] Step 1:

[0889] Users select suspected fake news articles or videos from their own devices and send them to the server. The input is the text data or video file of the news article, and the output is the data sent to the server. Users can either drag and drop files into a dedicated upload form or click the select button to select a file.

[0890] Step 2:

[0891] The server distributes data received from users to the analysis modules. The input is text data or video data sent by the user, and the output is data sent to each analysis module. The server first checks the data format and then sends text data to the fake news analysis module and video data to the fake video analysis module.

[0892] Step 3:

[0893] The server sends the received news article to the fake news analysis module. The input is the text data of the news article, and the output is a credibility score. The generative AI model analyzes the news article's language patterns, context, and writing style. Specifically, it checks for specific keywords, sentence structure, and contextual consistency within the text, and returns the analysis result to the server as a credibility score.

[0894] Step 4:

[0895] The server sends the received video to the fake video analysis module. The input is the video data, and the output is a credibility score. The analysis module divides the video into frames and analyzes the video and audio data of each frame individually. Specifically, video analysis detects iris movement and changes in facial expression, and audio analysis meticulously analyzes the tone and patterns of the voice. The generative AI model integrates this data, determines whether the video is authentic, and sends a credibility score back to the server.

[0896] Step 5:

[0897] The server sends the analysis results from the fake news analysis module and fake video analysis module to the data integration means. The input is the analysis results of the news article and video, and the output is an overall credibility score. The data integration means integrates the credibility score of the news article and the credibility score of the video to calculate an overall credibility score. The integration includes weighting each score and analyzing correlations.

[0898] Step 6:

[0899] The server uses emotion recognition means to analyze the user's emotional state in real time and optimize the display method of the analysis results. The input is the user's facial expression data and voice tone, and the output is optimized display content. The emotion recognition means analyzes the user's facial expression, voice tone, and input patterns to understand the emotional state the user is in when receiving information. The user's emotional state is monitored through a camera and microphone, and the display content is adjusted based on the analysis results.

[0900] Step 7:

[0901] The server displays the final reliability score and analysis results on the user's device via the user interface means. The input is the overall reliability score and analysis results, and the output is the results displayed on the user's device. The user can check the analysis results through their device and receive a detailed explanation. Specifically, the results display screen will display information such as "This news article is unreliable and therefore likely to be fake news," and if the emotion recognition means detects the user's anxiety, additional explanations and answers to questions will be displayed.

[0902] (Application example 2)

[0903] 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."

[0904] In modern society, fake news and fake videos have become a serious problem because many news articles and videos spread instantly via the Internet. In particular, in the advertising industry, there is a high risk that companies' trust will be damaged if advertisements containing unreliable information are delivered to consumers. The present invention aims to provide a system that can accurately identify such fake news and fake videos and check the reliability of advertisements in real time.

[0905] The identification processing 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 server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing the news articles, fake video analysis means for analyzing the videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and an emotion engine for recognizing the user's emotional state in real time and optimizing the presentation method of the analysis results. This increases the reliability of information in advertisements, reduces corporate risks due to incorrect information, and enables the creation of an environment in which users can view advertisements with peace of mind.

[0906] The "server means" refers to a device that distributes data received from user terminals to processing and analysis modules.

[0907] A "fake news analysis tool" is a device that analyzes the language patterns, context, and writing style of news articles to identify fake news.

[0908] A "generative AI model" is an artificial intelligence model that is trained on historical datasets and analyzes news articles and videos.

[0909] A "fake video analysis means" is a device that analyzes video on a frame-by-frame basis and detects iris movements and changes in facial expression.

[0910] A "multimodal data integration means" is a device that integrates the results of fake news analysis and fake video analysis and calculates an overall reliability score.

[0911] The "user interface means" refers to a device that displays the analysis results on a user terminal, allowing the user to confirm the results.

[0912] An "emotion engine" is a device that analyzes a user's facial expressions and tone of voice in real time to recognize the user's emotional state.

[0913] A "credibility score" is a numerical representation of the reliability of a news article or video based on its veracity.

[0914] This invention is a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. The system includes a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion engine.

[0915] First, the user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0916] The server means has the function of receiving news articles or videos from user terminals and distributing them to the fake news analysis means and fake video analysis means.

[0917] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of received news articles. This generative AI model is trained based on past datasets and can determine the veracity of news with high accuracy. For example, if a user uploads an article about "a politician's statement" from their device, the server sends the article to the fake news analysis method, where the generative AI model analyzes the language patterns and context. As a result, it determines that "this news article has low credibility" and generates a credibility score.

[0918] The fake video analysis method divides the received video into frames and analyzes the video and audio data of each frame. Specifically, the video analysis uses OpenCV to detect iris movements and changes in facial expressions, and the audio analysis uses LibROSA to analyze voice tone and patterns. The generative AI model integrates this data and determines the authenticity of the video. For example, if a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis method, which analyzes the video and audio. If the iris movements and changes in facial expressions are unnatural, it is determined that "this video is likely to be fake," and a reliability score is generated.

[0919] The emotion engine recognizes the user's emotional state in real time while the user is checking the analysis results. The emotion engine understands the emotional state of the user when receiving information by analyzing the user's facial expressions and tone of voice. For example, if the emotion engine detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the user interface means will provide the analysis results in a more understandable format or display additional explanations to help the user understand.

[0920] The multimodal data integration means receives the analysis results of the fake news analysis means and the fake video analysis means from the server, and integrates these results to calculate an overall credibility score. For example, if a user simultaneously uploads a news article and its related video, the server integrates these analysis results to generate an overall credibility score. This score is a comprehensive assessment of the authenticity of the news article and the video.

[0921] The user interface means has the function of displaying the final reliability score and analysis results on the user's device. The server receives this and displays the results on the user's device. Because the emotion engine detects the user's emotional state, the display format of the results is adjusted according to the user's emotional state. For example, when the user checks the results on their device, the analysis result may say, "This news article is unreliable, so it is likely to be fake news." At the same time, the analysis result for the video may also say, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to questions may be displayed to help the user understand.

