system

Abstract

Provide a system. A real-time fact-checking system that uses natural language processing technology to evaluate the veracity of news articles and images. means for receiving news articles or image data from a terminal; means for preprocessing the received data to extract text data and text information within images; A means of analyzing the extracted text data and information in the images and assessing their veracity by comparing them with reliable databases and other reliable sources; and A means for determining whether the news or image is true or false based on the evaluation result; means for generating summary information based on the judgment results and facts and transmitting the summary information to a terminal; A system including:

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, there has been an increase in potentially misleading fake information in news articles and images on the Internet, creating a need for real-time evaluation of its veracity. Such fake information has a high potential for social impact, making it important to provide reliable information quickly. However, conventional methods require manual checking of large amounts of information, which requires time and effort, making real-time fact-checking difficult. To address this issue, there is a need for the development of an efficient and accurate fact-checking system. [Means for solving the problem]

[0005] The present invention provides a real-time fact-checking system that uses natural language processing technology to evaluate the veracity of news articles and images. The system comprises the following means: First, a means for receiving news articles or image data from a terminal. Next, a means for preprocessing the received data and extracting text data or text information from the image is provided. Further, a means for analyzing the extracted text data or information from the image and comparing it with a reliable database or other reliable information source to evaluate its veracity is provided. Next, a means for determining whether the news or image is true or false based on the evaluation results is provided. Finally, a means for generating summary information based on the determination results and facts and sending it to the terminal is provided. This enables the rapid identification of fake information and the provision of accurate information, which is expected to have the effect of reducing its social impact.

[0006] A "news article" is a textual report or information provided over the Internet or through other media.

[0007] An "image" is still image data in a digital format that contains visual information.

[0008] A "terminal" is a digital device used by a user, such as a computer, smartphone, or tablet.

[0009] A "server" is a computer system that processes data and provides services over a network.

[0010] "Natural language processing technology" is a technology that enables computers to understand, interpret, and generate human language.

[0011] "Fact-checking" is the process of verifying and assessing the truthfulness and accuracy of information provided.

[0012] "Preprocessing" is the process of removing unnecessary information from news articles and image data and converting them into a format suitable for analysis.

[0013] "Text data" refers to information expressed as a string of characters, including text information in news articles and images.

[0014] Optical character recognition (OCR) is a technology that converts character information in an image into digital text data.

[0015] A "database" is a collection of data compiled from reliable sources and used for cross-checking.

[0016] The "rating score" is a numerical indicator of the veracity of a news article or image.

[0017] A "summary" is a shortened, easy-to-understand summary of factual information related to a news article or image.

[0018] "Matching" is the process of comparing extracted text data or information in images with a reliable database to confirm a match. [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 uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system performs a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[0041] System Overview

[0042] The system includes the following main components to assess the veracity of news articles and images:

[0043] 1. Data Entry Module

[0044] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[0045] The terminal transmits the input data to the server.

[0046] 2. Data Preprocessing Module

[0047] The server removes HTML tags and unnecessary data from the received news articles and extracts the pure text data.

[0048] The server extracts text information from the image data using OCR technology and analyzes the content of the image using image recognition technology.

[0049] 3. Fact-checking module

[0050] The server analyzes the extracted text data and text information in the image using natural language processing technology.

[0051] The server compares the text data with reliable databases and news sources to assess its veracity.

[0052] 4. Judgment result generation module

[0053] Based on the results of the fact check, the server determines whether the news or image is true or false.

[0054] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[0055] 5. Result Delivery Module

[0056] The server transmits the generated judgment result and summary information to the user's terminal.

[0057] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[0058] Program processing

[0059] The program processing of this system will be explained in natural language below.

[0060] 1. Receiving Data

[0061] A user uses a terminal to input the URL or image data of a news article into the system.

[0062] The terminal sends the input data to the server.

[0063] 2. Data Preprocessing

[0064] The server retrieves the text data of the news article from the received URL.

[0065] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[0066] For image data, the server uses OCR technology to extract text information within the image.

[0067] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[0068] 3. Conduct fact-checks

[0069] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[0070] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[0071] 4. Generating the judgment result

[0072] The server generates a score based on the fact-check results to determine whether a news article or image is true or false.

[0073] The server generates news summaries based on reliable information.

[0074] 5. Providing results

[0075] The server transmits the judgment result and summary information to the user's terminal.

[0076] Users can view the results on their devices and verify the authenticity of the information.

[0077] Specific examples

[0078] 1. News article example

[0079] The user enters the URL of a news article that says, "A politician committed fraud."

[0080] The server retrieves the news article from the URL and preprocesses the text data.

[0081] The server uses natural language processing to analyze the content of news articles and compare them with a reliable database.

[0082] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[0083] Users can check the judgment results and summary on their device.

[0084] 2. Example images

[0085] Users upload "photos of specific incidents."

[0086] The server analyzes the image data using OCR technology and extracts text information.

[0087] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0088] The server determines that the image is misleading and generates a factual explanation.

[0089] Users can check the judgment results and summary explanation on their terminal.

[0090] As described above, the present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and images in real time and quickly provide accurate information to users, allowing users to make decisions based on reliable information without being misled by potentially misleading information.

[0091] The processing flow will be explained below.

[0092] Step 1:

[0093] The user inputs the URL or image data of a news article into the device.

[0094] Step 2:

[0095] The device sends the entered URL and image data to the server.

[0096] Step 3:

[0097] The server scrapes text data from the URL of the news article received.

[0098] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[0099] Step 4:

[0100] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[0101] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[0102] Step 5:

[0103] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3 (registered trademark)) to understand the context and content of the article.

[0104] Step 6:

[0105] The server compares the analyzed data with reliable databases and news sources (e.g., public fact-checking sites) to assess the veracity of the information.

[0106] Step 7:

[0107] Based on the results of the fact check, the server scores the news article or image as true or false.

[0108] Step 8:

[0109] The server generates an accurate news summary, including judgments and facts from reliable sources.

[0110] Step 9:

[0111] The server transmits the generated judgment result and summary information to the user's terminal.

[0112] Step 10:

[0113] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0114] Example 1

[0115] 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."

[0116] Conventional fact-checking systems for news articles and image data often took a long time to evaluate the veracity of articles or based their judgments on unreliable sources. As a result, users were at a higher risk of making decisions based on incorrect information, making it difficult to provide reliable information. Furthermore, inaccurate analysis was not possible due to insufficient preprocessing, such as removing HTML tags from news articles or using optical character recognition technology for image data.

[0117] 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.

[0118] In this invention, the server includes: a means for a user to input a URL or image data of a news article from a terminal and transmit the data from the terminal to the server; a means for acquiring text data from the URL of the news article received by the server or extracting text information from the image data using optical character recognition technology; a means for preprocessing the acquired text data and removing HTML tags and unnecessary data; a means for analyzing the preprocessed text data or image data using natural language processing technology and comparing it with a reliable database or other reliable information source to evaluate its veracity; a means for determining whether the news article or image is true or false based on the comparison result; and a means for generating summary information based on the determination result and facts and transmitting it to the user's terminal. This enables the veracity of news articles and image data to be evaluated with high accuracy in real time and provides accurate information to users quickly.

[0119] A "user" is a person who inputs the URL of a news article or image data into the system via a terminal.

[0120] A "terminal" is a device that receives news article URLs and image data from users and sends them to a server.

[0121] "Server" means a device that processes and analyzes data received from users, evaluates its authenticity, and generates and provides results.

[0122] A "news article" is a web page containing news information.

[0123] "URL" refers to the web address of a news article.

[0124] "Image data" refers to an image file provided by the user.

[0125] "Optical character recognition technology" is a technology that analyzes and extracts text information from images.

[0126] An "HTML tag" is an element of a markup language that defines the structure and content of a web page.

[0127] "Natural language processing technology" is a technology that enables computers to understand, generate, and analyze human language.

[0128] A "database" is a system for systematically storing and managing information.

[0129] "Veracity" refers to whether a news article or image data is based on facts.

[0130] A "judgment score" is a numerical value that is displayed as a result of evaluating the veracity of a news article or image data.

[0131] "Summary information" is information that briefly summarizes the judgment results of news articles or image data.

[0132] "Preprocessing" is the process of preparing data before analysis or evaluation.

[0133] A "reliable database" is a database built on accurate and reliable sources of information.

[0134] "Evaluation" means analyzing data and determining its veracity.

[0135] "Fact checking" is the process of verifying whether the contents of a news article or image data are based on facts.

[0136] This invention is a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system executes a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[0137] The system is roughly divided into a server, a terminal, and a user. First, a user inputs the URL of a news article and image data from their own terminal (PC, smartphone, tablet, etc.). The input data is sent from the terminal to the server.

[0138] The server performs various preprocessing operations on the received data. For news article URLs, web scraping techniques are used to remove HTML tags and obtain pure text data (e.g., using libraries such as BeautifulSoup). For image data, optical character recognition techniques (e.g., Tesseract) are used to extract text information, and image recognition techniques (e.g., OpenCV or TENSORFLOW®) are used to analyze the image content.

[0139] The server then analyzes the preprocessed text data and the text information in the images using natural language processing techniques (e.g., spaCy or NLTK). The analysis results are then cross-referenced with reliable databases and other trusted sources. This cross-reference uses APIs and web crawlers to obtain real-time data (e.g., Wikipedia API or official news API).

[0140] As a result of the matching, the server evaluates the veracity of the news article or image data and generates a judgment score. The server then generates a summary based on reliable information and provides it to the user in an easy-to-understand format. A generative AI model (e.g., GPT-3) is used to generate this summary.

[0141] Finally, the server sends the judgment result and summary information to the user's device. The user can view the judgment result and summary information on their device and confirm the authenticity of the news or image data. This series of processes allows the user to make decisions based on reliable information without being misled by potentially misleading information.

[0142] Specific example explanation

[0143] News article example

[0144] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[0145] 2. The device sends this URL to the server.

[0146] 3. The server retrieves the news article from the URL and preprocesses the text data.

[0147] 4. The server uses natural language processing to analyze the content of the news article and compare it with a reliable database.

[0148] 5. The server determines that the article contains some errors and generates a summary news based on accurate information.

[0149] 6. The user checks the judgment results and summary on the terminal.

[0150] Image example

[0151] 1. The user uploads a photo of a specific incident.

[0152] 2. The terminal sends this image data to the server.

[0153] 3. The server analyzes the image data using OCR technology and extracts text information.

[0154] 4. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0155] 5. The server determines that the image is misleading and generates a factual explanation.

[0156] 6. The user checks the judgment result and summary explanation on the terminal.

[0157] Prompt Sentence Examples

[0158] Example news article prompt:

[0159] Enter the URL of the following news article and our system will analyze its content and rate its veracity:

[0160] Example: https: / / example.com / news / article123

[0161] Image prompt example:

[0162] Please upload the following image and our system will analyze its content and rate its authenticity:

[0163] Example: (Drag and drop image file)

[0164] The above is an embodiment of the present invention. The present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and image data in real time and quickly provide accurate information to users.

[0165] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0166] Step 1: Receiving Data

[0167] The user enters the URL or image data of a news article into an input form on the device, for example, by copying and pasting the URL of a news site or by dragging and dropping an image file.

[0168] The terminal receives the input data and sends it to the server as an HTTP request. Specifically, it converts the input data into JSON format and generates a POST request. This request is added to the processing queue.

[0169] Input: News article URL or image data

[0170] Output: HTTP request to the server

[0171] Step 2: Data preprocessing (analysis of news articles)

[0172] The server retrieves the web page content from the URL of the news article received using a web scraping technique (e.g., BeautifulSoup).

[0173] The server removes unnecessary information such as HTML tags and advertisements from the HTML pages it retrieves, and extracts the text data. Specifically, it uses an HTML parser to filter out tag information and extract only the pure text data.

[0174] Input: News article URL

[0175] Output: Preprocessed text data

[0176] Step 2: Data preprocessing (analysis of image data)

[0177] The server uses optical character recognition technology (e.g., Tesseract) on the received image data to extract text information.

[0178] The server uses image recognition technology (e.g., OpenCV or TensorFlow) to analyze the content of the image and identify relevant information, such as identifying faces and objects contained within the image.

[0179] Input: Image data

[0180] Output: Extracted text information and image content data

[0181] Step 3: Conduct a fact check

[0182] The server analyzes the preprocessed text data or the text information in the images using natural language processing techniques (e.g., spaCy or NLTK), performing syntactic and semantic analysis of the text to extract important keywords and phrases.

[0183] The server cross-checks the parsed data against reliable databases and other trusted sources, using APIs and web crawlers to retrieve real-time data (e.g., Wikipedia API or official news APIs).

[0184] Input: Preprocessed text or image data

[0185] Output: Matching result

[0186] Step 4: Generate a decision result

[0187] The server generates a score based on the fact-check results to evaluate the veracity of the news article or image. Specifically, it calculates the degree of match between the matching results and calculates a score ranging from 0 to 100.

[0188] The server uses a generative AI model (e.g., GPT-3) to generate a summary based on reliable information, including a rating of veracity and relevant factual information.

[0189] Input: Matching result

[0190] Output: Judgment score and summary information

[0191] Step 5: Delivering results

[0192] The server sends the generated judgment result and summary information to the user's device using an HTTP response. Specifically, the generated information is encoded in JSON format and returned as an HTTP response.

[0193] The user checks the judgment results and summary information on the device. For example, the results are displayed in the display area of ​​a web browser. Based on this, the user can determine the authenticity of the news article or image.

[0194] Input: Judgment score and summary information

[0195] Output: Providing results to the user

[0196] (Application example 1)

[0197] 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."

[0198] Conventional systems for assessing the veracity of news articles and images have difficulty assessing the credibility of information entered by users in real time, and providing assessment results quickly via smartphones has been a particular challenge. Furthermore, there is no well-established mechanism for notifying users of misinformation in real time, and immediate comparison with reliable sources is often not performed. This creates a risk that users will make decisions based on incorrect information.

[0199] 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.

[0200] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and comparing it with a reliable database or other reliable information sources to evaluate their veracity, means for determining whether the news or images are true or false based on the evaluation results, means for generating summary information based on the evaluation results and facts and sending it to the terminal, and means for notifying the user of the generated summary information and the evaluation results in real time via an application installed on the smartphone. This enables immediate evaluation of the veracity of news articles and images and the provision of the results in real time.

[0201] A "news article" is a document created to inform the public of various events or information.

[0202] "Image data" is visual information stored in digital format, including photographs and illustrations.

[0203] A "terminal" is a device used by a user to enter information and view results, typically a smartphone or computer.

[0204] "Means for receiving" refers to a mechanism for acquiring news articles and image data sent from a terminal.

[0205] "Preprocessing" refers to the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[0206] "Text data" is digital data that contains text information.

[0207] Optical character recognition (OCR) is a technology for extracting text information from an image as digital character data.

[0208] "Natural language processing technology" is a technology that allows computers to understand and process human language.

[0209] A "trusted database" is a data repository that stores reliable information and primarily contains data from trusted sources.

[0210] A "source" is a data source that provides particular information.

[0211] A "means for assessing veracity" is a mechanism for determining whether the data is true by analyzing preprocessed data and comparing it with reliable sources of information.

[0212] "Evaluation results" refer to the results of a truthfulness analysis, indicating whether the news or image is true or false.

[0213] "Summary information" is information created based on the evaluation results in a concise, easy-to-understand format.

[0214] "Means of real-time notification" refers to a mechanism for instantly conveying the generated summary information and evaluation results to users, and is primarily implemented through smartphone applications.

[0215] An "application" is a software program that runs on a computer to perform a particular function.

[0216] A description will be given of an embodiment of the present invention. A system for evaluating the veracity of news articles and images in real time and providing the results to a user comprises the following elements.

[0217] 1. System Configuration

[0218] This real-time fact-checking system mainly includes the following components:

[0219] Data Entry Module

[0220] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[0221] Data entered from the terminal is transmitted to the server.

[0222] Data Preprocessing Module

[0223] The server retrieves text data from the URL of the received news article and removes unnecessary information such as HTML tags and advertisements.

[0224] For image data, the server uses optical character recognition technology (OCR) to extract text information within the image.

[0225] The text data is automatically cleaned and converted into a format suitable for analysis.

[0226] Fact Checking Module

[0227] The server analyzes the cleaned text data or information in the images using natural language processing techniques.

[0228] The recognized text data is checked against reliable databases and other trusted sources to assess its veracity.

[0229] Judgment result generation module

[0230] The veracity of news and images is determined based on the results of fact-checking.

[0231] Judgments and fact-based summaries are automatically generated.

[0232] Result Delivery Module

[0233] The server transmits the generated judgment result and summary information to the user's terminal.

[0234] An application installed on a smartphone notifies the user of this information in real time.

[0235] The user can view the results and their summary through the terminal.

[0236] 2. Hardware and Software Requirements

[0237] The system uses a variety of hardware and software tools.