[0922] Example prompt sentence:

[0923] Please check whether the contents of the advertised article below are trustworthy.

[0924] "Politician A announced a new policy that he said would dramatically improve the economy."

[0925] Result: This advert has low credibility. A similar pattern was found in many fake news stories. Credibility score: 30%

[0926] In this way, the system of the present invention achieves high-accuracy identification of fake news and fake videos, and further optimizes the presentation method of the analysis results according to the user's emotional state, thereby providing users with accurate and safe information and increasing the reliability of information in advertisements.

[0927] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0928] Step 1:

[0929] A user uploads a news article or video from a terminal. The input data includes a news article text file or video file. The terminal uploads this data and sends it to the server. The server receives this data.

[0930] Step 2:

[0931] The server distributes the received data to the analysis modules. If a news article is received, the data is sent to the fake news analysis means, and if a video is received, the data is sent to the fake video analysis means. The input data is a news article or a video file. The output is the data to be processed by each analysis module.

[0932] Step 3:

[0933] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of news articles. The text of the news article is used as input, and the generative AI model performs data analysis. The output is a credibility score based on the analyzed language patterns, context, and writing style.

[0934] Step 4:

[0935] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. The video file is used as input. Specifically, OpenCV is used to detect iris movement and changes in facial expressions, and LibROSA is used to analyze voice tone and patterns. The output is a reliability score based on the video and audio analysis results.

[0936] Step 5:

[0937] The server transmits the analysis results from the fake news analysis means and the fake video analysis means to the multimodal data integration means. The input data are reliability scores. The multimodal data integration means calculates an overall reliability score based on these results. The output is an integrated overall reliability score.

[0938] Step 6:

[0939] The user interface means displays the received overall reliability score on the user terminal. The input data is the reliability score. When the user checks the results, the emotion engine analyzes the user's emotional state in real time and optimizes the display format. Specifically, it analyzes the user's facial expressions and tone of voice and selects a method for presenting the results according to their emotional state. The output is a display of the optimized reliability score.

[0940] Step 7:

[0941] The user checks the results on their device. The input data is the analyzed confidence score and its detailed explanation. If the emotion engine detects anxiety or discomfort while the user is viewing the analysis results, it provides additional explanations and details to help the user understand. The output is a display of the detailed analysis results according to the user's emotional state.

[0942] 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.

[0943] 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.

[0944] 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.

[0945] [Fourth embodiment]

[0946] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0947] 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.

[0948] 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).

[0949] 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.

[0950] 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.

[0951] 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).

[0952] 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.

[0953] 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.

[0954] 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.

[0955] 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.

[0956] 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.

[0957] 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.

[0958] 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."

[0959] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, and a user interface means.

[0960] Program processing

[0961] The user sends information suspected to be fake news or fake video from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[0962] Fake News Analysis Module

[0963] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0964] Examples:

[0965] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[0966] Fake video analysis module

[0967] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[0968] Examples:

[0969] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[0970] Multimodal Data Integration Module

[0971] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0972] Examples:

[0973] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[0974] User Interface Module

[0975] The final confidence score and analysis results are sent to the user interface module, which receives them and displays them on the user's terminal.

[0976] Examples:

[0977] When the user checks the results on their device, the analysis result will show "This news article is likely to be unreliable." At the same time, the analysis result for the video will show "This video is likely fake." This will help users make decisions based on accurate information, without being misled by false information.

[0978] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[0979] The processing flow will be explained below.

[0980] Step 1:

[0981] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[0982] Step 2:

[0983] The server distributes the data received from users to either the fake news analysis module or the fake video analysis module, verifying the data format and content and performing appropriate preprocessing (e.g., text cleaning and video frame segmentation).

[0984] Step 3:

[0985] The server sends news articles to a fake news analysis module, which uses a generative AI model to analyze the article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[0986] Step 4:

[0987] The server sends the video to a fake video analysis module, which analyzes the video and audio data separately. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis analyzes the tone and patterns of the voice. A generative AI model combines these data to determine whether the video is authentic or fake.

[0988] Step 5:

[0989] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[0990] Step 6:

[0991] The server sends the final reliability score and analysis results to a user interface module, which displays the results on a user interface and provides the results to the user terminal in a visually easy-to-understand format.

[0992] Step 7:

[0993] Users can check the analysis results displayed on their device and make decisions based on them. The analysis results are presented with a specific reliability score and the basis for the analysis, allowing users to refer to this information and obtain accurate information without being misled by incorrect information.

[0994] Example 1

[0995] 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."

[0996] Today, there are many news articles and videos on the Internet, some of which contain false information such as fake news and videos. This makes it difficult for users to access accurate information, which can have a negative impact on decision-making. In particular, in the case of political and economic news, which have a great social impact, the spread of false information increases the risk of causing confusion and anxiety. Therefore, a system is needed that allows users to easily determine the authenticity of information on the Internet.

[0997] 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.

[0998] In this invention, the server includes an information processing device that analyzes news articles or videos received from a user terminal, an information analysis device that analyzes the news articles, a video analysis device that analyzes the videos, a data integration device that integrates the analysis results of the news articles and videos, and a display device that displays the integration results on the user terminal, thereby enabling users to determine the authenticity of the news articles and videos with high accuracy and make decisions based on accurate information.

[0999] A "user terminal" is a device through which a user enters information and communicates with the system.

[1000] An "information processing device" is a device that has the function of distributing news articles and videos received from user terminals to analysis modules.

[1001] An "information analysis device" is a device that analyzes received news articles using language patterns, context, and writing style.

[1002] A "video analysis device" is a device that analyzes the video and audio of received video on a frame-by-frame basis to determine its authenticity.

[1003] The "data integration device" is a device that integrates the analysis results of news articles and videos and calculates an overall reliability score.

[1004] The "display device" is a device that displays the analysis results and integration results on a user terminal.

[1005] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can accurately determine the authenticity of news articles and videos.