[0238] Hardware

[0239] Devices: smartphones, computers

[0240] Server: A server with high-performance processing capabilities

[0241] software

[0242] OCR technology: software for extracting text information from images (e.g., Tesseract OCR)

[0243] Natural language processing libraries: Tools for analyzing news articles and text data (e.g., SpaCy and NLTK)

[0244] Database connection software: Libraries for communicating with reliable databases

[0245] Smartphone app: An application that provides users with real-time information

[0246] Generative AI models: AI models used to generate summaries and analysis results for text data.

[0247] 3. Specific Examples

[0248] Specific usage scenarios include:

[0249] News article example

[0250] The user enters the URL of a news article that purports to say that a certain politician committed fraud.

[0251] The server retrieves the news article from the URL and preprocesses the text data.

[0252] The content of news articles is analyzed using natural language processing and compared with a reliable database.

[0253] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[0254] Users can check the assessment results and summary in real time using a smartphone application.

[0255] Image example

[0256] Users upload "photos of specific incidents."

[0257] The server analyzes the image data using OCR technology and extracts text information.

[0258] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0259] The server determines that the image is misleading and generates a factual explanation.

[0260] Users can check the assessment results and summary explanations in real time using a smartphone application.

[0261] 4. Examples of prompts

[0262] An example of a prompt that a user might enter into the system is:

[0263] "https: / / example.com / news-article Check whether this news article is true or not."

[0264] "Please verify this image is authentic."

[0265] This allows users to identify misinformation in real time and make quick decisions based on trusted sources of information.

[0266] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0267] Step 1:

[0268] User enters the URL or image data of a news article

[0269] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[0270] Input: News article URL or image data

[0271] Output: Input data is sent from the device to the server

[0272] Step 2:

[0273] The server preprocesses the received data

[0274] The server retrieves text data from the URLs of received news articles, removing unnecessary information such as HTML tags and advertisements, and extracting pure text data.

[0275] For image data, OCR technology is used to extract text information within the image.

[0276] Input: User-entered news article URL or image data

[0277] Output: Preprocessed text data or text information in images

[0278] Step 3:

[0279] The server analyzes the data using natural language processing technology

[0280] The server then analyzes the preprocessed text data and text information in the images using natural language processing techniques, which includes extracting words, understanding context, and summarizing information.

[0281] Input: Preprocessed text data or text information in images

[0282] Output: Parsed text data

[0283] Step 4:

[0284] The server checks the authenticity against a reliable database

[0285] The server then compares the parsed text data with reliable databases and other trusted sources, assessing the authenticity of the news article or image.

[0286] Input: Parsed text data

[0287] Output: Veracity evaluation result

[0288] Step 5:

[0289] The server generates the judgment result and summary information

[0290] The server generates a score based on the fact-check evaluation results to determine whether a news article or image is true or false, and creates a fact-based summary based on the evaluation results.

[0291] Input: Veracity assessment result

[0292] Output: Judgment results and summary information

[0293] Step 6:

[0294] The server sends the results to the device

[0295] The server then sends the generated judgment results and summary information to the user's device, and an application installed on the smartphone notifies the user in real time.

[0296] Input: Judgment results and summary information

[0297] Output: Notification and display on the user's terminal

[0298] Step 7:

[0299] User checks the results

[0300] Users can check the results and summary information sent from the server through a smartphone application, allowing them to determine the authenticity of news articles and images.

[0301] Input: Notification from the server and displayed results

[0302] Output: User's understanding and judgment

[0303] The above steps realize a system that evaluates the veracity of news articles and images in real time and provides appropriate information to users.

[0304] 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.

[0305] The present invention relates to a system that uses natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provides results based on the user's emotions. This system performs a series of processes, including preprocessing, analysis, fact-checking, emotion recognition, and generating and providing judgment results for news articles or image data.

[0306] System Overview

[0307] The system includes the following main components:

[0308] 1. Data Entry Module

[0309] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[0310] The terminal transmits the input data to the server.

[0311] 2. Data Preprocessing Module

[0312] The server removes unnecessary data such as HTML tags and advertisements from the received news articles and extracts pure text data.

[0313] The server uses optical character recognition (OCR) technology on the image data to extract text information within the image.

[0314] The server analyzes the content of the image using image recognition technology.

[0315] 3. Fact-checking module

[0316] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[0317] The server checks the text data against reliable databases and other trusted sources to assess its veracity.

[0318] 4. Emotion Engine Module

[0319] The server analyzes the user's voice data and facial expression data acquired through the user's terminal and recognizes emotions.

[0320] The server adjusts the method of providing the judgment results and summary information based on the user's emotions recognized by the emotion engine.

[0321] 5. Judgment Result Generation Module

[0322] The server determines whether the news or image is true or false based on the results of the fact check.

[0323] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[0324] 6. Result Delivery Module

[0325] The server transmits the generated judgment result and summary information to the user's terminal.

[0326] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[0327] Program processing

[0328] The program processing of this system will be explained in natural language below.

[0329] 1. Receiving Data

[0330] A user uses a terminal to input the URL or image data of a news article into the system.

[0331] The terminal sends the input data to the server.

[0332] 2. Data Preprocessing

[0333] The server obtains the text data of the news article from the received URL.

[0334] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[0335] The server analyzes the image data using OCR technology and extracts text information.

[0336] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[0337] 3. Conduct fact-checks

[0338] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[0339] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[0340] 4. Emotion Recognition by Emotion Engine

[0341] The server uses an emotion engine to analyze the user's voice data and facial expression data acquired through the user's device, and recognizes the user's current emotions.

[0342] For example, the emotional state of the user, such as whether they are surprised, angry, or relaxed, is grasped.

[0343] 5. Generating the judgment result

[0344] Based on the results of the fact check, the server generates a score that determines whether the news article or image is true or false.

[0345] The server generates a news summary based on reliable information, including judgments and facts from reliable sources.

[0346] 6. Tailored delivery of results

[0347] The server adjusts the display method of the judgment result and summary information based on the user's emotional state recognized by the emotion engine.

[0348] For example, if the user is surprised, provide more detailed information to make it easier for the user to understand.

[0349] 7. Providing Results

[0350] The server transmits the generated judgment result and summary information to the user's terminal.

[0351] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0352] Specific examples

[0353] 1. News article example

[0354] A user enters the URL of a news article that says, "A politician committed fraud."

[0355] The server retrieves the news article from the URL and preprocesses the text data.

[0356] The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[0357] The server determines that this article contains some errors and generates a summary news based on accurate information.

[0358] The server recognizes the user's current emotion and, if they are surprised for example, provides a result with detailed supporting information.

[0359] Users can check the judgment results and summary on their device.

[0360] 2. Example images

[0361] A user uploads a "photo of a specific incident."

[0362] The server analyzes the image data using OCR technology and extracts text information.

[0363] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0364] The server determines that the image is misleading and generates a factual explanation.

[0365] The server recognizes the user's current emotions and, for example, if they are anxious, provides results that are thoughtfully reassuring.

[0366] Users can check the judgment results and summary explanation on their terminal.

[0367] As described above, the present invention is a system that utilizes natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provide information that is individually tailored to the user's emotions. As a result, users can make decisions based on reliable information without being misled by potentially misleading information. Furthermore, by taking the user's emotions into consideration, a more user-friendly and easy-to-use system can be realized.

[0368] The processing flow will be explained below.

[0369] Step 1:

[0370] The user inputs the URL or image data of a news article into the device.

[0371] Step 2:

[0372] The device sends the entered URL and image data to the server.

[0373] Step 3:

[0374] The server scrapes text data from the URL of the news article received.

[0375] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[0376] Step 4:

[0377] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[0378] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[0379] Step 5:

[0380] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[0381] Step 6:

[0382] The server cross-checks the parsed data against reliable databases and other trusted sources (e.g., public fact-checking sites) to assess the veracity of the information.

[0383] Step 7:

[0384] Based on the results of the fact check, the server scores the news article or image as true or false.

[0385] Step 8:

[0386] The server uses an emotion engine to analyze the voice and facial expression data it receives from the user's device and recognizes the user's current emotions.

[0387] Step 9:

[0388] The server adjusts the method of providing the judgment results and summary information based on the user's emotional state.

[0389] For example, if the user is surprised, add more information in an easy-to-understand format.

[0390] Step 10:

[0391] The server generates an accurate news summary, including judgments and facts from reliable sources.

[0392] Step 11:

[0393] The server transmits the generated judgment result and summary information to the user's terminal.

[0394] Step 12:

[0395] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0396] Example 2

[0397] 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."

[0398] In modern society, quickly and accurately assessing the veracity of news articles and images has become an important issue. The inability to instantly determine the reliability of such information can lead to the spread of false information and misleading images, which can have a significant impact on society. Furthermore, as the amount of information increases, it is becoming increasingly difficult for users to quickly verify the authenticity of information. Furthermore, information provided does not take into account the user's feelings, which can lead to misunderstandings and anxiety.

[0399] 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.

[0400] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their authenticity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data to recognize emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts and transmitting it to the terminal in a form that corresponds to the user's emotional state. This makes it possible to quickly and accurately evaluate the authenticity of news articles and images and provide information that corresponds to the user's emotions.

[0401] A "news article" is text data containing current information and news reports that are primarily published on websites and digital media.

[0402] "Image data" refers to file formats (e.g., JPEG, PNG) that contain visual information stored in a digital format.

[0403] A "terminal" is an electronic device such as a computer or smartphone that is used to input news articles and image data and view the results.

[0404] "Server" is a central control device that provides resources for data processing and analysis and manages the operation of the entire system.

[0405] "Preprocessing" refers to the process of removing unnecessary information from news articles or image data and converting them into a format suitable for analysis.

[0406] "Text data" refers to purely textual information extracted from news articles or image data.

[0407] "Optical character recognition technology" refers to technology (e.g., OCR) used to convert text in an image into digital characters.

[0408] "Natural language processing technology" refers to computer science technology for analyzing the meaning of text data and understanding its context and content.

[0409] A "trusted database" is a data repository containing verified information sources that can be used to assess veracity.

[0410] "Emotion recognition technology" is a technology that analyzes a user's voice data and facial expression data to understand their emotional state.

[0411] "Decision result" refers to the conclusion reached after evaluating the veracity of news articles or image data.

[0412] "Summary information" refers to information that succinctly summarizes the important points of a news article or image data.

[0413] A "reliable source" is a source or database that is widely recognized as providing truthful information.

[0414] This invention relates to a real-time fact-checking system that utilizes natural language processing and emotion recognition technologies. This system evaluates the veracity of news articles and image data and provides information according to the user's emotions, thereby realizing the provision of highly reliable information.

[0415] System configuration and operation

[0416] The system includes the following major components:

[0417] server

[0418] Device (PC or smartphone)

[0419] Receiving data

[0420] The user inputs the URL and image data of the news article into the terminal, which then sends it to the server.

[0421] Data Preprocessing

[0422] The server downloads the HTML data from the URL of the received news article and extracts the text data using an HTML parser (e.g., BeautifulSoup). It then removes unnecessary information, such as HTML tags and advertisements, from the text data. For image data, it extracts the text information from the image using an OCR engine (e.g., Tesseract OCR) and analyzes the image content using an image recognition model (e.g., ResNet, YOLO).

[0423] Data analysis and evaluation

[0424] The server inputs the preprocessed text data and information in the images into natural language processing models (e.g., BERT, GPT-3) to analyze the context and content, and then compares the information with reliable databases (e.g., Wikidata, FactCheck.org) and news sources via API to assess the veracity of the information.

[0425] Emotion recognition

[0426] The server inputs the voice data and facial expression images sent from the device into an emotion recognition engine (e.g., OpenFace, IBM Watson (registered trademark) Emotion Recognition) to understand the user's emotional state. For example, it recognizes emotions such as surprise, anger, and relaxation.

[0427] Generation of judgment results and summary information

[0428] The server determines whether a news article or image is true or false based on the fact-check results and generates a truthfulness score. The server then uses natural language generation technology (e.g., GPT-3) to create a news summary, which includes the results of the judgment and facts based on reliable sources.

[0429] Providing information

[0430] The server adjusts the display method of the judgment results and summary information according to the user's emotional state. For example, if the user is surprised, it provides more detailed information, and if the user is relaxed, it provides simple information. The generated information is sent to the terminal, where the user can view it.

[0431] Specific examples

[0432] News article example

[0433] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[0434] 2. The server retrieves the news article from the URL and preprocesses the text data.

[0435] 3. The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[0436] 4. The server determines that this article contains some errors and generates a summary news article based on accurate information.

[0437] 5. The server recognizes the user's emotion and provides a result with detailed supporting information if, for example, they are surprised.

[0438] 6. The user checks the judgment results and summary on the terminal.

[0439] Image example

[0440] 1. A user uploads a photo of a specific incident.

[0441] 2. The server analyzes the image data using OCR technology and extracts text information.

[0442] 3. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0443] 4. The server determines that the image is potentially misleading and generates a factual explanation.

[0444] 5. The server recognizes the user's emotions and provides reassuring information if, for example, the user is worried.

[0445] 6. The user checks the judgment result and summary explanation on the terminal.

[0446] Example prompt sentence:

[0447] The user is prompted to enter information in a format such as "Please enter the URL of a news article" or "Please upload image data."

[0448] As described above, the present invention makes it possible to evaluate the veracity of news articles and images in real time and provide information adjusted according to the user's emotions.

[0449] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0450] Program processing steps

[0451] Step 1: Receiving Data

[0452] A user uses a device to input a URL or image data of a news article.

[0453] Specific actions: Paste or upload the URL of a news article or image file into the input form displayed on the device.

[0454] Input: News article URL or image data.

[0455] Output: The received URL or image data.

[0456] Step 2: Sending data

[0457] The terminal transmits the input data to the server.

[0458] Specific operation: The device sends an HTTP POST request to the server, including the received URL or image data.

[0459] Input: Received URL or image data.

[0460] Output: URL or image data received by the server.

[0461] Step 3: Acquiring and preprocessing news articles

[0462] The server retrieves the text data of the news article from the received URL and removes unnecessary information.

[0463] What happens: The server uses an HTML parser (e.g. BeautifulSoup) to extract the text data and remove HTML tags and advertisements.

[0464] Input: The received URL.

[0465] Output: Pure text data.

[0466] Step 4: OCR analysis of image data

[0467] The image data received by the server is analyzed using an OCR engine to extract text information.

[0468] What happens: The server uses an OCR engine (e.g., Tesseract OCR) to extract text information from the image.

[0469] Input: Received image data.

[0470] Output: The extracted text information.

[0471] Step 5: Image content analysis

[0472] The server analyzes the content of the image using an image recognition model.

[0473] Specific operation: The server inputs image data into an image recognition model (e.g., ResNet, YOLO) and identifies the content of the image (people, objects, scenes, etc.).

[0474] Input: Received image data.

[0475] Output: Analyzed image content.

[0476] Step 6: Analyzing text and image data

[0477] The server analyzes the preprocessed text data and the extracted text in the image using natural language processing technology.

[0478] Specific operation: The server inputs text data into a natural language processing model (e.g., BERT, GPT-3) and analyzes the context and content.

[0479] Input: Pure text data and extracted text information.

[0480] Output: Semantic information of the parsed text and image data.

[0481] Step 7: Veracity Assessment

[0482] The server checks the parsed data against a reliable database to assess its veracity.

[0483] Specific operation: The server collates the parsed data via the database API and evaluates its authenticity.

[0484] Input: Parsed text and image data.

[0485] Output: A veracity rating score.

[0486] Step 8: Recognize emotions

[0487] The server analyzes the user's voice data and facial expression data to recognize the user's emotional state.

[0488] Specific operation: The server inputs voice data and facial expression data into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to recognize emotions.

[0489] Input: User's voice and facial expression data.

[0490] Output: The perceived emotional state of the user.

[0491] Step 9: Generate decision results and summary information

[0492] The server generates a judgment result based on the truthfulness evaluation and creates a summarized news.

[0493] Specific operation: The server generates a news summary using natural language generation technology (e.g., GPT-3). The server creates a news summary based on the judgment results.

[0494] Input: Truthfulness rating score.

[0495] Output: Judgment results and summary news.

[0496] Step 10: Coordinating input

[0497] The server provides the results in a coordinated manner.

[0498] Specific behavior: The server adjusts the way information is displayed depending on the user's emotional state, choosing whether to provide detailed or concise information.

[0499] Input: The perceived emotional state of the user and the judgment result.

[0500] Output: Adjusted judgement results and summary news.

[0501] Step 11: Sending the results

[0502] The server transmits the generated judgment result and summary information to the user's terminal.

[0503] Specific behavior: The server sends the result using push notification or HTTP response.

[0504] Input: Adjusted judgement results and summary news.

[0505] Output: The results sent to the terminal.

[0506] Step 12: View the results

[0507] The user checks the judgment results and summary information on the terminal.

[0508] Specific operations: The user uses the device to view the received judgment results and summary information.

[0509] Input: The result sent to the terminal.

[0510] Output: User understanding and judgment.

[0511] This concludes the program processing flow for this system, including the specific operations performed at each step. This enables the system to quickly and accurately evaluate the veracity of news articles and images, and provide information tailored to the user's emotions.