[1006] A "prompt statement" is an instruction statement that causes the generative AI model to perform analysis.

[1007] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of an information processing device, an information analysis device, a video analysis device, a data integration device, and a display device.

[1008] Information processing device

[1009] The server receives news articles and videos sent by users from their devices. Users have the ability to upload news article text files and video files. The received data is sorted into the appropriate analysis devices depending on whether it is a news article or a video.

[1010] Information analysis device

[1011] The information analysis device analyzes the text of news articles. This process utilizes a generative AI model. The generative AI model is trained based on past datasets and analyzes the language patterns, context, and writing style of news articles. This allows it to determine the veracity of news articles with high accuracy and returns the results to the server as a reliability score.

[1012] For example, if a user uploads an article about a politician's remarks from their device, the server sends the article to an information analysis device, where the generative AI model analyzes the language patterns and context. As a result, it determines that the news article has low reliability and generates a reliability score.

[1013] Video analysis equipment

[1014] Video analysis devices analyze videos. In this process, the video is divided into frames, and the video and audio data of each frame is analyzed. Video analysis uses technology to detect iris movement and changes in facial expressions. Audio analysis uses technology to analyze voice tone and patterns. A generative AI model combines these data to determine the authenticity of the video. The results are sent back to the server as a reliability score.

[1015] For example, if a user uploads a video introducing a new product from their device, the server sends the video to a video analysis device, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the server determines that the video is likely to be fake and generates a reliability score.

[1016] Data integration device

[1017] The server receives the analysis results from the information analysis device and the video analysis device and sends them to the data integration device. This device integrates the analysis results of the news article and the video and calculates an overall reliability score. This comprehensively evaluates the authenticity of the news article and the video.

[1018] display device

[1019] The final reliability score and analysis results are sent to a display device, and the server receives them and displays them on the user's device. The user can check the results through their device, and the analysis results will be displayed as "This news article is likely to be unreliable" or "This video is likely to be fake." This will allow users to make decisions based on accurate information, without being misled by false information.

[1020] In this way, the system of the present invention can achieve highly accurate identification of fake news and fake videos and provide accurate information to users.

[1021] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1022] Step 1:

[1023] Users upload suspected fake news or fake videos from their own devices. They select a text file or video file of the news article and click the send button. The input is the text file or video file of the news article, and the output is that this data is sent to the server.

[1024] Specific behavior:

[1025] 1. The user opens a browser on their device and accesses the system's web page.

[1026] 2. Drag and drop your news article or video file or click the "Choose File" button to upload it.

[1027] 3. Once the upload is complete, the user clicks the "Submit" button.

[1028] Step 2:

[1029] The server receives the data sent by the user and distributes the news articles and videos to the corresponding analysis modules. The input is the text file or video file of the news article sent by the user, and the output is the distributed data.

[1030] Specific behavior:

[1031] 1. The server receives the transmitted data.

[1032] 2. Determine the type of data (text or video).

[1033] 3. Text data is sent to the information analysis device, and video data is sent to the video analysis device.

[1034] Step 3:

[1035] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The input is the news article text, and the output is a credibility score.

[1036] Specific behavior:

[1037] 1. The text of a news article is sent to an information analysis device.

[1038] 2. The generative AI model receives a prompt to analyze the news article and begins the analysis.

[1039] 3. Language patterns, context, and writing style are analyzed.

[1040] 4. A confidence score is calculated and sent back to the server.

[1041] Step 4:

[1042] The server sends the video to the fake video analysis module, which splits the video into frames and analyzes the video and audio data. The input is the video file, and the output is a confidence score.

[1043] Specific behavior:

[1044] 1. The video file is sent to the video analyzer.

[1045] 2. The video is divided into frames.

[1046] 3. The video data is analyzed using technology that detects iris movement and changes in facial expression.

[1047] 4. Audio data is analyzed for tone and patterns of voice.

[1048] 5. A generative AI model combines the video and audio data to calculate a reliability score.

[1049] 6. The confidence score is sent back to the server.

[1050] Step 5:

[1051] The server receives the analysis results from the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module. The input is the analyzed credibility score, and the output is the integrated overall credibility score.

[1052] Specific behavior:

[1053] 1. The credibility scores of the news article and the video arrive at the server.

[1054] 2. These scores are sent to a data aggregator.

[1055] 3. The data aggregator combines the credibility scores of the news article and the video to calculate an overall credibility score.

[1056] Step 6:

[1057] The final reliability score and analysis results are sent from the server to the user interface module, which receives them and displays the results on the user's terminal. The input is the overall reliability score, and the output is the display of the analysis results.

[1058] Specific behavior:

[1059] 1. The overall reliability score and analysis results are sent to the user interface module.

[1060] 2. The user interface module processes the results to display them on the user terminal.

[1061] 3. The user checks the results on their device and sees messages such as "This news article is likely to be unreliable" or "This video may be fake."

[1062] In this way, the system can achieve high accuracy in identifying fake news and fake videos and provide accurate information to users.

[1063] (Application example 1)

[1064] 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."

[1065] In recent years, a large number of news articles and videos have been circulating on the Internet, and many of them contain fake news and videos. Such misinformation can lead to incorrect perceptions and judgments among users, potentially causing social unrest. Therefore, there is a need for a method to accurately determine the authenticity of news articles and videos and provide users with accurate information. In particular, for content distribution services, it is important to have a function that can determine the reliability of the content viewed by users in real time and immediately display a reliability score.

[1066] 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.

[1067] In this invention, the server includes server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing news articles, fake video analysis means for analyzing videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and means for determining the authenticity of the news articles and videos viewed by the user in real time and displaying a reliability score. This allows users to check the reliability of the content they view in real time, enabling them to make decisions based on accurate information without being misled by misinformation.

[1068] A "user terminal" is a device operated by a user, and includes a smartphone, tablet, personal computer, and the like.

[1069] The "server means" is a central processing unit that performs analysis and data processing, and has the function of communicating with user terminals via a network.