[0512] (Application example 2)

[0513] 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."

[0514] In modern society, the proliferation of fake news and fabricated images has led to the problem of many people making decisions based on incorrect information. This problem is a factor that causes social confusion and misunderstanding. Furthermore, while user emotions have a significant impact on determining the authenticity of information, current systems provide uniform information without taking user emotions into consideration, which reduces user understanding and satisfaction.

[0515] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0516] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their veracity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data and recognizing emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts, adjusting the display method of the result based on the user's emotional state, and transmitting the summary information to the terminal. This makes it easier for users to obtain individually tailored, reliable information without being misled by erroneous information.

[0517] A "news article" is text data and image data containing news information distributed online or offline.

[0518] "Image data" means files containing visual information stored and transmitted in digital format.

[0519] "Natural language processing technology" is a technology that uses computers to understand, interpret, and generate human language.

[0520] A "real-time fact-checking system" is a system that instantly evaluates the veracity of information and provides the results immediately.

[0521] "Data preprocessing" is the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[0522] "Text data" is a data format that includes information about sentences and characters.

[0523] "Extract" refers to the operation of extracting necessary parts from data.

[0524] A "database" is a collection of information that systematically stores large amounts of information and makes it searchable and usable.

[0525] A "reliable source" is an entity or medium that provides information that is highly accurate, reliable, and has been assessed.

[0526] "Emotion recognition" is a technology that analyzes voice data and facial expression data to determine a user's emotional state.

[0527] "Evaluation results" are the conclusions or grades reached after determining the truth or falsity of the information.

[0528] "Summary information" is the result of summarizing a large amount of information in a concise manner.

[0529] A "terminal" is a device such as a computer, smartphone, or tablet that is connected to the Internet or a network.

[0530] "Result presentation method" refers to the means or format in which information is presented to the user.

[0531] The system that realizes this application example evaluates the veracity of news articles and images in real time and provides results according to the user's emotions. Below is a natural language explanation of the program processing of this system.

[0532] The server first receives the URL or image data of the news article from the user's device. The device then sends the input data to the server. This process requires an internet connection, and the user can use a device such as a smartphone or tablet.

[0533] The server retrieves the text data of the news article from the received URL, and then uses a data preprocessing module to remove unnecessary information such as HTML tags and advertisements, converting it into pure text data. This process uses BeautifulSoup as a scraping library and Tesseract OCR as optical character recognition (OCR) technology.

[0534] The server then analyzes the extracted text data and textual information within the images and uses natural language processing techniques (using Hugging Face's Transformer library for its NLP model) to assess their veracity by matching them with trusted databases and other reliable sources, such as public databases and established news sources.

[0535] The server also includes an emotion recognition module to analyze the user's voice and facial expression data to recognize the user's emotional state. For this purpose, it uses OpenCV and the DeepFace library, which allows it to understand the user's emotions, such as surprised, angry, relaxed, etc.

[0536] The server determines whether a news item or image is true or false based on the fact-check results and generates a score. It then generates a news summary based on reliable information, including the results of the fact-check and facts from reliable sources. The server then adjusts the display of the results and summary information based on the user's emotional state, as recognized by the emotion engine. For example, if the user is surprised, the server provides more detailed information to make it easier for the user to understand.

[0537] The results are then sent from the server to the user's device, where the user can view the results and summary information to verify the authenticity of the news article or image.

[0538] Specific examples

[0539] As a concrete example, the following scenario can be considered.

[0540] A user enters the URL of a news article about a politician committing fraud. The server retrieves the news article from the URL and preprocesses the text data. Next, the server analyzes the text using natural language processing technology and compares it with a reliable database. The server determines that the article contains some errors and generates a summary news based on accurate information. Furthermore, the server recognizes the user's current emotions and provides results with detailed supplementary information if the user is surprised. The user can check the results and summary on their device.

[0541] Prompt Sentence Examples

[0542] The user inputs the URL or image data of a news article. The system receives the URL or image data and performs preprocessing. It then analyzes the data using natural language processing techniques and compares it with a reliable database. It recognizes the user's emotion (e.g., joy) and generates and displays results according to that emotion.

[0543] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0544] Step 1:

[0545] The user inputs the URL and image data of the news article via the terminal.

[0546] Input: News article URL or image data

[0547] How it works: A user provides data through an input form on a device such as a smartphone or tablet.

[0548] Output: The entered data is sent from the terminal to the server.

[0549] Step 2:

[0550] The server receives the URL or image data of the news article.

[0551] Input: URL of news article or image data sent from the device

[0552] How it works: The server receives the HTTP request and stores the data locally.

[0553] Output: The received data is stored on the server and handed over to the next process.

[0554] Step 3:

[0555] The server preprocesses the data received.

[0556] Input: URL or image data of a saved news article

[0557] How it works: For news articles, it uses BeautifulSoup to remove HTML tags and ads and extract pure text data. For image data, it uses Tesseract OCR to extract text information from images.

[0558] Output: Preprocessed text data or text information in images.

[0559] Step 4:

[0560] The server analyzes the preprocessed data.

[0561] Input: Preprocessed text data or text information in images

[0562] How it works: Uses Hugging Face's Transformer library for natural language processing to understand the context and content of data.

[0563] Output: Parsed data.

[0564] Step 5:

[0565] The server checks the parsed data against a reliable database to assess its veracity.

[0566] Input: Parsed data

[0567] How it works: The parsed data is compared to public databases and established news sources to generate a credibility score.

[0568] Output: A truthfulness score.

[0569] Step 6:

[0570] The server analyzes the user's voice data and facial expression data to recognize their emotional state.

[0571] Input: User's voice data and facial expression data

[0572] How it works: It uses OpenCV and DeepFace libraries to analyze user emotions and identify their emotional state.

[0573] Output: The user's emotional state.

[0574] Step 7:

[0575] The server makes a decision based on the results of the fact check.

[0576] Input: Truthfulness score, user emotional state

[0577] How it works: Based on the evaluation results, it determines whether a news item or image is true or false, generates an appropriate score, and adjusts how the results are displayed based on the user's emotional state.

[0578] Output: Judgment results and how they are displayed.

[0579] Step 8:

[0580] The server generates the judgment result and summary information and transmits them to the terminal.

[0581] Input: Judgment result, display method

[0582] How it works: The server generates news summaries based on reliable information, presents them in a format that is easy for users to understand, and sends the adjusted results to the user's device.

[0583] Output: The judgment result and summary information sent to the user's terminal.

[0584] Step 9:

[0585] The user views the provided results and summary.

[0586] Input: Judgment result and summary information sent from the server

[0587] How it works: Users check the information displayed on their smartphones, tablets, or other devices to verify the authenticity of news articles and images.

[0588] Output: User understanding and judgment.

[0589] 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.

[0590] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0591] 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.

[0592] [Second embodiment]

[0593] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0594] 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.

[0595] 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).

[0596] 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.

[0597] 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.

[0598] 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).

[0599] 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.

[0600] 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.

[0601] 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.

[0602] 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.

[0603] In the smart glasses 214, 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.

[0604] 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."

[0605] The present invention relates to a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system performs a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[0606] System Overview

[0607] The system includes the following main components to assess the veracity of news articles and images:

[0608] 1. Data Entry Module

[0609] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[0610] The terminal transmits the input data to the server.

[0611] 2. Data Preprocessing Module

[0612] The server removes HTML tags and unnecessary data from the received news articles and extracts the pure text data.

[0613] The server extracts text information from the image data using OCR technology and analyzes the content of the image using image recognition technology.

[0614] 3. Fact-checking module

[0615] The server analyzes the extracted text data and text information in the image using natural language processing technology.

[0616] The server compares the text data with reliable databases and news sources to assess its veracity.

[0617] 4. Judgment result generation module

[0618] Based on the results of the fact check, the server determines whether the news or image is true or false.

[0619] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[0620] 5. Result Delivery Module

[0621] The server transmits the generated judgment result and summary information to the user's terminal.

[0622] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[0623] Program processing

[0624] The program processing of this system will be explained in natural language below.

[0625] 1. Receiving Data

[0626] A user uses a terminal to input the URL or image data of a news article into the system.

[0627] The terminal sends the input data to the server.

[0628] 2. Data Preprocessing

[0629] The server retrieves the text data of the news article from the received URL.

[0630] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[0631] For image data, the server uses OCR technology to extract text information within the image.

[0632] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[0633] 3. Conduct fact-checks

[0634] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[0635] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[0636] 4. Generating the judgment result

[0637] The server generates a score based on the fact-check results to determine whether a news article or image is true or false.

[0638] The server generates news summaries based on reliable information.

[0639] 5. Providing results

[0640] The server transmits the judgment result and summary information to the user's terminal.

[0641] Users can view the results on their devices and verify the authenticity of the information.

[0642] Specific examples

[0643] 1. News article example

[0644] The user enters the URL of a news article that says, "A politician committed fraud."

[0645] The server retrieves the news article from the URL and preprocesses the text data.

[0646] The server uses natural language processing to analyze the content of news articles and compare them with a reliable database.

[0647] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[0648] Users can check the judgment results and summary on their device.

[0649] 2. Example images

[0650] Users upload "photos of specific incidents."

[0651] The server analyzes the image data using OCR technology and extracts text information.

[0652] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0653] The server determines that the image is misleading and generates a factual explanation.

[0654] Users can check the judgment results and summary explanation on their terminal.

[0655] As described above, the present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and images in real time and quickly provide accurate information to users, allowing users to make decisions based on reliable information without being misled by potentially misleading information.

[0656] The processing flow will be explained below.

[0657] Step 1:

[0658] The user inputs the URL or image data of a news article into the device.

[0659] Step 2:

[0660] The device sends the entered URL and image data to the server.

[0661] Step 3:

[0662] The server scrapes text data from the URL of the news article received.

[0663] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[0664] Step 4:

[0665] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[0666] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[0667] Step 5:

[0668] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[0669] Step 6:

[0670] The server compares the analyzed data with reliable databases and news sources (e.g., public fact-checking sites) to assess the veracity of the information.

[0671] Step 7:

[0672] Based on the results of the fact check, the server scores the news article or image as true or false.

[0673] Step 8:

[0674] The server generates an accurate news summary, including judgments and facts from reliable sources.

[0675] Step 9:

[0676] The server transmits the generated judgment result and summary information to the user's terminal.

[0677] Step 10:

[0678] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0679] Example 1

[0680] 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."

[0681] Conventional fact-checking systems for news articles and image data often took a long time to evaluate the veracity of articles or based their judgments on unreliable sources. As a result, users were at a higher risk of making decisions based on incorrect information, making it difficult to provide reliable information. Furthermore, inaccurate analysis was not possible due to insufficient preprocessing, such as removing HTML tags from news articles or using optical character recognition technology for image data.

[0682] 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.

[0683] In this invention, the server includes: a means for a user to input a URL or image data of a news article from a terminal and transmit the data from the terminal to the server; a means for acquiring text data from the URL of the news article received by the server or extracting text information from the image data using optical character recognition technology; a means for preprocessing the acquired text data and removing HTML tags and unnecessary data; a means for analyzing the preprocessed text data or image data using natural language processing technology and comparing it with a reliable database or other reliable information source to evaluate its veracity; a means for determining whether the news article or image is true or false based on the comparison result; and a means for generating summary information based on the determination result and facts and transmitting it to the user's terminal. This enables the veracity of news articles and image data to be evaluated with high accuracy in real time and provides accurate information to users quickly.

[0684] A "user" is a person who inputs the URL of a news article or image data into the system via a terminal.

[0685] A "terminal" is a device that receives news article URLs and image data from users and sends them to a server.

[0686] "Server" means a device that processes and analyzes data received from users, evaluates its authenticity, and generates and provides results.

[0687] A "news article" is a web page containing news information.

[0688] "URL" refers to the web address of a news article.

[0689] "Image data" refers to an image file provided by the user.

[0690] "Optical character recognition technology" is a technology that analyzes and extracts text information from images.

[0691] An "HTML tag" is an element of a markup language that defines the structure and content of a web page.

[0692] "Natural language processing technology" is a technology that enables computers to understand, generate, and analyze human language.

[0693] A "database" is a system for systematically storing and managing information.

[0694] "Veracity" refers to whether a news article or image data is based on facts.

[0695] A "judgment score" is a numerical value that is displayed as a result of evaluating the veracity of a news article or image data.

[0696] "Summary information" is information that briefly summarizes the judgment results of news articles or image data.

[0697] "Preprocessing" is the process of preparing data before analysis or evaluation.

[0698] A "reliable database" is a database built on accurate and reliable sources of information.

[0699] "Evaluation" means analyzing data and determining its veracity.

[0700] "Fact checking" is the process of verifying whether the contents of a news article or image data are based on facts.

[0701] This invention is a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system executes a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[0702] The system is roughly divided into a server, a terminal, and a user. First, a user inputs the URL of a news article and image data from their own terminal (PC, smartphone, tablet, etc.). The input data is sent from the terminal to the server.

[0703] The server performs various preprocessing operations on the received data. For news article URLs, web scraping techniques are used to remove HTML tags and obtain pure text data (e.g., using libraries such as BeautifulSoup). For image data, optical character recognition techniques (e.g., Tesseract) are used to extract text information, and image recognition techniques (e.g., OpenCV or TensorFlow) are used to analyze the image content.

[0704] The server then analyzes the preprocessed text data and the text information in the images using natural language processing techniques (e.g., spaCy or NLTK). The analysis results are then cross-referenced with reliable databases and other trusted sources. This cross-reference uses APIs and web crawlers to obtain real-time data (e.g., Wikipedia API or official news API).

[0705] As a result of the matching, the server evaluates the veracity of the news article or image data and generates a judgment score. The server then generates a summary based on reliable information and provides it to the user in an easy-to-understand format. A generative AI model (e.g., GPT-3) is used to generate this summary.

[0706] Finally, the server sends the judgment result and summary information to the user's device. The user can view the judgment result and summary information on their device and confirm the authenticity of the news or image data. This series of processes allows the user to make decisions based on reliable information without being misled by potentially misleading information.

[0707] Specific example explanation

[0708] News article example

[0709] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[0710] 2. The device sends this URL to the server.

[0711] 3. The server retrieves the news article from the URL and preprocesses the text data.

[0712] 4. The server uses natural language processing to analyze the content of the news article and compare it with a reliable database.

[0713] 5. The server determines that the article contains some errors and generates a summary news based on accurate information.

[0714] 6. The user checks the judgment results and summary on the terminal.

[0715] Image example

[0716] 1. The user uploads a photo of a specific incident.

[0717] 2. The terminal sends this image data to the server.

[0718] 3. The server analyzes the image data using OCR technology and extracts text information.

[0719] 4. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0720] 5. The server determines that the image is misleading and generates a factual explanation.

[0721] 6. The user checks the judgment result and summary explanation on the terminal.

[0722] Prompt Sentence Examples

[0723] Example news article prompt:

[0724] Enter the URL of the following news article and our system will analyze its content and rate its veracity:

[0725] Example: https: / / example.com / news / article123

[0726] Image prompt example:

[0727] Please upload the following image and our system will analyze its content and rate its authenticity:

[0728] Example: (Drag and drop image file)

[0729] The above is an embodiment of the present invention. The present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and image data in real time and quickly provide accurate information to users.

[0730] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0731] Step 1: Receiving Data

[0732] The user enters the URL or image data of a news article into an input form on the device, for example, by copying and pasting the URL of a news site or by dragging and dropping an image file.

[0733] The terminal receives the input data and sends it to the server as an HTTP request. Specifically, it converts the input data into JSON format and generates a POST request. This request is added to the processing queue.

[0734] Input: News article URL or image data

[0735] Output: HTTP request to the server

[0736] Step 2: Data preprocessing (analysis of news articles)

[0737] The server retrieves the web page content from the URL of the news article received using a web scraping technique (e.g., BeautifulSoup).

[0738] The server removes unnecessary information such as HTML tags and advertisements from the HTML pages it retrieves, and extracts the text data. Specifically, it uses an HTML parser to filter out tag information and extract only the pure text data.

[0739] Input: News article URL

[0740] Output: Preprocessed text data

[0741] Step 2: Data preprocessing (analysis of image data)

[0742] The server uses optical character recognition technology (e.g., Tesseract) on the received image data to extract text information.

[0743] The server uses image recognition technology (e.g., OpenCV or TensorFlow) to analyze the content of the image and identify relevant information, such as identifying faces and objects contained within the image.

[0744] Input: Image data

[0745] Output: Extracted text information and image content data

[0746] Step 3: Conduct a fact check

[0747] The server analyzes the preprocessed text data or the text information in the images using natural language processing techniques (e.g., spaCy or NLTK), performing syntactic and semantic analysis of the text to extract important keywords and phrases.

[0748] The server cross-checks the parsed data against reliable databases and other trusted sources, using APIs and web crawlers to retrieve real-time data (e.g., Wikipedia API or official news APIs).

[0749] Input: Preprocessed text or image data

[0750] Output: Matching result

[0751] Step 4: Generate a decision result

[0752] The server generates a score based on the fact-check results to evaluate the veracity of the news article or image. Specifically, it calculates the degree of match between the matching results and calculates a score ranging from 0 to 100.

[0753] The server uses a generative AI model (e.g., GPT-3) to generate a summary based on reliable information, including a rating of veracity and relevant factual information.

[0754] Input: Matching result

[0755] Output: Judgment score and summary information

[0756] Step 5: Delivering results

[0757] The server sends the generated judgment result and summary information to the user's device using an HTTP response. Specifically, the generated information is encoded in JSON format and returned as an HTTP response.

[0758] The user checks the judgment results and summary information on the device. For example, the results are displayed in the display area of ​​a web browser. Based on this, the user can determine the authenticity of the news article or image.

[0759] Input: Judgment score and summary information

[0760] Output: Providing results to the user

[0761] (Application example 1)

[0762] 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."

[0763] Conventional systems for assessing the veracity of news articles and images have difficulty assessing the credibility of information entered by users in real time, and providing assessment results quickly via smartphones has been a particular challenge. Furthermore, there is no well-established mechanism for notifying users of misinformation in real time, and immediate comparison with reliable sources is often not performed. This creates a risk that users will make decisions based on incorrect information.

[0764] 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.

[0765] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and comparing it with a reliable database or other reliable information sources to evaluate their veracity, means for determining whether the news or images are true or false based on the evaluation results, means for generating summary information based on the evaluation results and facts and sending it to the terminal, and means for notifying the user of the generated summary information and the evaluation results in real time via an application installed on the smartphone. This enables immediate evaluation of the veracity of news articles and images and the provision of the results in real time.

[0766] A "news article" is a document created to inform the public of various events or information.

[0767] "Image data" is visual information stored in digital format, including photographs and illustrations.

[0768] A "terminal" is a device used by a user to enter information and view results, typically a smartphone or computer.

[0769] "Means for receiving" refers to a mechanism for acquiring news articles and image data sent from a terminal.

[0770] "Preprocessing" refers to the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[0771] "Text data" is digital data that contains text information.

[0772] Optical character recognition (OCR) is a technology for extracting text information from an image as digital character data.

[0773] "Natural language processing technology" is a technology that allows computers to understand and process human language.

[0774] A "trusted database" is a data repository that stores reliable information and primarily contains data from trusted sources.

[0775] A "source" is a data source that provides particular information.

[0776] A "means for assessing veracity" is a mechanism for determining whether the data is true by analyzing preprocessed data and comparing it with reliable sources of information.

[0777] "Evaluation results" refer to the results of a truthfulness analysis, indicating whether the news or image is true or false.

[0778] "Summary information" is information created based on the evaluation results in a concise, easy-to-understand format.

[0779] "Means of real-time notification" refers to a mechanism for instantly conveying the generated summary information and evaluation results to users, and is primarily implemented through smartphone applications.

[0780] An "application" is a software program that runs on a computer to perform a particular function.

[0781] A description will be given of an embodiment of the present invention. A system for evaluating the veracity of news articles and images in real time and providing the results to a user comprises the following elements.

[0782] 1. System Configuration

[0783] This real-time fact-checking system mainly includes the following components:

[0784] Data Entry Module

[0785] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[0786] Data entered from the terminal is transmitted to the server.

[0787] Data Preprocessing Module

[0788] The server retrieves text data from the URL of the received news article and removes unnecessary information such as HTML tags and advertisements.

[0789] For image data, the server uses optical character recognition technology (OCR) to extract text information within the image.

[0790] The text data is automatically cleaned and converted into a format suitable for analysis.

[0791] Fact Checking Module

[0792] The server analyzes the cleaned text data or information in the images using natural language processing techniques.

[0793] The recognized text data is checked against reliable databases and other trusted sources to assess its veracity.

[0794] Judgment result generation module

[0795] The veracity of news and images is determined based on the results of fact-checking.

[0796] Judgments and fact-based summaries are automatically generated.

[0797] Result Delivery Module

[0798] The server transmits the generated judgment result and summary information to the user's terminal.

[0799] An application installed on a smartphone notifies the user of this information in real time.

[0800] The user can view the results and their summary through the terminal.

[0801] 2. Hardware and Software Requirements

[0802] The system uses a variety of hardware and software tools.

[0803] Hardware

[0804] Devices: smartphones, computers

[0805] Server: A server with high-performance processing capabilities

[0806] software

[0807] OCR technology: software for extracting text information from images (e.g., Tesseract OCR)

[0808] Natural language processing libraries: Tools for analyzing news articles and text data (e.g., SpaCy and NLTK)

[0809] Database connection software: Libraries for communicating with reliable databases

[0810] Smartphone app: An application that provides users with real-time information

[0811] Generative AI models: AI models used to generate summaries and analysis results for text data.

[0812] 3. Specific Examples

[0813] Specific usage scenarios include:

[0814] News article example

[0815] The user enters the URL of a news article that purports to say that a certain politician committed fraud.

[0816] The server retrieves the news article from the URL and preprocesses the text data.

[0817] The content of news articles is analyzed using natural language processing and compared with a reliable database.

[0818] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[0819] Users can check the assessment results and summary in real time using a smartphone application.

[0820] Image example

[0821] Users upload "photos of specific incidents."

[0822] The server analyzes the image data using OCR technology and extracts text information.

[0823] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0824] The server determines that the image is misleading and generates a factual explanation.

[0825] Users can check the assessment results and summary explanations in real time using a smartphone application.

[0826] 4. Examples of prompts

[0827] An example of a prompt that a user might enter into the system is:

[0828] "https: / / example.com / news-article Check whether this news article is true or not."

[0829] "Please verify this image is authentic."

[0830] This allows users to identify misinformation in real time and make quick decisions based on trusted sources of information.

[0831] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0832] Step 1:

[0833] User enters the URL or image data of a news article

[0834] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[0835] Input: News article URL or image data

[0836] Output: Input data is sent from the device to the server

[0837] Step 2:

[0838] The server preprocesses the received data

[0839] The server retrieves text data from the URLs of received news articles, removing unnecessary information such as HTML tags and advertisements, and extracting pure text data.

[0840] For image data, OCR technology is used to extract text information within the image.

[0841] Input: User-entered news article URL or image data

[0842] Output: Preprocessed text data or text information in images

[0843] Step 3:

[0844] The server analyzes the data using natural language processing technology

[0845] The server then analyzes the preprocessed text data and text information in the images using natural language processing techniques, which includes extracting words, understanding context, and summarizing information.

[0846] Input: Preprocessed text data or text information in images

[0847] Output: Parsed text data

[0848] Step 4:

[0849] The server checks the authenticity against a reliable database

[0850] The server then compares the parsed text data with reliable databases and other trusted sources, assessing the authenticity of the news article or image.

[0851] Input: Parsed text data

[0852] Output: Veracity evaluation result

[0853] Step 5:

[0854] The server generates the judgment result and summary information

[0855] The server generates a score based on the fact-check evaluation results to determine whether a news article or image is true or false, and creates a fact-based summary based on the evaluation results.

[0856] Input: Veracity assessment result

[0857] Output: Judgment results and summary information

[0858] Step 6:

[0859] The server sends the results to the device

[0860] The server then sends the generated judgment results and summary information to the user's device, and an application installed on the smartphone notifies the user in real time.

[0861] Input: Judgment results and summary information

[0862] Output: Notification and display on the user's terminal

[0863] Step 7:

[0864] User checks the results

[0865] Users can check the results and summary information sent from the server through a smartphone application, allowing them to determine the authenticity of news articles and images.

[0866] Input: Notification from the server and displayed results

[0867] Output: User's understanding and judgment

[0868] The above steps realize a system that evaluates the veracity of news articles and images in real time and provides appropriate information to users.

[0869] 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.

[0870] The present invention relates to a system that uses natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provides results based on the user's emotions. This system performs a series of processes, including preprocessing, analysis, fact-checking, emotion recognition, and generating and providing judgment results for news articles or image data.

[0871] System Overview

[0872] The system includes the following main components:

[0873] 1. Data Entry Module

[0874] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[0875] The terminal transmits the input data to the server.

[0876] 2. Data Preprocessing Module

[0877] The server removes unnecessary data such as HTML tags and advertisements from the received news articles and extracts pure text data.

[0878] The server uses optical character recognition (OCR) technology on the image data to extract text information within the image.

[0879] The server analyzes the content of the image using image recognition technology.

[0880] 3. Fact-checking module

[0881] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[0882] The server checks the text data against reliable databases and other trusted sources to assess its veracity.

[0883] 4. Emotion Engine Module

[0884] The server analyzes the user's voice data and facial expression data acquired through the user's terminal and recognizes emotions.

[0885] The server adjusts the method of providing the judgment results and summary information based on the user's emotions recognized by the emotion engine.

[0886] 5. Judgment Result Generation Module

[0887] The server determines whether the news or image is true or false based on the results of the fact check.

[0888] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[0889] 6. Result Delivery Module

[0890] The server transmits the generated judgment result and summary information to the user's terminal.

[0891] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[0892] Program processing

[0893] The program processing of this system will be explained in natural language below.

[0894] 1. Receiving Data

[0895] A user uses a terminal to input the URL or image data of a news article into the system.

[0896] The terminal sends the input data to the server.

[0897] 2. Data Preprocessing

[0898] The server obtains the text data of the news article from the received URL.

[0899] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[0900] The server analyzes the image data using OCR technology and extracts text information.

[0901] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[0902] 3. Conduct fact-checks

[0903] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[0904] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[0905] 4. Emotion Recognition by Emotion Engine

[0906] The server uses an emotion engine to analyze the user's voice data and facial expression data acquired through the user's device, and recognizes the user's current emotions.

[0907] For example, the emotional state of the user, such as whether they are surprised, angry, or relaxed, is grasped.

[0908] 5. Generating the judgment result

[0909] Based on the results of the fact check, the server generates a score that determines whether the news article or image is true or false.

[0910] The server generates a news summary based on reliable information, including judgments and facts from reliable sources.

[0911] 6. Tailored delivery of results

[0912] The server adjusts the display method of the judgment result and summary information based on the user's emotional state recognized by the emotion engine.

[0913] For example, if the user is surprised, provide more detailed information to make it easier for the user to understand.

[0914] 7. Providing Results

[0915] The server transmits the generated judgment result and summary information to the user's terminal.

[0916] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0917] Specific examples

[0918] 1. News article example

[0919] A user enters the URL of a news article that says, "A politician committed fraud."

[0920] The server retrieves the news article from the URL and preprocesses the text data.

[0921] The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[0922] The server determines that this article contains some errors and generates a summary news based on accurate information.

[0923] The server recognizes the user's current emotion and, if they are surprised for example, provides a result with detailed supporting information.

[0924] Users can check the judgment results and summary on their device.

[0925] 2. Example images

[0926] A user uploads a "photo of a specific incident."

[0927] The server analyzes the image data using OCR technology and extracts text information.

[0928] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[0929] The server determines that the image is misleading and generates a factual explanation.

[0930] The server recognizes the user's current emotions and, for example, if they are anxious, provides results that are thoughtfully reassuring.

[0931] Users can check the judgment results and summary explanation on their terminal.

[0932] As described above, the present invention is a system that utilizes natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provide information that is individually tailored to the user's emotions. As a result, users can make decisions based on reliable information without being misled by potentially misleading information. Furthermore, by taking the user's emotions into consideration, a more user-friendly and easy-to-use system can be realized.

[0933] The processing flow will be explained below.

[0934] Step 1:

[0935] The user inputs the URL or image data of a news article into the device.

[0936] Step 2:

[0937] The device sends the entered URL and image data to the server.

[0938] Step 3:

[0939] The server scrapes text data from the URL of the news article received.

[0940] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[0941] Step 4:

[0942] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[0943] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[0944] Step 5:

[0945] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[0946] Step 6:

[0947] The server cross-checks the parsed data against reliable databases and other trusted sources (e.g., public fact-checking sites) to assess the veracity of the information.

[0948] Step 7:

[0949] Based on the results of the fact check, the server scores the news article or image as true or false.

[0950] Step 8:

[0951] The server uses an emotion engine to analyze the voice and facial expression data it receives from the user's device and recognizes the user's current emotions.

[0952] Step 9:

[0953] The server adjusts the method of providing the judgment results and summary information based on the user's emotional state.

[0954] For example, if the user is surprised, add more information in an easy-to-understand format.

[0955] Step 10:

[0956] The server generates an accurate news summary, including judgments and facts from reliable sources.

[0957] Step 11:

[0958] The server transmits the generated judgment result and summary information to the user's terminal.

[0959] Step 12:

[0960] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[0961] Example 2

[0962] 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."

[0963] In modern society, quickly and accurately assessing the veracity of news articles and images has become an important issue. The inability to instantly determine the reliability of such information can lead to the spread of false information and misleading images, which can have a significant impact on society. Furthermore, as the amount of information increases, it is becoming increasingly difficult for users to quickly verify the authenticity of information. Furthermore, information provided does not take into account the user's feelings, which can lead to misunderstandings and anxiety.

[0964] 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.

[0965] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their authenticity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data to recognize emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts and transmitting it to the terminal in a form that corresponds to the user's emotional state. This makes it possible to quickly and accurately evaluate the authenticity of news articles and images and provide information that corresponds to the user's emotions.

[0966] A "news article" is text data containing current information and news reports that are primarily published on websites and digital media.

[0967] "Image data" refers to file formats (e.g., JPEG, PNG) that contain visual information stored in a digital format.

[0968] A "terminal" is an electronic device such as a computer or smartphone that is used to input news articles and image data and view the results.

[0969] "Server" is a central control device that provides resources for data processing and analysis and manages the operation of the entire system.

[0970] "Preprocessing" refers to the process of removing unnecessary information from news articles or image data and converting them into a format suitable for analysis.

[0971] "Text data" refers to purely textual information extracted from news articles or image data.

[0972] "Optical character recognition technology" refers to technology (e.g., OCR) used to convert text in an image into digital characters.

[0973] "Natural language processing technology" refers to computer science technology for analyzing the meaning of text data and understanding its context and content.

[0974] A "trusted database" is a data repository containing verified information sources that can be used to assess veracity.

[0975] "Emotion recognition technology" is a technology that analyzes a user's voice data and facial expression data to understand their emotional state.

[0976] "Decision result" refers to the conclusion reached after evaluating the veracity of news articles or image data.

[0977] "Summary information" refers to information that succinctly summarizes the important points of a news article or image data.

[0978] A "reliable source" is a source or database that is widely recognized as providing truthful information.

[0979] This invention relates to a real-time fact-checking system that utilizes natural language processing and emotion recognition technologies. This system evaluates the veracity of news articles and image data and provides information according to the user's emotions, thereby realizing the provision of highly reliable information.

[0980] System configuration and operation

[0981] The system includes the following major components:

[0982] server

[0983] Device (PC or smartphone)

[0984] Receiving data

[0985] The user inputs the URL and image data of the news article into the terminal, which then sends it to the server.

[0986] Data Preprocessing

[0987] The server downloads the HTML data from the URL of the received news article and extracts the text data using an HTML parser (e.g., BeautifulSoup). It then removes unnecessary information, such as HTML tags and advertisements, from the text data. For image data, it extracts the text information from the image using an OCR engine (e.g., Tesseract OCR) and analyzes the image content using an image recognition model (e.g., ResNet, YOLO).

[0988] Data analysis and evaluation

[0989] The server inputs the preprocessed text data and information in the images into natural language processing models (e.g., BERT, GPT-3) to analyze the context and content, and then compares the information with reliable databases (e.g., Wikidata, FactCheck.org) and news sources via API to assess the veracity of the information.

[0990] Emotion recognition

[0991] The server inputs the voice data and facial expression images sent from the device into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to understand the user's emotional state. For example, it recognizes emotions such as surprise, anger, or relaxation.

[0992] Generation of judgment results and summary information

[0993] The server determines whether a news article or image is true or false based on the fact-check results and generates a truthfulness score. The server then uses natural language generation technology (e.g., GPT-3) to create a news summary, which includes the results of the judgment and facts based on reliable sources.

[0994] Providing information

[0995] The server adjusts the display method of the judgment results and summary information according to the user's emotional state. For example, if the user is surprised, it provides more detailed information, and if the user is relaxed, it provides simple information. The generated information is sent to the terminal, where the user can view it.

[0996] Specific examples

[0997] News article example

[0998] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[0999] 2. The server retrieves the news article from the URL and preprocesses the text data.

[1000] 3. The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[1001] 4. The server determines that this article contains some errors and generates a summary news article based on accurate information.

[1002] 5. The server recognizes the user's emotion and provides a result with detailed supporting information if, for example, they are surprised.

[1003] 6. The user checks the judgment results and summary on the terminal.

[1004] Image example

[1005] 1. A user uploads a photo of a specific incident.

[1006] 2. The server analyzes the image data using OCR technology and extracts text information.

[1007] 3. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1008] 4. The server determines that the image is potentially misleading and generates a factual explanation.

[1009] 5. The server recognizes the user's emotions and provides reassuring information if, for example, the user is worried.