[1070] The "fake news analysis tool" is an analytical module that determines the authenticity of news articles and has the ability to analyze language patterns, context, and writing style.

[1071] The "fake video analysis means" is an analysis module that determines the authenticity of a video, and includes means for analyzing the video on a frame-by-frame basis and detecting iris movement and changes in facial expression.

[1072] The "multimodal data integration means" has the function of integrating the analysis results obtained from the fake news analysis means and the fake video analysis means and calculating an overall reliability score.

[1073] The "user interface means" is an interface that displays the analysis results to the user, and refers to a screen or application that the user operates.

[1074] "Real-time judgment" is a process that instantly analyzes the authenticity of news articles and videos viewed by users and immediately provides the results to the users.

[1075] A "trust score" is a number generated based on the analysis results that indicates the reliability of a news article or video.

[1076] MODE FOR CARRYING OUT THE INVENTION

[1077] The present invention provides a system for analyzing news articles or videos sent from a user terminal and evaluating their reliability, which includes a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, a user interface unit, and a unit for displaying a reliability score in real time.

[1078] System Configuration

[1079] Server means:

[1080] The server distributes news articles or videos received from user terminals to the analysis module. The server functions as a central processing unit with high-performance processing capabilities, and receives and transmits data in real time via the network.

[1081] Fake news analysis methods:

[1082] The fake news analysis tool analyzes the language patterns, context, and writing style of news articles using a specific generative AI model. The generative AI model is trained on a large number of past news articles and can determine the veracity of news articles with high accuracy. The server receives the analysis results as a credibility score.

[1083] Fake video analysis methods:

[1084] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. Video analysis detects iris movement and changes in facial expressions, while audio analysis analyzes voice tone and patterns. A generative AI model integrates these data to determine the authenticity of the video. The server receives the analysis results as a reliability score.

[1085] Multimodal data integration methods:

[1086] The multimodal data integration method integrates the results obtained from fake news analysis and fake video analysis to calculate a comprehensive credibility score, which enables accurate credibility assessment that takes into account not only the judgment results of a single news article or video, but also the relationship between the two.

[1087] User Interface Methods:

[1088] The user interface means plays a role in displaying the analysis results to the user. The analysis results are displayed on the user terminal in real time, allowing the user to immediately confirm the reliability of the content.

[1089] Example of operation

[1090] When a user uploads a news article (path / to / content / news.txt) from their smartphone, the server routes the news article to the fake news analysis tool, which analyzes language patterns and context using a generative AI model. The resulting trust score is displayed in real time on the user's device, such as "News Trust Score: 85." Similarly, when a user uploads a video, the fake video analysis tool analyzes the video and audio, and displays a trust score.

[1091] Prompt Sentence Examples

[1092] An example of a prompt sentence to input to a generative AI model is, "Please rate the credibility of the following news article." By providing the text of the news article along with this prompt sentence to the generative AI model, the model analyzes the article's language patterns and context and returns a credibility score.

[1093] As described above, the system of the present invention can accurately determine the authenticity of news articles and videos viewed by users and display a reliability score in real time, allowing users to make decisions based on accurate information without being misled by misinformation.

[1094] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1095] Step 1:

[1096] The user selects a news article or video from a user device such as a smartphone and sends it to the server. The input is a text file or video file of the news article, which the server receives. Specifically, the user presses the upload button in the application, selects a file from the file selection screen, and sends it.

[1097] Step 2:

[1098] The server distributes the received data to the analysis modules. It receives news articles or video files as input, and distributes them to the fake news analysis means if they are news articles, and to the fake video analysis means if they are videos. Specifically, the server determines the type of file it receives and transfers it to the corresponding analysis module.

[1099] Step 3:

[1100] The fake news analysis method analyzes the content of news articles. It receives the text of a news article as input and uses a generative AI model to analyze the language patterns, context, and writing style. It generates a credibility score as output. Specifically, the text of a news article is input into the generative AI model, and a credibility score is returned as the analysis result.

[1101] Step 4:

[1102] The fake video analysis method analyzes videos frame by frame and simultaneously analyzes audio data. It receives a video file as input and analyzes the video and audio data for each frame. It uses a generative AI model to determine iris movement, facial expression changes, and vocal tone and patterns, and generates a credibility score as output. Specifically, the video is divided into frames, and the data for each frame is analyzed.

[1103] Step 5:

[1104] The multimodal data integration means integrates the results from the fake news analysis means and the fake video analysis means. It receives the credibility scores of news articles and videos as input and calculates an overall credibility score. It produces this overall score as output. Specifically, multiple credibility scores are input into the integration algorithm, and a final credibility score is calculated.

[1105] Step 6:

[1106] The user interface means displays the calculated trustworthiness score on the user's terminal. It receives the overall trustworthiness score from the server as input and converts it into a format to be displayed on the user's terminal. As an output, the trustworthiness score is displayed on the user's screen. Specifically, the trustworthiness score is displayed on the application screen in a format such as "News Trust Score: 85."

[1107] 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.

[1108] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is composed of a server, a fake news analysis unit, a fake video analysis unit, a multimodal data integration unit, and a user interface unit. Furthermore, the present invention is equipped with an emotion engine that recognizes the user's emotions and incorporates a function to optimize the presentation of the analysis results.

[1109] Program processing

[1110] The user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[1111] Fake News Analysis Module

[1112] The server sends the received news articles to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The generative AI model is trained on historical datasets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[1113] Examples:

[1114] When a user uploads an article about a politician's remarks from their device, the server sends the article to the fake news analysis module, where the generative AI model analyzes language patterns and context. As a result, it determines that the news article has low credibility and generates a credibility score.

[1115] Fake video analysis module

[1116] The server sends the received video to the fake video analysis module, which divides the video into frames and analyzes the video and audio data of each frame. Video analysis uses technology to detect iris movement and changes in facial expressions, while audio analysis analyzes the tone and patterns of voice. A generative AI model integrates these data to determine whether the video is authentic or not.