[1010] 6. The user checks the judgment result and summary explanation on the terminal.

[1011] Example prompt sentence:

[1012] The user is prompted to enter information in a format such as "Please enter the URL of a news article" or "Please upload image data."

[1013] As described above, the present invention makes it possible to evaluate the veracity of news articles and images in real time and provide information adjusted according to the user's emotions.

[1014] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1015] Program processing steps

[1016] Step 1: Receiving Data

[1017] A user uses a device to input a URL or image data of a news article.

[1018] Specific actions: Paste or upload the URL of a news article or image file into the input form displayed on the device.

[1019] Input: News article URL or image data.

[1020] Output: The received URL or image data.

[1021] Step 2: Sending data

[1022] The terminal transmits the input data to the server.

[1023] Specific operation: The device sends an HTTP POST request to the server, including the received URL or image data.

[1024] Input: Received URL or image data.

[1025] Output: URL or image data received by the server.

[1026] Step 3: Acquiring and preprocessing news articles

[1027] The server retrieves the text data of the news article from the received URL and removes unnecessary information.

[1028] What happens: The server uses an HTML parser (e.g. BeautifulSoup) to extract the text data and remove HTML tags and advertisements.

[1029] Input: The received URL.

[1030] Output: Pure text data.

[1031] Step 4: OCR analysis of image data

[1032] The image data received by the server is analyzed using an OCR engine to extract text information.

[1033] What happens: The server uses an OCR engine (e.g., Tesseract OCR) to extract text information from the image.

[1034] Input: Received image data.

[1035] Output: The extracted text information.

[1036] Step 5: Image content analysis

[1037] The server analyzes the content of the image using an image recognition model.

[1038] Specific operation: The server inputs image data into an image recognition model (e.g., ResNet, YOLO) and identifies the content of the image (people, objects, scenes, etc.).

[1039] Input: Received image data.

[1040] Output: Analyzed image content.

[1041] Step 6: Analyzing text and image data

[1042] The server analyzes the preprocessed text data and the extracted text in the image using natural language processing technology.

[1043] Specific operation: The server inputs text data into a natural language processing model (e.g., BERT, GPT-3) and analyzes the context and content.

[1044] Input: Pure text data and extracted text information.

[1045] Output: Semantic information of the parsed text and image data.

[1046] Step 7: Veracity Assessment

[1047] The server checks the parsed data against a reliable database to assess its veracity.

[1048] Specific operation: The server collates the parsed data via the database API and evaluates its authenticity.

[1049] Input: Parsed text and image data.

[1050] Output: A veracity rating score.

[1051] Step 8: Recognize emotions

[1052] The server analyzes the user's voice data and facial expression data to recognize the user's emotional state.

[1053] Specific operation: The server inputs voice data and facial expression data into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to recognize emotions.

[1054] Input: User's voice and facial expression data.

[1055] Output: The perceived emotional state of the user.

[1056] Step 9: Generate decision results and summary information

[1057] The server generates a judgment result based on the truthfulness evaluation and creates a summarized news.

[1058] Specific operation: The server generates a news summary using natural language generation technology (e.g., GPT-3). The server creates a news summary based on the judgment results.

[1059] Input: Truthfulness rating score.

[1060] Output: Judgment results and summary news.

[1061] Step 10: Coordinating input

[1062] The server provides the results in a coordinated manner.

[1063] Specific behavior: The server adjusts the way information is displayed depending on the user's emotional state, choosing whether to provide detailed or concise information.

[1064] Input: The perceived emotional state of the user and the judgment result.

[1065] Output: Adjusted judgement results and summary news.

[1066] Step 11: Sending the results

[1067] The server transmits the generated judgment result and summary information to the user's terminal.

[1068] Specific behavior: The server sends the result using push notification or HTTP response.

[1069] Input: Adjusted judgement results and summary news.

[1070] Output: The results sent to the terminal.

[1071] Step 12: View the results

[1072] The user checks the judgment results and summary information on the terminal.

[1073] Specific operations: The user uses the device to view the received judgment results and summary information.

[1074] Input: The result sent to the terminal.

[1075] Output: User understanding and judgment.

[1076] This concludes the program processing flow for this system, including the specific operations performed at each step. This enables the system to quickly and accurately evaluate the veracity of news articles and images, and provide information tailored to the user's emotions.

[1077] (Application example 2)

[1078] 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."

[1079] In modern society, the proliferation of fake news and fabricated images has led to the problem of many people making decisions based on incorrect information. This problem is a factor that causes social confusion and misunderstanding. Furthermore, while user emotions have a significant impact on determining the authenticity of information, current systems provide uniform information without taking user emotions into consideration, which reduces user understanding and satisfaction.

[1080] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1081] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their veracity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data and recognizing emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts, adjusting the display method of the result based on the user's emotional state, and transmitting the summary information to the terminal. This makes it easier for users to obtain individually tailored, reliable information without being misled by erroneous information.

[1082] A "news article" is text data and image data containing news information distributed online or offline.

[1083] "Image data" means files containing visual information stored and transmitted in digital format.

[1084] "Natural language processing technology" is a technology that uses computers to understand, interpret, and generate human language.

[1085] A "real-time fact-checking system" is a system that instantly evaluates the veracity of information and provides the results immediately.

[1086] "Data preprocessing" is the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[1087] "Text data" is a data format that includes information about sentences and characters.

[1088] "Extract" refers to the operation of extracting necessary parts from data.

[1089] A "database" is a collection of information that systematically stores large amounts of information and makes it searchable and usable.

[1090] A "reliable source" is an entity or medium that provides information that is highly accurate, reliable, and has been assessed.

[1091] "Emotion recognition" is a technology that analyzes voice data and facial expression data to determine a user's emotional state.

[1092] "Evaluation results" are the conclusions or grades reached after determining the truth or falsity of the information.

[1093] "Summary information" is the result of summarizing a large amount of information in a concise manner.

[1094] A "terminal" is a device such as a computer, smartphone, or tablet that is connected to the Internet or a network.

[1095] "Result presentation method" refers to the means or format in which information is presented to the user.

[1096] The system that realizes this application example evaluates the veracity of news articles and images in real time and provides results according to the user's emotions. Below is a natural language explanation of the program processing of this system.

[1097] The server first receives the URL or image data of the news article from the user's device. The device then sends the input data to the server. This process requires an internet connection, and the user can use a device such as a smartphone or tablet.

[1098] The server retrieves the text data of the news article from the received URL, and then uses a data preprocessing module to remove unnecessary information such as HTML tags and advertisements, converting it into pure text data. This process uses BeautifulSoup as a scraping library and Tesseract OCR as optical character recognition (OCR) technology.

[1099] The server then analyzes the extracted text data and textual information within the images and uses natural language processing techniques (using Hugging Face's Transformer library for its NLP model) to assess their veracity by matching them with trusted databases and other reliable sources, such as public databases and established news sources.

[1100] The server also includes an emotion recognition module to analyze the user's voice and facial expression data to recognize the user's emotional state. For this purpose, it uses OpenCV and the DeepFace library, which allows it to understand the user's emotions, such as surprised, angry, relaxed, etc.

[1101] The server determines whether a news item or image is true or false based on the fact-check results and generates a score. It then generates a news summary based on reliable information, including the results of the fact-check and facts from reliable sources. The server then adjusts the display of the results and summary information based on the user's emotional state, as recognized by the emotion engine. For example, if the user is surprised, the server provides more detailed information to make it easier for the user to understand.

[1102] The results are then sent from the server to the user's device, where the user can view the results and summary information to verify the authenticity of the news article or image.

[1103] Specific examples

[1104] As a concrete example, the following scenario can be considered.

[1105] A user enters the URL of a news article about a politician committing fraud. The server retrieves the news article from the URL and preprocesses the text data. Next, the server analyzes the text using natural language processing technology and compares it with a reliable database. The server determines that the article contains some errors and generates a summary news based on accurate information. Furthermore, the server recognizes the user's current emotions and provides results with detailed supplementary information if the user is surprised. The user can check the results and summary on their device.

[1106] Prompt Sentence Examples

[1107] The user inputs the URL or image data of a news article. The system receives the URL or image data and performs preprocessing. It then analyzes the data using natural language processing techniques and compares it with a reliable database. It recognizes the user's emotion (e.g., joy) and generates and displays results according to that emotion.

[1108] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1109] Step 1:

[1110] The user inputs the URL and image data of the news article via the terminal.

[1111] Input: News article URL or image data

[1112] How it works: A user provides data through an input form on a device such as a smartphone or tablet.

[1113] Output: The entered data is sent from the terminal to the server.

[1114] Step 2:

[1115] The server receives the URL or image data of the news article.

[1116] Input: URL of news article or image data sent from the device

[1117] How it works: The server receives the HTTP request and stores the data locally.

[1118] Output: The received data is stored on the server and handed over to the next process.

[1119] Step 3:

[1120] The server preprocesses the data received.

[1121] Input: URL or image data of a saved news article

[1122] How it works: For news articles, it uses BeautifulSoup to remove HTML tags and ads and extract pure text data. For image data, it uses Tesseract OCR to extract text information from images.

[1123] Output: Preprocessed text data or text information in images.

[1124] Step 4:

[1125] The server analyzes the preprocessed data.

[1126] Input: Preprocessed text data or text information in images

[1127] How it works: Uses Hugging Face's Transformer library for natural language processing to understand the context and content of data.

[1128] Output: Parsed data.

[1129] Step 5:

[1130] The server checks the parsed data against a reliable database to assess its veracity.

[1131] Input: Parsed data

[1132] How it works: The parsed data is compared to public databases and established news sources to generate a credibility score.

[1133] Output: A truthfulness score.

[1134] Step 6:

[1135] The server analyzes the user's voice data and facial expression data to recognize their emotional state.

[1136] Input: User's voice data and facial expression data

[1137] How it works: It uses OpenCV and DeepFace libraries to analyze user emotions and identify their emotional state.

[1138] Output: The user's emotional state.

[1139] Step 7:

[1140] The server makes a decision based on the results of the fact check.

[1141] Input: Truthfulness score, user emotional state

[1142] How it works: Based on the evaluation results, it determines whether a news item or image is true or false, generates an appropriate score, and adjusts how the results are displayed based on the user's emotional state.

[1143] Output: Judgment results and how they are displayed.

[1144] Step 8:

[1145] The server generates the judgment result and summary information and transmits them to the terminal.

[1146] Input: Judgment result, display method

[1147] How it works: The server generates news summaries based on reliable information, presents them in a format that is easy for users to understand, and sends the adjusted results to the user's device.

[1148] Output: The judgment result and summary information sent to the user's terminal.

[1149] Step 9:

[1150] The user views the provided results and summary.

[1151] Input: Judgment result and summary information sent from the server

[1152] How it works: Users check the information displayed on their smartphones, tablets, or other devices to verify the authenticity of news articles and images.

[1153] Output: User understanding and judgment.

[1154] 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.

[1155] 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.

[1156] 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.

[1157] [Third embodiment]

[1158] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1159] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1160] 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).

[1161] 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.

[1162] 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.

[1163] 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).

[1164] 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.

[1165] 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.

[1166] 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.

[1167] 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.

[1168] 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.

[1169] 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."

[1170] The present invention relates to a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system performs a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[1171] System Overview

[1172] The system includes the following main components to assess the veracity of news articles and images:

[1173] 1. Data Entry Module

[1174] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[1175] The terminal transmits the input data to the server.

[1176] 2. Data Preprocessing Module

[1177] The server removes HTML tags and unnecessary data from the received news articles and extracts the pure text data.

[1178] The server extracts text information from the image data using OCR technology and analyzes the content of the image using image recognition technology.

[1179] 3. Fact-checking module

[1180] The server analyzes the extracted text data and text information in the image using natural language processing technology.

[1181] The server compares the text data with reliable databases and news sources to assess its veracity.

[1182] 4. Judgment result generation module

[1183] Based on the results of the fact check, the server determines whether the news or image is true or false.

[1184] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[1185] 5. Result Delivery Module

[1186] The server transmits the generated judgment result and summary information to the user's terminal.

[1187] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[1188] Program processing

[1189] The program processing of this system will be explained in natural language below.

[1190] 1. Receiving Data

[1191] A user uses a terminal to input the URL or image data of a news article into the system.

[1192] The terminal sends the input data to the server.

[1193] 2. Data Preprocessing

[1194] The server retrieves the text data of the news article from the received URL.

[1195] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[1196] For image data, the server uses OCR technology to extract text information within the image.

[1197] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[1198] 3. Conduct fact-checks

[1199] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[1200] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[1201] 4. Generating the judgment result

[1202] The server generates a score based on the fact-check results to determine whether a news article or image is true or false.

[1203] The server generates news summaries based on reliable information.

[1204] 5. Providing results

[1205] The server transmits the judgment result and summary information to the user's terminal.

[1206] Users can view the results on their devices and verify the authenticity of the information.

[1207] Specific examples

[1208] 1. News article example

[1209] The user enters the URL of a news article that says, "A politician committed fraud."

[1210] The server retrieves the news article from the URL and preprocesses the text data.

[1211] The server uses natural language processing to analyze the content of news articles and compare them with a reliable database.

[1212] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[1213] Users can check the judgment results and summary on their device.

[1214] 2. Example images

[1215] Users upload "photos of specific incidents."

[1216] The server analyzes the image data using OCR technology and extracts text information.

[1217] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1218] The server determines that the image is misleading and generates a factual explanation.

[1219] Users can check the judgment results and summary explanation on their terminal.

[1220] As described above, the present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and images in real time and quickly provide accurate information to users, allowing users to make decisions based on reliable information without being misled by potentially misleading information.

[1221] The processing flow will be explained below.

[1222] Step 1:

[1223] The user inputs the URL or image data of a news article into the device.

[1224] Step 2:

[1225] The device sends the entered URL and image data to the server.

[1226] Step 3:

[1227] The server scrapes text data from the URL of the news article received.

[1228] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[1229] Step 4:

[1230] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[1231] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[1232] Step 5:

[1233] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[1234] Step 6:

[1235] The server compares the analyzed data with reliable databases and news sources (e.g., public fact-checking sites) to assess the veracity of the information.

[1236] Step 7:

[1237] Based on the results of the fact check, the server scores the news article or image as true or false.

[1238] Step 8:

[1239] The server generates an accurate news summary, including judgments and facts from reliable sources.

[1240] Step 9:

[1241] The server transmits the generated judgment result and summary information to the user's terminal.

[1242] Step 10:

[1243] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[1244] Example 1

[1245] 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."

[1246] Conventional fact-checking systems for news articles and image data often took a long time to evaluate the veracity of articles or based their judgments on unreliable sources. As a result, users were at a higher risk of making decisions based on incorrect information, making it difficult to provide reliable information. Furthermore, inaccurate analysis was not possible due to insufficient preprocessing, such as removing HTML tags from news articles or using optical character recognition technology for image data.

[1247] 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.

[1248] In this invention, the server includes: a means for a user to input a URL or image data of a news article from a terminal and transmit the data from the terminal to the server; a means for acquiring text data from the URL of the news article received by the server or extracting text information from the image data using optical character recognition technology; a means for preprocessing the acquired text data and removing HTML tags and unnecessary data; a means for analyzing the preprocessed text data or image data using natural language processing technology and comparing it with a reliable database or other reliable information source to evaluate its veracity; a means for determining whether the news article or image is true or false based on the comparison result; and a means for generating summary information based on the determination result and facts and transmitting it to the user's terminal. This enables the veracity of news articles and image data to be evaluated with high accuracy in real time and provides accurate information to users quickly.

[1249] A "user" is a person who inputs the URL of a news article or image data into the system via a terminal.

[1250] A "terminal" is a device that receives news article URLs and image data from users and sends them to a server.

[1251] "Server" means a device that processes and analyzes data received from users, evaluates its authenticity, and generates and provides results.

[1252] A "news article" is a web page containing news information.

[1253] "URL" refers to the web address of a news article.

[1254] "Image data" refers to an image file provided by the user.

[1255] "Optical character recognition technology" is a technology that analyzes and extracts text information from images.

[1256] An "HTML tag" is an element of a markup language that defines the structure and content of a web page.

[1257] "Natural language processing technology" is a technology that enables computers to understand, generate, and analyze human language.

[1258] A "database" is a system for systematically storing and managing information.

[1259] "Veracity" refers to whether a news article or image data is based on facts.

[1260] A "judgment score" is a numerical value that is displayed as a result of evaluating the veracity of a news article or image data.

[1261] "Summary information" is information that briefly summarizes the judgment results of news articles or image data.

[1262] "Preprocessing" is the process of preparing data before analysis or evaluation.

[1263] A "reliable database" is a database built on accurate and reliable sources of information.

[1264] "Evaluation" means analyzing data and determining its veracity.

[1265] "Fact checking" is the process of verifying whether the contents of a news article or image data are based on facts.

[1266] This invention is a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system executes a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[1267] The system is roughly divided into a server, a terminal, and a user. First, a user inputs the URL of a news article and image data from their own terminal (PC, smartphone, tablet, etc.). The input data is sent from the terminal to the server.

[1268] The server performs various preprocessing operations on the received data. For news article URLs, web scraping techniques are used to remove HTML tags and obtain pure text data (e.g., using libraries such as BeautifulSoup). For image data, optical character recognition techniques (e.g., Tesseract) are used to extract text information, and image recognition techniques (e.g., OpenCV or TensorFlow) are used to analyze the image content.

[1269] The server then analyzes the preprocessed text data and the text information in the images using natural language processing techniques (e.g., spaCy or NLTK). The analysis results are then cross-referenced with reliable databases and other trusted sources. This cross-reference uses APIs and web crawlers to obtain real-time data (e.g., Wikipedia API or official news API).

[1270] As a result of the matching, the server evaluates the veracity of the news article or image data and generates a judgment score. The server then generates a summary based on reliable information and provides it to the user in an easy-to-understand format. A generative AI model (e.g., GPT-3) is used to generate this summary.

[1271] Finally, the server sends the judgment result and summary information to the user's device. The user can view the judgment result and summary information on their device and confirm the authenticity of the news or image data. This series of processes allows the user to make decisions based on reliable information without being misled by potentially misleading information.

[1272] Specific example explanation

[1273] News article example

[1274] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[1275] 2. The device sends this URL to the server.

[1276] 3. The server retrieves the news article from the URL and preprocesses the text data.

[1277] 4. The server uses natural language processing to analyze the content of the news article and compare it with a reliable database.

[1278] 5. The server determines that the article contains some errors and generates a summary news based on accurate information.

[1279] 6. The user checks the judgment results and summary on the terminal.

[1280] Image example

[1281] 1. The user uploads a photo of a specific incident.

[1282] 2. The terminal sends this image data to the server.

[1283] 3. The server analyzes the image data using OCR technology and extracts text information.

[1284] 4. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1285] 5. The server determines that the image is misleading and generates a factual explanation.

[1286] 6. The user checks the judgment result and summary explanation on the terminal.

[1287] Prompt Sentence Examples

[1288] Example news article prompt:

[1289] Enter the URL of the following news article and our system will analyze its content and rate its veracity:

[1290] Example: https: / / example.com / news / article123

[1291] Image prompt example:

[1292] Please upload the following image and our system will analyze its content and rate its authenticity:

[1293] Example: (Drag and drop image file)

[1294] The above is an embodiment of the present invention. The present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and image data in real time and quickly provide accurate information to users.

[1295] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1296] Step 1: Receiving Data

[1297] The user enters the URL or image data of a news article into an input form on the device, for example, by copying and pasting the URL of a news site or by dragging and dropping an image file.

[1298] The terminal receives the input data and sends it to the server as an HTTP request. Specifically, it converts the input data into JSON format and generates a POST request. This request is added to the processing queue.

[1299] Input: News article URL or image data

[1300] Output: HTTP request to the server

[1301] Step 2: Data preprocessing (analysis of news articles)

[1302] The server retrieves the web page content from the URL of the news article received using a web scraping technique (e.g., BeautifulSoup).

[1303] The server removes unnecessary information such as HTML tags and advertisements from the HTML pages it retrieves, and extracts the text data. Specifically, it uses an HTML parser to filter out tag information and extract only the pure text data.

[1304] Input: News article URL

[1305] Output: Preprocessed text data

[1306] Step 2: Data preprocessing (analysis of image data)

[1307] The server uses optical character recognition technology (e.g., Tesseract) on the received image data to extract text information.

[1308] The server uses image recognition technology (e.g., OpenCV or TensorFlow) to analyze the content of the image and identify relevant information, such as identifying faces and objects contained within the image.

[1309] Input: Image data

[1310] Output: Extracted text information and image content data

[1311] Step 3: Conduct a fact check

[1312] The server analyzes the preprocessed text data or the text information in the images using natural language processing techniques (e.g., spaCy or NLTK), performing syntactic and semantic analysis of the text to extract important keywords and phrases.

[1313] The server cross-checks the parsed data against reliable databases and other trusted sources, using APIs and web crawlers to retrieve real-time data (e.g., Wikipedia API or official news APIs).

[1314] Input: Preprocessed text or image data

[1315] Output: Matching result

[1316] Step 4: Generate a decision result

[1317] The server generates a score based on the fact-check results to evaluate the veracity of the news article or image. Specifically, it calculates the degree of match between the matching results and calculates a score ranging from 0 to 100.

[1318] The server uses a generative AI model (e.g., GPT-3) to generate a summary based on reliable information, including a rating of veracity and relevant factual information.

[1319] Input: Matching result

[1320] Output: Judgment score and summary information

[1321] Step 5: Delivering results

[1322] The server sends the generated judgment result and summary information to the user's device using an HTTP response. Specifically, the generated information is encoded in JSON format and returned as an HTTP response.

[1323] The user checks the judgment results and summary information on the device. For example, the results are displayed in the display area of ​​a web browser. Based on this, the user can determine the authenticity of the news article or image.

[1324] Input: Judgment score and summary information

[1325] Output: Providing results to the user

[1326] (Application example 1)

[1327] 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."

[1328] Conventional systems for assessing the veracity of news articles and images have difficulty assessing the credibility of information entered by users in real time, and providing assessment results quickly via smartphones has been a particular challenge. Furthermore, there is no well-established mechanism for notifying users of misinformation in real time, and immediate comparison with reliable sources is often not performed. This creates a risk that users will make decisions based on incorrect information.

[1329] 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.

[1330] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and comparing it with a reliable database or other reliable information sources to evaluate their veracity, means for determining whether the news or images are true or false based on the evaluation results, means for generating summary information based on the evaluation results and facts and sending it to the terminal, and means for notifying the user of the generated summary information and the evaluation results in real time via an application installed on the smartphone. This enables immediate evaluation of the veracity of news articles and images and the provision of the results in real time.

[1331] A "news article" is a document created to inform the public of various events or information.

[1332] "Image data" is visual information stored in digital format, including photographs and illustrations.

[1333] A "terminal" is a device used by a user to enter information and view results, typically a smartphone or computer.

[1334] "Means for receiving" refers to a mechanism for acquiring news articles and image data sent from a terminal.

[1335] "Preprocessing" refers to the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[1336] "Text data" is digital data that contains text information.

[1337] Optical character recognition (OCR) is a technology for extracting text information from an image as digital character data.

[1338] "Natural language processing technology" is a technology that allows computers to understand and process human language.

[1339] A "trusted database" is a data repository that stores reliable information and primarily contains data from trusted sources.

[1340] A "source" is a data source that provides particular information.

[1341] A "means for assessing veracity" is a mechanism for determining whether the data is true by analyzing preprocessed data and comparing it with reliable sources of information.

[1342] "Evaluation results" refer to the results of a truthfulness analysis, indicating whether the news or image is true or false.

[1343] "Summary information" is information created based on the evaluation results in a concise, easy-to-understand format.

[1344] "Means of real-time notification" refers to a mechanism for instantly conveying the generated summary information and evaluation results to users, and is primarily implemented through smartphone applications.

[1345] An "application" is a software program that runs on a computer to perform a particular function.

[1346] A description will be given of an embodiment of the present invention. A system for evaluating the veracity of news articles and images in real time and providing the results to a user comprises the following elements.

[1347] 1. System Configuration

[1348] This real-time fact-checking system mainly includes the following components:

[1349] Data Entry Module

[1350] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[1351] Data entered from the terminal is transmitted to the server.

[1352] Data Preprocessing Module

[1353] The server retrieves text data from the URL of the received news article and removes unnecessary information such as HTML tags and advertisements.

[1354] For image data, the server uses optical character recognition technology (OCR) to extract text information within the image.

[1355] The text data is automatically cleaned and converted into a format suitable for analysis.

[1356] Fact Checking Module

[1357] The server analyzes the cleaned text data or information in the images using natural language processing techniques.

[1358] The recognized text data is checked against reliable databases and other trusted sources to assess its veracity.

[1359] Judgment result generation module

[1360] The veracity of news and images is determined based on the results of fact-checking.

[1361] Judgments and fact-based summaries are automatically generated.

[1362] Result Delivery Module

[1363] The server transmits the generated judgment result and summary information to the user's terminal.

[1364] An application installed on a smartphone notifies the user of this information in real time.

[1365] The user can view the results and their summary through the terminal.

[1366] 2. Hardware and Software Requirements

[1367] The system uses a variety of hardware and software tools.

[1368] Hardware

[1369] Devices: smartphones, computers

[1370] Server: A server with high-performance processing capabilities

[1371] software

[1372] OCR technology: software for extracting text information from images (e.g., Tesseract OCR)

[1373] Natural language processing libraries: Tools for analyzing news articles and text data (e.g., SpaCy and NLTK)

[1374] Database connection software: Libraries for communicating with reliable databases

[1375] Smartphone app: An application that provides users with real-time information

[1376] Generative AI models: AI models used to generate summaries and analysis results for text data.

[1377] 3. Specific Examples

[1378] Specific usage scenarios include:

[1379] News article example

[1380] The user enters the URL of a news article that purports to say that a certain politician committed fraud.

[1381] The server retrieves the news article from the URL and preprocesses the text data.

[1382] The content of news articles is analyzed using natural language processing and compared with a reliable database.

[1383] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[1384] Users can check the assessment results and summary in real time using a smartphone application.

[1385] Image example

[1386] Users upload "photos of specific incidents."

[1387] The server analyzes the image data using OCR technology and extracts text information.

[1388] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1389] The server determines that the image is misleading and generates a factual explanation.

[1390] Users can check the assessment results and summary explanations in real time using a smartphone application.

[1391] 4. Examples of prompts

[1392] An example of a prompt that a user might enter into the system is:

[1393] "https: / / example.com / news-article Check whether this news article is true or not."

[1394] "Please verify this image is authentic."

[1395] This allows users to identify misinformation in real time and make quick decisions based on trusted sources of information.

[1396] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1397] Step 1:

[1398] User enters the URL or image data of a news article

[1399] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[1400] Input: News article URL or image data

[1401] Output: Input data is sent from the device to the server

[1402] Step 2:

[1403] The server preprocesses the received data

[1404] The server retrieves text data from the URLs of received news articles, removing unnecessary information such as HTML tags and advertisements, and extracting pure text data.

[1405] For image data, OCR technology is used to extract text information within the image.

[1406] Input: User-entered news article URL or image data

[1407] Output: Preprocessed text data or text information in images

[1408] Step 3:

[1409] The server analyzes the data using natural language processing technology

[1410] The server then analyzes the preprocessed text data and text information in the images using natural language processing techniques, which includes extracting words, understanding context, and summarizing information.

[1411] Input: Preprocessed text data or text information in images

[1412] Output: Parsed text data

[1413] Step 4:

[1414] The server checks the authenticity against a reliable database

[1415] The server then compares the parsed text data with reliable databases and other trusted sources, assessing the authenticity of the news article or image.

[1416] Input: Parsed text data

[1417] Output: Veracity evaluation result

[1418] Step 5:

[1419] The server generates the judgment result and summary information

[1420] The server generates a score based on the fact-check evaluation results to determine whether a news article or image is true or false, and creates a fact-based summary based on the evaluation results.

[1421] Input: Veracity assessment result

[1422] Output: Judgment results and summary information

[1423] Step 6:

[1424] The server sends the results to the device

[1425] The server then sends the generated judgment results and summary information to the user's device, and an application installed on the smartphone notifies the user in real time.

[1426] Input: Judgment results and summary information

[1427] Output: Notification and display on the user's terminal

[1428] Step 7:

[1429] User checks the results

[1430] Users can check the results and summary information sent from the server through a smartphone application, allowing them to determine the authenticity of news articles and images.

[1431] Input: Notification from the server and displayed results

[1432] Output: User's understanding and judgment

[1433] The above steps realize a system that evaluates the veracity of news articles and images in real time and provides appropriate information to users.

[1434] 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.

[1435] The present invention relates to a system that uses natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provides results based on the user's emotions. This system performs a series of processes, including preprocessing, analysis, fact-checking, emotion recognition, and generating and providing judgment results for news articles or image data.

[1436] System Overview

[1437] The system includes the following main components:

[1438] 1. Data Entry Module

[1439] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[1440] The terminal transmits the input data to the server.

[1441] 2. Data Preprocessing Module

[1442] The server removes unnecessary data such as HTML tags and advertisements from the received news articles and extracts pure text data.

[1443] The server uses optical character recognition (OCR) technology on the image data to extract text information within the image.

[1444] The server analyzes the content of the image using image recognition technology.

[1445] 3. Fact-checking module

[1446] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[1447] The server checks the text data against reliable databases and other trusted sources to assess its veracity.

[1448] 4. Emotion Engine Module

[1449] The server analyzes the user's voice data and facial expression data acquired through the user's terminal and recognizes emotions.

[1450] The server adjusts the method of providing the judgment results and summary information based on the user's emotions recognized by the emotion engine.

[1451] 5. Judgment Result Generation Module

[1452] The server determines whether the news or image is true or false based on the results of the fact check.

[1453] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[1454] 6. Result Delivery Module

[1455] The server transmits the generated judgment result and summary information to the user's terminal.

[1456] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[1457] Program processing

[1458] The program processing of this system will be explained in natural language below.

[1459] 1. Receiving Data

[1460] A user uses a terminal to input the URL or image data of a news article into the system.

[1461] The terminal sends the input data to the server.

[1462] 2. Data Preprocessing

[1463] The server obtains the text data of the news article from the received URL.

[1464] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[1465] The server analyzes the image data using OCR technology and extracts text information.

[1466] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[1467] 3. Conduct fact-checks

[1468] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[1469] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[1470] 4. Emotion Recognition by Emotion Engine

[1471] The server uses an emotion engine to analyze the user's voice data and facial expression data acquired through the user's device, and recognizes the user's current emotions.

[1472] For example, the emotional state of the user, such as whether they are surprised, angry, or relaxed, is grasped.

[1473] 5. Generating the judgment result

[1474] Based on the results of the fact check, the server generates a score that determines whether the news article or image is true or false.

[1475] The server generates a news summary based on reliable information, including judgments and facts from reliable sources.

[1476] 6. Tailored delivery of results

[1477] The server adjusts the display method of the judgment result and summary information based on the user's emotional state recognized by the emotion engine.

[1478] For example, if the user is surprised, provide more detailed information to make it easier for the user to understand.

[1479] 7. Providing Results

[1480] The server transmits the generated judgment result and summary information to the user's terminal.

[1481] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[1482] Specific examples

[1483] 1. News article example

[1484] A user enters the URL of a news article that says, "A politician committed fraud."

[1485] The server retrieves the news article from the URL and preprocesses the text data.

[1486] The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[1487] The server determines that this article contains some errors and generates a summary news based on accurate information.

[1488] The server recognizes the user's current emotion and, if they are surprised for example, provides a result with detailed supporting information.

[1489] Users can check the judgment results and summary on their device.

[1490] 2. Example images

[1491] A user uploads a "photo of a specific incident."

[1492] The server analyzes the image data using OCR technology and extracts text information.

[1493] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1494] The server determines that the image is misleading and generates a factual explanation.

[1495] The server recognizes the user's current emotions and, for example, if they are anxious, provides results that are thoughtfully reassuring.

[1496] Users can check the judgment results and summary explanation on their terminal.

[1497] As described above, the present invention is a system that utilizes natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provide information that is individually tailored to the user's emotions. As a result, users can make decisions based on reliable information without being misled by potentially misleading information. Furthermore, by taking the user's emotions into consideration, a more user-friendly and easy-to-use system can be realized.

[1498] The processing flow will be explained below.

[1499] Step 1:

[1500] The user inputs the URL or image data of a news article into the device.

[1501] Step 2:

[1502] The device sends the entered URL and image data to the server.

[1503] Step 3:

[1504] The server scrapes text data from the URL of the news article received.

[1505] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[1506] Step 4:

[1507] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[1508] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[1509] Step 5:

[1510] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[1511] Step 6:

[1512] The server cross-checks the parsed data against reliable databases and other trusted sources (e.g., public fact-checking sites) to assess the veracity of the information.

[1513] Step 7:

[1514] Based on the results of the fact check, the server scores the news article or image as true or false.

[1515] Step 8:

[1516] The server uses an emotion engine to analyze the voice and facial expression data it receives from the user's device and recognizes the user's current emotions.

[1517] Step 9:

[1518] The server adjusts the method of providing the judgment results and summary information based on the user's emotional state.

[1519] For example, if the user is surprised, add more information in an easy-to-understand format.

[1520] Step 10:

[1521] The server generates an accurate news summary, including judgments and facts from reliable sources.

[1522] Step 11:

[1523] The server transmits the generated judgment result and summary information to the user's terminal.

[1524] Step 12:

[1525] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[1526] Example 2

[1527] 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."

[1528] In modern society, quickly and accurately assessing the veracity of news articles and images has become an important issue. The inability to instantly determine the reliability of such information can lead to the spread of false information and misleading images, which can have a significant impact on society. Furthermore, as the amount of information increases, it is becoming increasingly difficult for users to quickly verify the authenticity of information. Furthermore, information provided does not take into account the user's feelings, which can lead to misunderstandings and anxiety.