[1117] Examples:

[1118] When a user uploads a "new product introduction video" from their device, the server sends the video to a fake video analysis module, which analyzes the video and audio. If the iris movements or changes in facial expressions are unnatural, the module determines that the video is likely to be fake and generates a reliability score.

[1119] Emotion Engine

[1120] Furthermore, an emotion engine is incorporated into the user interface means to recognize the user's emotional state in real time. The emotion engine analyzes the user's facial expressions, voice tone, and input patterns to understand the user's emotional state when receiving information. This allows the method of presenting the analysis results to be optimized according to the user's emotions.

[1121] Examples:

[1122] When the emotion engine detects anxiety or discomfort from the user's facial expression or voice while the user is viewing the analysis results, the user interface means provides the analysis results in a more easily understandable format or displays additional explanations to help the user understand.

[1123] Multimodal Data Integration Module

[1124] The server receives the analysis results from the fake news analysis module and the fake video analysis module and sends them to the multimodal data integration module, which integrates the analysis results of the news articles and videos to calculate an overall credibility score.

[1125] Examples:

[1126] When a user uploads a news article and its associated video at the same time, the server combines the results of these analyses to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[1127] User Interface Module

[1128] The final confidence score and analysis results are sent to the user interface module. The server receives them and displays the results on the user's device. The emotional engine detects the user's emotional state, so the display format of the results is adjusted according to the user's emotional state.

[1129] Examples:

[1130] When the user checks the results on their device, the analysis result will display, "This news article is unreliable and is likely fake news." At the same time, the analysis result for the video will display, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[1131] In this way, the system of the present invention achieves highly accurate identification of fake news and fake videos, and furthermore, by optimizing the way in which the analysis results are presented according to the user's emotional state, it is possible to provide accurate information to users and prevent them from being misled by false information.

[1132] The processing flow will be explained below.

[1133] Step 1:

[1134] The user sends a news article suspected of being fake news or a video suspected of being fake news from their device to the server. The user then selects and uploads the text of the news article or the video file.

[1135] Step 2:

[1136] The server receives the data sent by the user, determines whether it is a news article or a video, and then assigns it to either the fake news analysis module or the fake video analysis module depending on the determined data format.

[1137] Step 3:

[1138] The server sends the news article to a fake news analysis module, which uses a generative AI model to analyze the news article's language patterns, context, and writing style. The result of this analysis is a credibility score, which is returned to the server.

[1139] Step 4:

[1140] The server sends the video to a fake video analysis module, which divides the video into frames and detects changes in iris movement and facial expressions in each frame. It also analyzes the audio data, analyzing the tone and patterns of the voice. Based on this data, the generative AI model determines whether the video is authentic, and a reliability score is generated and sent back to the server.

[1141] Step 5:

[1142] The server receives the analysis results provided by the fake news analysis module and the fake video analysis module, and sends these results to the multimodal data integration module, which integrates the analysis results of the news article and the video to calculate an overall credibility score.

[1143] Step 6:

[1144] The server recognizes the user's emotional state in real time through the emotion engine, which grasps the user's emotional state by analyzing the user's facial expressions, voice tone, and input patterns through the user interface means.

[1145] Step 7:

[1146] The server sends the final confidence score and analysis results to the user interface module, which presents the analysis results to the user in the most appropriate manner based on feedback from the emotion engine.

[1147] Step 8:

[1148] The user checks the analysis results on their device, which display a credibility score and supporting reasons. If the emotion engine detects a negative emotion from the user, the analysis results will provide further detailed explanations and supplementary information. For example, it may say, "This news article has low credibility and is likely fake news," providing details of the language patterns and writing style that support this.

[1149] This allows users to identify unreliable information and make decisions based on accurate information. The introduction of an emotion engine improves user understanding and makes analysis results more easily accepted.

[1150] Example 2

[1151] 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."

[1152] Conventional systems for determining the authenticity of news articles and videos are biased toward analyzing a single data mode (only articles or videos), making it difficult to integrate and analyze multimodal data. Furthermore, when users receive analysis results, they are presented in a uniform display format without taking into account their level of understanding or emotional state, which can result in insufficient information being provided to the user. Furthermore, technologies for improving the reliability of analysis results are limited. There is a need for a system that can resolve these issues, achieve high-accuracy identification of fake news and fake videos, and optimize the presentation of analysis results according to the user's emotional state.

[1153] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1154] In this invention, the server includes a processing device that analyzes text data or video data received from a user device, a false information analysis device that analyzes the text data, a video analysis device that analyzes the video data, a display device that displays the analysis results and integration results on the user device, and an emotion recognition device incorporated in the display device that recognizes the user's emotional state in real time and optimizes the presentation method of the analysis results. This enables multimodal data integration that integrates the analysis results of news articles and videos, realizing optimal information presentation according to the user's emotional state. Furthermore, high-precision analysis using a generative AI model enables reliable identification of fake news and fake videos.

[1155] A "user terminal" is a computing device operated by a user, and is a device for transmitting news articles and video data to a server.

[1156] "Text data" refers to information in text format such as news articles.

[1157] "Video data" refers to frame-by-frame video information such as news videos and other video files.

[1158] "Processor means" refers to a computing device for analyzing data received from a user terminal and is a means for identifying audio and video data.

[1159] A "disinformation analysis tool" is a device for determining the authenticity of received text data, and uses a generative AI model to analyze language patterns, context, and writing style.

[1160] The "video analysis means" is a device for determining the authenticity of received video data, analyzing each frame and detecting iris movements and changes in facial expression.

[1161] The "data integration means" is a device that integrates the analysis results obtained by the false information analysis means and the video analysis means and calculates an overall reliability score.

[1162] The "display means" is a device that displays the analysis results and the integration results on a user terminal.

[1163] The "emotion recognition means" is a device that recognizes the user's emotional state in real time and optimizes the method of presenting the analysis results.

[1164] A "generative AI model" is an artificial intelligence model that is trained based on past datasets and can determine truth or falsehood with high accuracy.

[1165] The present invention relates to a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. This system is mainly composed of a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion recognition means. Each means will be described in detail below.