[1529] 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.

[1530] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their authenticity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data to recognize emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts and transmitting it to the terminal in a form that corresponds to the user's emotional state. This makes it possible to quickly and accurately evaluate the authenticity of news articles and images and provide information that corresponds to the user's emotions.

[1531] A "news article" is text data containing current information and news reports that are primarily published on websites and digital media.

[1532] "Image data" refers to file formats (e.g., JPEG, PNG) that contain visual information stored in a digital format.

[1533] A "terminal" is an electronic device such as a computer or smartphone that is used to input news articles and image data and view the results.

[1534] "Server" is a central control device that provides resources for data processing and analysis and manages the operation of the entire system.

[1535] "Preprocessing" refers to the process of removing unnecessary information from news articles or image data and converting them into a format suitable for analysis.

[1536] "Text data" refers to purely textual information extracted from news articles or image data.

[1537] "Optical character recognition technology" refers to technology (e.g., OCR) used to convert text in an image into digital characters.

[1538] "Natural language processing technology" refers to computer science technology for analyzing the meaning of text data and understanding its context and content.

[1539] A "trusted database" is a data repository containing verified information sources that can be used to assess veracity.

[1540] "Emotion recognition technology" is a technology that analyzes a user's voice data and facial expression data to understand their emotional state.

[1541] "Decision result" refers to the conclusion reached after evaluating the veracity of news articles or image data.

[1542] "Summary information" refers to information that succinctly summarizes the important points of a news article or image data.

[1543] A "reliable source" is a source or database that is widely recognized as providing truthful information.

[1544] This invention relates to a real-time fact-checking system that utilizes natural language processing and emotion recognition technologies. This system evaluates the veracity of news articles and image data and provides information according to the user's emotions, thereby realizing the provision of highly reliable information.

[1545] System configuration and operation

[1546] The system includes the following major components:

[1547] server

[1548] Device (PC or smartphone)

[1549] Receiving data

[1550] The user inputs the URL and image data of the news article into the terminal, which then sends it to the server.

[1551] Data Preprocessing

[1552] The server downloads the HTML data from the URL of the received news article and extracts the text data using an HTML parser (e.g., BeautifulSoup). It then removes unnecessary information, such as HTML tags and advertisements, from the text data. For image data, it extracts the text information from the image using an OCR engine (e.g., Tesseract OCR) and analyzes the image content using an image recognition model (e.g., ResNet, YOLO).

[1553] Data analysis and evaluation

[1554] The server inputs the preprocessed text data and information in the images into natural language processing models (e.g., BERT, GPT-3) to analyze the context and content, and then compares the information with reliable databases (e.g., Wikidata, FactCheck.org) and news sources via API to assess the veracity of the information.

[1555] Emotion recognition

[1556] The server inputs the voice data and facial expression images sent from the device into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to understand the user's emotional state. For example, it recognizes emotions such as surprise, anger, or relaxation.

[1557] Generation of judgment results and summary information

[1558] The server determines whether a news article or image is true or false based on the fact-check results and generates a truthfulness score. The server then uses natural language generation technology (e.g., GPT-3) to create a news summary, which includes the results of the judgment and facts based on reliable sources.

[1559] Providing information

[1560] The server adjusts the display method of the judgment results and summary information according to the user's emotional state. For example, if the user is surprised, it provides more detailed information, and if the user is relaxed, it provides simple information. The generated information is sent to the terminal, where the user can view it.

[1561] Specific examples

[1562] News article example

[1563] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[1564] 2. The server retrieves the news article from the URL and preprocesses the text data.

[1565] 3. The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[1566] 4. The server determines that this article contains some errors and generates a summary news article based on accurate information.

[1567] 5. The server recognizes the user's emotion and provides a result with detailed supporting information if, for example, they are surprised.

[1568] 6. The user checks the judgment results and summary on the terminal.

[1569] Image example

[1570] 1. A user uploads a photo of a specific incident.

[1571] 2. The server analyzes the image data using OCR technology and extracts text information.

[1572] 3. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1573] 4. The server determines that the image is potentially misleading and generates a factual explanation.

[1574] 5. The server recognizes the user's emotions and provides reassuring information if, for example, the user is worried.

[1575] 6. The user checks the judgment result and summary explanation on the terminal.

[1576] Example prompt sentence:

[1577] The user is prompted to enter information in a format such as "Please enter the URL of a news article" or "Please upload image data."

[1578] As described above, the present invention makes it possible to evaluate the veracity of news articles and images in real time and provide information adjusted according to the user's emotions.

[1579] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1580] Program processing steps

[1581] Step 1: Receiving Data

[1582] A user uses a device to input a URL or image data of a news article.

[1583] Specific actions: Paste or upload the URL of a news article or image file into the input form displayed on the device.

[1584] Input: News article URL or image data.

[1585] Output: The received URL or image data.

[1586] Step 2: Sending data

[1587] The terminal transmits the input data to the server.

[1588] Specific operation: The device sends an HTTP POST request to the server, including the received URL or image data.

[1589] Input: Received URL or image data.

[1590] Output: URL or image data received by the server.

[1591] Step 3: Acquiring and preprocessing news articles

[1592] The server retrieves the text data of the news article from the received URL and removes unnecessary information.

[1593] What happens: The server uses an HTML parser (e.g. BeautifulSoup) to extract the text data and remove HTML tags and advertisements.

[1594] Input: The received URL.

[1595] Output: Pure text data.

[1596] Step 4: OCR analysis of image data

[1597] The image data received by the server is analyzed using an OCR engine to extract text information.

[1598] What happens: The server uses an OCR engine (e.g., Tesseract OCR) to extract text information from the image.

[1599] Input: Received image data.

[1600] Output: The extracted text information.

[1601] Step 5: Image content analysis

[1602] The server analyzes the content of the image using an image recognition model.

[1603] Specific operation: The server inputs image data into an image recognition model (e.g., ResNet, YOLO) and identifies the content of the image (people, objects, scenes, etc.).

[1604] Input: Received image data.

[1605] Output: Analyzed image content.

[1606] Step 6: Analyzing text and image data

[1607] The server analyzes the preprocessed text data and the extracted text in the image using natural language processing technology.

[1608] Specific operation: The server inputs text data into a natural language processing model (e.g., BERT, GPT-3) and analyzes the context and content.

[1609] Input: Pure text data and extracted text information.

[1610] Output: Semantic information of the parsed text and image data.

[1611] Step 7: Veracity Assessment

[1612] The server checks the parsed data against a reliable database to assess its veracity.

[1613] Specific operation: The server collates the parsed data via the database API and evaluates its authenticity.

[1614] Input: Parsed text and image data.

[1615] Output: A veracity rating score.

[1616] Step 8: Recognize emotions

[1617] The server analyzes the user's voice data and facial expression data to recognize the user's emotional state.

[1618] Specific operation: The server inputs voice data and facial expression data into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to recognize emotions.

[1619] Input: User's voice and facial expression data.

[1620] Output: The perceived emotional state of the user.

[1621] Step 9: Generate decision results and summary information

[1622] The server generates a judgment result based on the truthfulness evaluation and creates a summarized news.

[1623] Specific operation: The server generates a news summary using natural language generation technology (e.g., GPT-3). The server creates a news summary based on the judgment results.

[1624] Input: Truthfulness rating score.

[1625] Output: Judgment results and summary news.

[1626] Step 10: Coordinating input

[1627] The server provides the results in a coordinated manner.

[1628] Specific behavior: The server adjusts the way information is displayed depending on the user's emotional state, choosing whether to provide detailed or concise information.

[1629] Input: The perceived emotional state of the user and the judgment result.

[1630] Output: Adjusted judgement results and summary news.

[1631] Step 11: Sending the results

[1632] The server transmits the generated judgment result and summary information to the user's terminal.

[1633] Specific behavior: The server sends the result using push notification or HTTP response.

[1634] Input: Adjusted judgement results and summary news.

[1635] Output: The results sent to the terminal.

[1636] Step 12: View the results

[1637] The user checks the judgment results and summary information on the terminal.

[1638] Specific operations: The user uses the device to view the received judgment results and summary information.

[1639] Input: The result sent to the terminal.

[1640] Output: User understanding and judgment.

[1641] This concludes the program processing flow for this system, including the specific operations performed at each step. This enables the system to quickly and accurately evaluate the veracity of news articles and images, and provide information tailored to the user's emotions.

[1642] (Application example 2)

[1643] 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."

[1644] In modern society, the proliferation of fake news and fabricated images has led to the problem of many people making decisions based on incorrect information. This problem is a factor that causes social confusion and misunderstanding. Furthermore, while user emotions have a significant impact on determining the authenticity of information, current systems provide uniform information without taking user emotions into consideration, which reduces user understanding and satisfaction.

[1645] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1646] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their veracity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data and recognizing emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts, adjusting the display method of the result based on the user's emotional state, and transmitting the summary information to the terminal. This makes it easier for users to obtain individually tailored, reliable information without being misled by erroneous information.

[1647] A "news article" is text data and image data containing news information distributed online or offline.

[1648] "Image data" means files containing visual information stored and transmitted in digital format.

[1649] "Natural language processing technology" is a technology that uses computers to understand, interpret, and generate human language.

[1650] A "real-time fact-checking system" is a system that instantly evaluates the veracity of information and provides the results immediately.

[1651] "Data preprocessing" is the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[1652] "Text data" is a data format that includes information about sentences and characters.

[1653] "Extract" refers to the operation of extracting necessary parts from data.

[1654] A "database" is a collection of information that systematically stores large amounts of information and makes it searchable and usable.

[1655] A "reliable source" is an entity or medium that provides information that is highly accurate, reliable, and has been assessed.

[1656] "Emotion recognition" is a technology that analyzes voice data and facial expression data to determine a user's emotional state.

[1657] "Evaluation results" are the conclusions or grades reached after determining the truth or falsity of the information.

[1658] "Summary information" is the result of summarizing a large amount of information in a concise manner.

[1659] A "terminal" is a device such as a computer, smartphone, or tablet that is connected to the Internet or a network.

[1660] "Result presentation method" refers to the means or format in which information is presented to the user.

[1661] The system that realizes this application example evaluates the veracity of news articles and images in real time and provides results according to the user's emotions. Below is a natural language explanation of the program processing of this system.

[1662] The server first receives the URL or image data of the news article from the user's device. The device then sends the input data to the server. This process requires an internet connection, and the user can use a device such as a smartphone or tablet.

[1663] The server retrieves the text data of the news article from the received URL, and then uses a data preprocessing module to remove unnecessary information such as HTML tags and advertisements, converting it into pure text data. This process uses BeautifulSoup as a scraping library and Tesseract OCR as optical character recognition (OCR) technology.

[1664] The server then analyzes the extracted text data and textual information within the images and uses natural language processing techniques (using Hugging Face's Transformer library for its NLP model) to assess their veracity by matching them with trusted databases and other reliable sources, such as public databases and established news sources.

[1665] The server also includes an emotion recognition module to analyze the user's voice and facial expression data to recognize the user's emotional state. For this purpose, it uses OpenCV and the DeepFace library, which allows it to understand the user's emotions, such as surprised, angry, relaxed, etc.

[1666] The server determines whether a news item or image is true or false based on the fact-check results and generates a score. It then generates a news summary based on reliable information, including the results of the fact-check and facts from reliable sources. The server then adjusts the display of the results and summary information based on the user's emotional state, as recognized by the emotion engine. For example, if the user is surprised, the server provides more detailed information to make it easier for the user to understand.

[1667] The results are then sent from the server to the user's device, where the user can view the results and summary information to verify the authenticity of the news article or image.

[1668] Specific examples

[1669] As a concrete example, the following scenario can be considered.

[1670] A user enters the URL of a news article about a politician committing fraud. The server retrieves the news article from the URL and preprocesses the text data. Next, the server analyzes the text using natural language processing technology and compares it with a reliable database. The server determines that the article contains some errors and generates a summary news based on accurate information. Furthermore, the server recognizes the user's current emotions and provides results with detailed supplementary information if the user is surprised. The user can check the results and summary on their device.

[1671] Prompt Sentence Examples

[1672] The user inputs the URL or image data of a news article. The system receives the URL or image data and performs preprocessing. It then analyzes the data using natural language processing techniques and compares it with a reliable database. It recognizes the user's emotion (e.g., joy) and generates and displays results according to that emotion.

[1673] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1674] Step 1:

[1675] The user inputs the URL and image data of the news article via the terminal.

[1676] Input: News article URL or image data

[1677] How it works: A user provides data through an input form on a device such as a smartphone or tablet.

[1678] Output: The entered data is sent from the terminal to the server.

[1679] Step 2:

[1680] The server receives the URL or image data of the news article.

[1681] Input: URL of news article or image data sent from the device

[1682] How it works: The server receives the HTTP request and stores the data locally.

[1683] Output: The received data is stored on the server and handed over to the next process.

[1684] Step 3:

[1685] The server preprocesses the data received.

[1686] Input: URL or image data of a saved news article

[1687] How it works: For news articles, it uses BeautifulSoup to remove HTML tags and ads and extract pure text data. For image data, it uses Tesseract OCR to extract text information from images.

[1688] Output: Preprocessed text data or text information in images.

[1689] Step 4:

[1690] The server analyzes the preprocessed data.

[1691] Input: Preprocessed text data or text information in images

[1692] How it works: Uses Hugging Face's Transformer library for natural language processing to understand the context and content of data.

[1693] Output: Parsed data.

[1694] Step 5:

[1695] The server checks the parsed data against a reliable database to assess its veracity.

[1696] Input: Parsed data

[1697] How it works: The parsed data is compared to public databases and established news sources to generate a credibility score.

[1698] Output: A truthfulness score.

[1699] Step 6:

[1700] The server analyzes the user's voice data and facial expression data to recognize their emotional state.

[1701] Input: User's voice data and facial expression data

[1702] How it works: It uses OpenCV and DeepFace libraries to analyze user emotions and identify their emotional state.

[1703] Output: The user's emotional state.

[1704] Step 7:

[1705] The server makes a decision based on the results of the fact check.

[1706] Input: Truthfulness score, user emotional state

[1707] How it works: Based on the evaluation results, it determines whether a news item or image is true or false, generates an appropriate score, and adjusts how the results are displayed based on the user's emotional state.

[1708] Output: Judgment results and how they are displayed.

[1709] Step 8:

[1710] The server generates the judgment result and summary information and transmits them to the terminal.

[1711] Input: Judgment result, display method

[1712] How it works: The server generates news summaries based on reliable information, presents them in a format that is easy for users to understand, and sends the adjusted results to the user's device.

[1713] Output: The judgment result and summary information sent to the user's terminal.

[1714] Step 9:

[1715] The user views the provided results and summary.

[1716] Input: Judgment result and summary information sent from the server

[1717] How it works: Users check the information displayed on their smartphones, tablets, or other devices to verify the authenticity of news articles and images.

[1718] Output: User understanding and judgment.

[1719] 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.

[1720] 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.

[1721] 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.

[1722] [Fourth embodiment]

[1723] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1724] 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.

[1725] 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).

[1726] 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.

[1727] 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.

[1728] 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).

[1729] 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.

[1730] 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.

[1731] 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.

[1732] 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.

[1733] 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.

[1734] 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.

[1735] 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."

[1736] The present invention relates to a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system performs a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[1737] System Overview

[1738] The system includes the following main components to assess the veracity of news articles and images:

[1739] 1. Data Entry Module

[1740] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[1741] The terminal transmits the input data to the server.

[1742] 2. Data Preprocessing Module

[1743] The server removes HTML tags and unnecessary data from the received news articles and extracts the pure text data.

[1744] The server extracts text information from the image data using OCR technology and analyzes the content of the image using image recognition technology.

[1745] 3. Fact-checking module

[1746] The server analyzes the extracted text data and text information in the image using natural language processing technology.

[1747] The server compares the text data with reliable databases and news sources to assess its veracity.

[1748] 4. Judgment result generation module

[1749] Based on the results of the fact check, the server determines whether the news or image is true or false.

[1750] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[1751] 5. Result Delivery Module

[1752] The server transmits the generated judgment result and summary information to the user's terminal.

[1753] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[1754] Program processing

[1755] The program processing of this system will be explained in natural language below.

[1756] 1. Receiving Data

[1757] A user uses a terminal to input the URL or image data of a news article into the system.

[1758] The terminal sends the input data to the server.

[1759] 2. Data Preprocessing

[1760] The server retrieves the text data of the news article from the received URL.

[1761] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[1762] For image data, the server uses OCR technology to extract text information within the image.

[1763] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[1764] 3. Conduct fact-checks

[1765] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[1766] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[1767] 4. Generating the judgment result

[1768] The server generates a score based on the fact-check results to determine whether a news article or image is true or false.

[1769] The server generates news summaries based on reliable information.

[1770] 5. Providing results

[1771] The server transmits the judgment result and summary information to the user's terminal.

[1772] Users can view the results on their devices and verify the authenticity of the information.