[1166] Server means:

[1167] The server has a processing device for analyzing text data (news articles) or video data (videos) received from user terminals. The server collects data from users and distributes it to the appropriate analysis module. Specifically, text data is sent to the fake news analysis means, and video data is sent to the fake video analysis means.

[1168] Fake news analysis methods:

[1169] The server analyzes the received news articles using a disinformation analysis methodology. This analysis methodology incorporates a generative AI model that performs detailed analysis of the news article's language patterns, context, and writing style. The generative AI model is trained on a large amount of previously collected data sets and can determine the veracity of news with high accuracy. The analysis results are returned to the server as a credibility score.

[1170] Examples:

[1171] When a user uploads a news article about a politician's remarks from their device, the server sends the article to a fake news analysis tool. The generative AI model analyzes language patterns and context, determining that the news article is low in credibility and generating a credibility score.

[1172] Example prompt sentence:

[1173] Please determine whether this news article is true or false.

[1174] Fake video analysis methods:

[1175] The server analyzes the received video using a video analysis method. The video analysis method analyzes the video frame by frame, analyzing iris movement, changes in facial expressions, and voice tone and patterns. The generative AI model integrates this data to determine the authenticity of the video. The analysis result is also returned to the server as a reliability score.

[1176] Examples:

[1177] When a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis means. The video analysis means analyzes the iris movement and changes in facial expressions, and the audio analysis means analyzes the tone and patterns of the voice. As a result, it is determined that "this video is likely to be fake," and a reliability score is generated.

[1178] Example prompt sentence:

[1179] Please tell me if this video is real or not

[1180] Multimodal data integration methods:

[1181] The server sends the analysis results from the false information analysis means and the video analysis means to the data integration means, which integrates the results of both analyses and calculates an overall reliability score, thereby matching the analysis results of the news article and the video, enabling a more comprehensive determination of whether the news article is true or false.

[1182] Examples:

[1183] When a user uploads a news article and its associated video at the same time, the server combines the analysis results of these data to generate an overall credibility score, which is a comprehensive assessment of the veracity of the news article and video.

[1184] Emotion recognition means:

[1185] The server is equipped with an emotion recognition system that optimizes the presentation of analysis results according to the user's emotional state. This system analyzes the user's facial expressions, tone of voice, and input patterns in real time to understand the user's emotional state when receiving information.

[1186] Examples:

[1187] If the emotion recognition means detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the analysis results will be presented in a more understandable format or additional explanations will be displayed to help the user understand.

[1188] User Interface Methods:

[1189] The server displays the final reliability score and analysis results on the user's device via a user interface. The user can check the analysis results and receive detailed explanations through the device. The emotion recognition device monitors the user's emotional state in real time and adjusts the display format accordingly.

[1190] Examples:

[1191] When the user checks the results on their device, the analysis results will display information such as "This news article is unreliable and is likely fake news." At the same time, the analysis results for the video will also display "This video is likely fake." If the emotion recognition method detects the user's anxiety, additional explanations about the analysis results and answers to any questions will be displayed to help the user understand.

[1192] Advantages of this embodiment:

[1193] The system of the present invention not only achieves high-accuracy identification of fake news and fake videos, but also provides an optimal display method that matches the user's emotional state, thereby providing users with more accurate and reliable information and preventing them from being misled by false information.

[1194] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1195] Step 1:

[1196] Users select suspected fake news articles or videos from their own devices and send them to the server. The input is the text data or video file of the news article, and the output is the data sent to the server. Users can either drag and drop files into a dedicated upload form or click the select button to select a file.

[1197] Step 2:

[1198] The server distributes data received from users to the analysis modules. The input is text data or video data sent by the user, and the output is data sent to each analysis module. The server first checks the data format and then sends text data to the fake news analysis module and video data to the fake video analysis module.

[1199] Step 3:

[1200] The server sends the received news article to the fake news analysis module. The input is the text data of the news article, and the output is a credibility score. The generative AI model analyzes the news article's language patterns, context, and writing style. Specifically, it checks for specific keywords, sentence structure, and contextual consistency within the text, and returns the analysis result to the server as a credibility score.

[1201] Step 4:

[1202] The server sends the received video to the fake video analysis module. The input is the video data, and the output is a credibility score. The analysis module divides the video into frames and analyzes the video and audio data of each frame individually. Specifically, video analysis detects iris movement and changes in facial expression, and audio analysis meticulously analyzes the tone and patterns of the voice. The generative AI model integrates this data, determines whether the video is authentic, and sends a credibility score back to the server.

[1203] Step 5:

[1204] The server sends the analysis results from the fake news analysis module and fake video analysis module to the data integration means. The input is the analysis results of the news article and video, and the output is an overall credibility score. The data integration means integrates the credibility score of the news article and the credibility score of the video to calculate an overall credibility score. The integration includes weighting each score and analyzing correlations.

[1205] Step 6:

[1206] The server uses emotion recognition means to analyze the user's emotional state in real time and optimize the display method of the analysis results. The input is the user's facial expression data and voice tone, and the output is optimized display content. The emotion recognition means analyzes the user's facial expression, voice tone, and input patterns to understand the emotional state the user is in when receiving information. The user's emotional state is monitored through a camera and microphone, and the display content is adjusted based on the analysis results.

[1207] Step 7:

[1208] The server displays the final reliability score and analysis results on the user's device via the user interface means. The input is the overall reliability score and analysis results, and the output is the results displayed on the user's device. The user can check the analysis results through their device and receive a detailed explanation. Specifically, the results display screen will display information such as "This news article is unreliable and therefore likely to be fake news," and if the emotion recognition means detects the user's anxiety, additional explanations and answers to questions will be displayed.

[1209] (Application example 2)

[1210] 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."

[1211] In modern society, fake news and fake videos have become a serious problem because many news articles and videos spread instantly via the Internet. In particular, in the advertising industry, there is a high risk that companies' trust will be damaged if advertisements containing unreliable information are delivered to consumers. The present invention aims to provide a system that can accurately identify such fake news and fake videos and check the reliability of advertisements in real time.

[1212] The identification processing 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 server means for analyzing news articles or videos received from a user terminal, fake news analysis means for analyzing the news articles, fake video analysis means for analyzing the videos, multimodal data integration means for integrating the analysis results of the news articles and videos, user interface means for displaying the integration results on the user terminal, and an emotion engine for recognizing the user's emotional state in real time and optimizing the presentation method of the analysis results. This increases the reliability of information in advertisements, reduces corporate risks due to incorrect information, and enables the creation of an environment in which users can view advertisements with peace of mind.

[1213] The "server means" refers to a device that distributes data received from user terminals to processing and analysis modules.

[1214] A "fake news analysis tool" is a device that analyzes the language patterns, context, and writing style of news articles to identify fake news.

[1215] A "generative AI model" is an artificial intelligence model that is trained on historical datasets and analyzes news articles and videos.

[1216] A "fake video analysis means" is a device that analyzes video on a frame-by-frame basis and detects iris movements and changes in facial expression.

[1217] A "multimodal data integration means" is a device that integrates the results of fake news analysis and fake video analysis and calculates an overall reliability score.

[1218] The "user interface means" refers to a device that displays the analysis results on a user terminal, allowing the user to confirm the results.

[1219] An "emotion engine" is a device that analyzes a user's facial expressions and tone of voice in real time to recognize the user's emotional state.

[1220] A "credibility score" is a numerical representation of the reliability of a news article or video based on its veracity.

[1221] This invention is a system that analyzes news articles and videos received from a user terminal and identifies fake news and fake videos. The system includes a server means, a fake news analysis means, a fake video analysis means, a multimodal data integration means, a user interface means, and an emotion engine.

[1222] First, the user sends a news article or video suspected of being fake news from their device to the server. At this time, the user selects and uploads the news article text or video file. The server then distributes the received data to the analysis module.

[1223] The server means has the function of receiving news articles or videos from user terminals and distributing them to the fake news analysis means and fake video analysis means.

[1224] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of received news articles. This generative AI model is trained based on past datasets and can determine the veracity of news with high accuracy. For example, if a user uploads an article about "a politician's statement" from their device, the server sends the article to the fake news analysis method, where the generative AI model analyzes the language patterns and context. As a result, it determines that "this news article has low credibility" and generates a credibility score.

[1225] The fake video analysis method divides the received video into frames and analyzes the video and audio data of each frame. Specifically, the video analysis uses OpenCV to detect iris movements and changes in facial expressions, and the audio analysis uses LibROSA to analyze voice tone and patterns. The generative AI model integrates this data and determines the authenticity of the video. For example, if a user uploads a "new product introduction video" from their device, the server sends the video to the fake video analysis method, which analyzes the video and audio. If the iris movements and changes in facial expressions are unnatural, it is determined that "this video is likely to be fake," and a reliability score is generated.

[1226] The emotion engine recognizes the user's emotional state in real time while the user is checking the analysis results. The emotion engine understands the emotional state of the user when receiving information by analyzing the user's facial expressions and tone of voice. For example, if the emotion engine detects anxiety or discomfort from the user's facial expressions or voice while the user is viewing the analysis results, the user interface means will provide the analysis results in a more understandable format or display additional explanations to help the user understand.

[1227] The multimodal data integration means receives the analysis results of the fake news analysis means and the fake video analysis means from the server, and integrates these results to calculate an overall credibility score. For example, if a user simultaneously uploads a news article and its related video, the server integrates these analysis results to generate an overall credibility score. This score is a comprehensive assessment of the authenticity of the news article and the video.

[1228] The user interface means has the function of displaying the final reliability score and analysis results on the user's device. The server receives this and displays the results on the user's device. Because the emotion engine detects the user's emotional state, the display format of the results is adjusted according to the user's emotional state. For example, when the user checks the results on their device, the analysis result may say, "This news article is unreliable, so it is likely to be fake news." At the same time, the analysis result for the video may also say, "This video is likely fake." If the emotion engine detects the user's anxiety, additional explanations about the analysis results and answers to questions may be displayed to help the user understand.

[1229] Example prompt sentence:

[1230] Please check whether the contents of the advertised article below are trustworthy.

[1231] "Politician A announced a new policy that he said would dramatically improve the economy."

[1232] Result: This advert has low credibility. A similar pattern was found in many fake news stories. Credibility score: 30%

[1233] In this way, the system of the present invention achieves high-accuracy identification of fake news and fake videos, and further optimizes the presentation method of the analysis results according to the user's emotional state, thereby providing users with accurate and safe information and increasing the reliability of information in advertisements.

[1234] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1235] Step 1:

[1236] A user uploads a news article or video from a terminal. The input data includes a news article text file or video file. The terminal uploads this data and sends it to the server. The server receives this data.

[1237] Step 2:

[1238] The server distributes the received data to the analysis modules. If a news article is received, the data is sent to the fake news analysis means, and if a video is received, the data is sent to the fake video analysis means. The input data is a news article or a video file. The output is the data to be processed by each analysis module.

[1239] Step 3:

[1240] The fake news analysis method uses a generative AI model to analyze the language patterns, context, and writing style of news articles. The text of the news article is used as input, and the generative AI model performs data analysis. The output is a credibility score based on the analyzed language patterns, context, and writing style.

[1241] Step 4:

[1242] The fake video analysis method divides the video into frames and analyzes the video and audio data of each frame. The video file is used as input. Specifically, OpenCV is used to detect iris movement and changes in facial expressions, and LibROSA is used to analyze voice tone and patterns. The output is a reliability score based on the video and audio analysis results.

[1243] Step 5:

[1244] The server transmits the analysis results from the fake news analysis means and the fake video analysis means to the multimodal data integration means. The input data are reliability scores. The multimodal data integration means calculates an overall reliability score based on these results. The output is an integrated overall reliability score.

[1245] Step 6:

[1246] The user interface means displays the received overall reliability score on the user terminal. The input data is the reliability score. When the user checks the results, the emotion engine analyzes the user's emotional state in real time and optimizes the display format. Specifically, it analyzes the user's facial expressions and tone of voice and selects a method for presenting the results according to their emotional state. The output is a display of the optimized reliability score.

[1247] Step 7:

[1248] The user checks the results on their device. The input data is the analyzed confidence score and its detailed explanation. If the emotion engine detects anxiety or discomfort while the user is viewing the analysis results, it provides additional explanations and details to help the user understand. The output is a display of the detailed analysis results according to the user's emotional state.

[1249] 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.

[1250] 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.

[1251] 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.

[1252] 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.

[1253] 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.

[1254] 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.

[1255] 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).

[1256] 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.

[1257] 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."

[1258] 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.

[1259] 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).

[1260] 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.

[1261] 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.

[1262] 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.

[1263] 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.

[1264] 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.

[1265] 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.

[1266] 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.

[1267] 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.

[1268] 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.

[1269] 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.

[1270] The following is further disclosed regarding the above embodiment.

[1271] (Claim 1)

[1272] a server means for analyzing news articles or videos received from a user terminal;

[1273] A fake news analysis means for analyzing the news article;

[1274] A fake video analysis means for analyzing the video;

[1275] a multimodal data integration means for integrating the analysis results of the news articles and videos;

[1276] a user interface means for displaying the integration result on a user terminal;

[1277] A system including:

[1278] (Claim 2)

[1279] 10. The system of claim 1, wherein the fake news analysis means includes a generative AI model that analyzes language patterns, context, and writing style of news articles.

[1280] (Claim 3)

[1281] The system of claim 1, wherein the fake video analysis means includes means for analyzing video footage frame by frame and detecting iris movement and changes in facial expression.

[1282] "Example 1"

[1283] (Claim 1)

[1284] an information processing device that analyzes news articles or videos received from a user terminal;

[1285] an information analysis device that analyzes the news article;

[1286] a video analysis device that analyzes the video;

[1287] a data integration device that integrates the analysis results of the news articles and videos;

[1288] a display device that displays the integration result on a user terminal;

[1289] A system including:

[1290] (Claim 2)

[1291] 10. The system of claim 1, wherein the information analysis device includes a generative AI model that analyzes language patterns, context, and writing style of news articles.

[1292] (Claim 3)

[1293] 2. The system according to claim 1, wherein the video analysis device includes means for analyzing video images frame by frame and detecting iris movements and changes in facial expression.

[1294] "Application Example 1"

[1295] (Claim 1)

[1296] a server means for analyzing news articles or videos received from a user terminal;

[1297] A fake news analysis means for analyzing the news article;

[1298] A fake video analysis means for analyzing the video;

[1299] a multimodal data integration means for integrating the analysis results of the news articles and videos;

[1300] a user interface means for displaying the integration result on a user terminal;

[1301] A means to determine the authenticity of news articles and videos viewed by users in real time and display a reliability score;

[1302] A system including:

[1303] (Claim 2)

[1304] 10. The system of claim 1, wherein the fake news analysis means includes a generative AI model that analyzes language patterns, context, and writing style of news articles.

[1305] (Claim 3)

[1306] The system of claim 1, wherein the fake video analysis means includes means for analyzing video footage frame by frame and detecting iris movement and changes in facial expression.

[1307] "Example 2: Combining Emotion Engines"

[1308] (Claim 1)

[1309] a processing unit for analyzing text data or video data received from a user terminal;

[1310] a false information analysis means for analyzing the text data;

[1311] a video analysis means for analyzing the video data;

[1312] a data integration means for integrating the analysis results of the text data and the video data;

[1313] a display means for displaying the analysis result and the integration result on a user terminal;

[1314] an emotion recognition means incorporated in the display means for recognizing the user's emotional state in real time and optimizing the presentation method of the analysis results;

[1315] A system including:

[1316] (Claim 2)

[1317] 10. The system of claim 1, wherein the disinformation analysis means includes a generative AI model that analyzes language patterns, context, and writing style of text data.

[1318] (Claim 3)

[1319] 2. The system according to claim 1, wherein the video analysis means includes means for analyzing the video frame by frame and detecting iris movements and changes in facial expression.

[1320] "Application example 2 when combining emotion engines"

[1321] (Claim 1)

[1322] a server means for analyzing news articles or videos received from a user terminal;

[1323] A fake news analysis means for analyzing the news article;

[1324] A fake video analysis means for analyzing the video;

[1325] a multimodal data integration means for integrating the analysis results of the news articles and videos;

[1326] a user interface means for displaying the integration result on a user terminal;

[1327] An emotion engine that recognizes the user's emotional state in real time and optimizes the way the analysis results are presented;

[1328] A system including:

[1329] (Claim 2)

[1330] 10. The system of claim 1, wherein the fake news analysis means includes a generative AI model that analyzes language patterns, context, and writing style of news articles.

[1331] (Claim 3)

[1332] The system of claim 1, wherein the fake video analysis means includes means for analyzing video footage frame by frame and detecting iris movement and changes in facial expression.

[1333] (Claim 4)

[1334] 2. The system of claim 1, wherein the emotion engine includes means for analyzing a user's facial expression patterns and vocal tone in real time.

[1335] (Claim 5)

[1336] 2. The system of claim 1, wherein the multimodal data integration means includes means for calculating an overall reliability score based on the analysis of the news article and the video. [Explanation of symbols]

[1337] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a server means for analyzing news articles or videos received from a user terminal; A fake news analysis means for analyzing the news article; A fake video analysis means for analyzing the video; a multimodal data integration means for integrating the analysis results of the news articles and videos; a user interface means for displaying the integration result on a user terminal; A system including:

2. 10. The system of claim 1, wherein the fake news analysis means includes a generative AI model that analyzes language patterns, context, and writing style of news articles.

3. The system according to claim 1 , wherein the fake video analysis means includes means for analyzing video images frame by frame and detecting iris movements and changes in facial expressions.

Images