[1773] Specific examples

[1774] 1. News article example

[1775] The user enters the URL of a news article that says, "A politician committed fraud."

[1776] The server retrieves the news article from the URL and preprocesses the text data.

[1777] The server uses natural language processing to analyze the content of news articles and compare them with a reliable database.

[1778] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[1779] Users can check the judgment results and summary on their device.

[1780] 2. Example images

[1781] Users upload "photos of specific incidents."

[1782] The server analyzes the image data using OCR technology and extracts text information.

[1783] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1784] The server determines that the image is misleading and generates a factual explanation.

[1785] Users can check the judgment results and summary explanation on their terminal.

[1786] As described above, the present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and images in real time and quickly provide accurate information to users, allowing users to make decisions based on reliable information without being misled by potentially misleading information.

[1787] The processing flow will be explained below.

[1788] Step 1:

[1789] The user inputs the URL or image data of a news article into the device.

[1790] Step 2:

[1791] The device sends the entered URL and image data to the server.

[1792] Step 3:

[1793] The server scrapes text data from the URL of the news article received.

[1794] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[1795] Step 4:

[1796] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[1797] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[1798] Step 5:

[1799] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[1800] Step 6:

[1801] The server compares the analyzed data with reliable databases and news sources (e.g., public fact-checking sites) to assess the veracity of the information.

[1802] Step 7:

[1803] Based on the results of the fact check, the server scores the news article or image as true or false.

[1804] Step 8:

[1805] The server generates an accurate news summary, including judgments and facts from reliable sources.

[1806] Step 9:

[1807] The server transmits the generated judgment result and summary information to the user's terminal.

[1808] Step 10:

[1809] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[1810] Example 1

[1811] 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."

[1812] Conventional fact-checking systems for news articles and image data often took a long time to evaluate the veracity of articles or based their judgments on unreliable sources. As a result, users were at a higher risk of making decisions based on incorrect information, making it difficult to provide reliable information. Furthermore, inaccurate analysis was not possible due to insufficient preprocessing, such as removing HTML tags from news articles or using optical character recognition technology for image data.

[1813] 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.

[1814] In this invention, the server includes: a means for a user to input a URL or image data of a news article from a terminal and transmit the data from the terminal to the server; a means for acquiring text data from the URL of the news article received by the server or extracting text information from the image data using optical character recognition technology; a means for preprocessing the acquired text data and removing HTML tags and unnecessary data; a means for analyzing the preprocessed text data or image data using natural language processing technology and comparing it with a reliable database or other reliable information source to evaluate its veracity; a means for determining whether the news article or image is true or false based on the comparison result; and a means for generating summary information based on the determination result and facts and transmitting it to the user's terminal. This enables the veracity of news articles and image data to be evaluated with high accuracy in real time and provides accurate information to users quickly.

[1815] A "user" is a person who inputs the URL of a news article or image data into the system via a terminal.

[1816] A "terminal" is a device that receives news article URLs and image data from users and sends them to a server.

[1817] "Server" means a device that processes and analyzes data received from users, evaluates its authenticity, and generates and provides results.

[1818] A "news article" is a web page containing news information.

[1819] "URL" refers to the web address of a news article.

[1820] "Image data" refers to an image file provided by the user.

[1821] "Optical character recognition technology" is a technology that analyzes and extracts text information from images.

[1822] An "HTML tag" is an element of a markup language that defines the structure and content of a web page.

[1823] "Natural language processing technology" is a technology that enables computers to understand, generate, and analyze human language.

[1824] A "database" is a system for systematically storing and managing information.

[1825] "Veracity" refers to whether a news article or image data is based on facts.

[1826] A "judgment score" is a numerical value that is displayed as a result of evaluating the veracity of a news article or image data.

[1827] "Summary information" is information that briefly summarizes the judgment results of news articles or image data.

[1828] "Preprocessing" is the process of preparing data before analysis or evaluation.

[1829] A "reliable database" is a database built on accurate and reliable sources of information.

[1830] "Evaluation" means analyzing data and determining its veracity.

[1831] "Fact checking" is the process of verifying whether the contents of a news article or image data are based on facts.

[1832] This invention is a system that uses natural language processing technology to evaluate the veracity of news articles and images in real time and provides the results to users. This system executes a series of processes, including preprocessing, analysis, fact-checking, and generating and providing judgment results for news articles or image data.

[1833] The system is roughly divided into a server, a terminal, and a user. First, a user inputs the URL of a news article and image data from their own terminal (PC, smartphone, tablet, etc.). The input data is sent from the terminal to the server.

[1834] The server performs various preprocessing operations on the received data. For news article URLs, web scraping techniques are used to remove HTML tags and obtain pure text data (e.g., using libraries such as BeautifulSoup). For image data, optical character recognition techniques (e.g., Tesseract) are used to extract text information, and image recognition techniques (e.g., OpenCV or TensorFlow) are used to analyze the image content.

[1835] The server then analyzes the preprocessed text data and the text information in the images using natural language processing techniques (e.g., spaCy or NLTK). The analysis results are then cross-referenced with reliable databases and other trusted sources. This cross-reference uses APIs and web crawlers to obtain real-time data (e.g., Wikipedia API or official news API).

[1836] As a result of the matching, the server evaluates the veracity of the news article or image data and generates a judgment score. The server then generates a summary based on reliable information and provides it to the user in an easy-to-understand format. A generative AI model (e.g., GPT-3) is used to generate this summary.

[1837] Finally, the server sends the judgment result and summary information to the user's device. The user can view the judgment result and summary information on their device and confirm the authenticity of the news or image data. This series of processes allows the user to make decisions based on reliable information without being misled by potentially misleading information.

[1838] Specific example explanation

[1839] News article example

[1840] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[1841] 2. The device sends this URL to the server.

[1842] 3. The server retrieves the news article from the URL and preprocesses the text data.

[1843] 4. The server uses natural language processing to analyze the content of the news article and compare it with a reliable database.

[1844] 5. The server determines that the article contains some errors and generates a summary news based on accurate information.

[1845] 6. The user checks the judgment results and summary on the terminal.

[1846] Image example

[1847] 1. The user uploads a photo of a specific incident.

[1848] 2. The terminal sends this image data to the server.

[1849] 3. The server analyzes the image data using OCR technology and extracts text information.

[1850] 4. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1851] 5. The server determines that the image is misleading and generates a factual explanation.

[1852] 6. The user checks the judgment result and summary explanation on the terminal.

[1853] Prompt Sentence Examples

[1854] Example news article prompt:

[1855] Enter the URL of the following news article and our system will analyze its content and rate its veracity:

[1856] Example: https: / / example.com / news / article123

[1857] Image prompt example:

[1858] Please upload the following image and our system will analyze its content and rate its authenticity:

[1859] Example: (Drag and drop image file)

[1860] The above is an embodiment of the present invention. The present invention is a system that utilizes natural language processing technology to evaluate the veracity of news articles and image data in real time and quickly provide accurate information to users.

[1861] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1862] Step 1: Receiving Data

[1863] The user enters the URL or image data of a news article into an input form on the device, for example, by copying and pasting the URL of a news site or by dragging and dropping an image file.

[1864] The terminal receives the input data and sends it to the server as an HTTP request. Specifically, it converts the input data into JSON format and generates a POST request. This request is added to the processing queue.

[1865] Input: News article URL or image data

[1866] Output: HTTP request to the server

[1867] Step 2: Data preprocessing (analysis of news articles)

[1868] The server retrieves the web page content from the URL of the news article received using a web scraping technique (e.g., BeautifulSoup).

[1869] The server removes unnecessary information such as HTML tags and advertisements from the HTML pages it retrieves, and extracts the text data. Specifically, it uses an HTML parser to filter out tag information and extract only the pure text data.

[1870] Input: News article URL

[1871] Output: Preprocessed text data

[1872] Step 2: Data preprocessing (analysis of image data)

[1873] The server uses optical character recognition technology (e.g., Tesseract) on the received image data to extract text information.

[1874] The server uses image recognition technology (e.g., OpenCV or TensorFlow) to analyze the content of the image and identify relevant information, such as identifying faces and objects contained within the image.

[1875] Input: Image data

[1876] Output: Extracted text information and image content data

[1877] Step 3: Conduct a fact check

[1878] The server analyzes the preprocessed text data or the text information in the images using natural language processing techniques (e.g., spaCy or NLTK), performing syntactic and semantic analysis of the text to extract important keywords and phrases.

[1879] The server cross-checks the parsed data against reliable databases and other trusted sources, using APIs and web crawlers to retrieve real-time data (e.g., Wikipedia API or official news APIs).

[1880] Input: Preprocessed text or image data

[1881] Output: Matching result

[1882] Step 4: Generate a decision result

[1883] The server generates a score based on the fact-check results to evaluate the veracity of the news article or image. Specifically, it calculates the degree of match between the matching results and calculates a score ranging from 0 to 100.

[1884] The server uses a generative AI model (e.g., GPT-3) to generate a summary based on reliable information, including a rating of veracity and relevant factual information.

[1885] Input: Matching result

[1886] Output: Judgment score and summary information

[1887] Step 5: Delivering results

[1888] The server sends the generated judgment result and summary information to the user's device using an HTTP response. Specifically, the generated information is encoded in JSON format and returned as an HTTP response.

[1889] The user checks the judgment results and summary information on the device. For example, the results are displayed in the display area of ​​a web browser. Based on this, the user can determine the authenticity of the news article or image.

[1890] Input: Judgment score and summary information

[1891] Output: Providing results to the user

[1892] (Application example 1)

[1893] 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."

[1894] Conventional systems for assessing the veracity of news articles and images have difficulty assessing the credibility of information entered by users in real time, and providing assessment results quickly via smartphones has been a particular challenge. Furthermore, there is no well-established mechanism for notifying users of misinformation in real time, and immediate comparison with reliable sources is often not performed. This creates a risk that users will make decisions based on incorrect information.

[1895] 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.

[1896] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and comparing it with a reliable database or other reliable information sources to evaluate their veracity, means for determining whether the news or images are true or false based on the evaluation results, means for generating summary information based on the evaluation results and facts and sending it to the terminal, and means for notifying the user of the generated summary information and the evaluation results in real time via an application installed on the smartphone. This enables immediate evaluation of the veracity of news articles and images and the provision of the results in real time.

[1897] A "news article" is a document created to inform the public of various events or information.

[1898] "Image data" is visual information stored in digital format, including photographs and illustrations.

[1899] A "terminal" is a device used by a user to enter information and view results, typically a smartphone or computer.

[1900] "Means for receiving" refers to a mechanism for acquiring news articles and image data sent from a terminal.

[1901] "Preprocessing" refers to the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[1902] "Text data" is digital data that contains text information.

[1903] Optical character recognition (OCR) is a technology for extracting text information from an image as digital character data.

[1904] "Natural language processing technology" is a technology that allows computers to understand and process human language.

[1905] A "trusted database" is a data repository that stores reliable information and primarily contains data from trusted sources.

[1906] A "source" is a data source that provides particular information.

[1907] A "means for assessing veracity" is a mechanism for determining whether the data is true by analyzing preprocessed data and comparing it with reliable sources of information.

[1908] "Evaluation results" refer to the results of a truthfulness analysis, indicating whether the news or image is true or false.

[1909] "Summary information" is information created based on the evaluation results in a concise, easy-to-understand format.

[1910] "Means of real-time notification" refers to a mechanism for instantly conveying the generated summary information and evaluation results to users, and is primarily implemented through smartphone applications.

[1911] An "application" is a software program that runs on a computer to perform a particular function.

[1912] A description will be given of an embodiment of the present invention. A system for evaluating the veracity of news articles and images in real time and providing the results to a user comprises the following elements.

[1913] 1. System Configuration

[1914] This real-time fact-checking system mainly includes the following components:

[1915] Data Entry Module

[1916] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[1917] Data entered from the terminal is transmitted to the server.

[1918] Data Preprocessing Module

[1919] The server retrieves text data from the URL of the received news article and removes unnecessary information such as HTML tags and advertisements.

[1920] For image data, the server uses optical character recognition technology (OCR) to extract text information within the image.

[1921] The text data is automatically cleaned and converted into a format suitable for analysis.

[1922] Fact Checking Module

[1923] The server analyzes the cleaned text data or information in the images using natural language processing techniques.

[1924] The recognized text data is checked against reliable databases and other trusted sources to assess its veracity.

[1925] Judgment result generation module

[1926] The veracity of news and images is determined based on the results of fact-checking.

[1927] Judgments and fact-based summaries are automatically generated.

[1928] Result Delivery Module

[1929] The server transmits the generated judgment result and summary information to the user's terminal.

[1930] An application installed on a smartphone notifies the user of this information in real time.

[1931] The user can view the results and their summary through the terminal.

[1932] 2. Hardware and Software Requirements

[1933] The system uses a variety of hardware and software tools.

[1934] Hardware

[1935] Devices: smartphones, computers

[1936] Server: A server with high-performance processing capabilities

[1937] software

[1938] OCR technology: software for extracting text information from images (e.g., Tesseract OCR)

[1939] Natural language processing libraries: Tools for analyzing news articles and text data (e.g., SpaCy and NLTK)

[1940] Database connection software: Libraries for communicating with reliable databases

[1941] Smartphone app: An application that provides users with real-time information

[1942] Generative AI models: AI models used to generate summaries and analysis results for text data.

[1943] 3. Specific Examples

[1944] Specific usage scenarios include:

[1945] News article example

[1946] The user enters the URL of a news article that purports to say that a certain politician committed fraud.

[1947] The server retrieves the news article from the URL and preprocesses the text data.

[1948] The content of news articles is analyzed using natural language processing and compared with a reliable database.

[1949] The server determines that the article contains some errors and generates a summary news article based on accurate information.

[1950] Users can check the assessment results and summary in real time using a smartphone application.

[1951] Image example

[1952] Users upload "photos of specific incidents."

[1953] The server analyzes the image data using OCR technology and extracts text information.

[1954] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[1955] The server determines that the image is misleading and generates a factual explanation.

[1956] Users can check the assessment results and summary explanations in real time using a smartphone application.

[1957] 4. Examples of prompts

[1958] An example of a prompt that a user might enter into the system is:

[1959] "https: / / example.com / news-article Check whether this news article is true or not."

[1960] "Please verify this image is authentic."

[1961] This allows users to identify misinformation in real time and make quick decisions based on trusted sources of information.

[1962] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1963] Step 1:

[1964] User enters the URL or image data of a news article

[1965] Users use their smartphones or computers to enter the URL or image data of news articles into the system.

[1966] Input: News article URL or image data

[1967] Output: Input data is sent from the device to the server

[1968] Step 2:

[1969] The server preprocesses the received data

[1970] The server retrieves text data from the URLs of received news articles, removing unnecessary information such as HTML tags and advertisements, and extracting pure text data.

[1971] For image data, OCR technology is used to extract text information within the image.

[1972] Input: User-entered news article URL or image data

[1973] Output: Preprocessed text data or text information in images

[1974] Step 3:

[1975] The server analyzes the data using natural language processing technology

[1976] The server then analyzes the preprocessed text data and text information in the images using natural language processing techniques, which includes extracting words, understanding context, and summarizing information.

[1977] Input: Preprocessed text data or text information in images

[1978] Output: Parsed text data

[1979] Step 4:

[1980] The server checks the authenticity against a reliable database

[1981] The server then compares the parsed text data with reliable databases and other trusted sources, assessing the authenticity of the news article or image.

[1982] Input: Parsed text data

[1983] Output: Veracity evaluation result

[1984] Step 5:

[1985] The server generates the judgment result and summary information

[1986] The server generates a score based on the fact-check evaluation results to determine whether a news article or image is true or false, and creates a fact-based summary based on the evaluation results.

[1987] Input: Veracity assessment result

[1988] Output: Judgment results and summary information

[1989] Step 6:

[1990] The server sends the results to the device

[1991] The server then sends the generated judgment results and summary information to the user's device, and an application installed on the smartphone notifies the user in real time.

[1992] Input: Judgment results and summary information

[1993] Output: Notification and display on the user's terminal

[1994] Step 7:

[1995] User checks the results

[1996] Users can check the results and summary information sent from the server through a smartphone application, allowing them to determine the authenticity of news articles and images.

[1997] Input: Notification from the server and displayed results

[1998] Output: User's understanding and judgment

[1999] The above steps realize a system that evaluates the veracity of news articles and images in real time and provides appropriate information to users.

[2000] 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.

[2001] The present invention relates to a system that uses natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provides results based on the user's emotions. This system performs a series of processes, including preprocessing, analysis, fact-checking, emotion recognition, and generating and providing judgment results for news articles or image data.

[2002] System Overview

[2003] The system includes the following main components:

[2004] 1. Data Entry Module

[2005] This is a module that allows users to input news article URLs and image data into the system via their terminal.

[2006] The terminal transmits the input data to the server.

[2007] 2. Data Preprocessing Module

[2008] The server removes unnecessary data such as HTML tags and advertisements from the received news articles and extracts pure text data.

[2009] The server uses optical character recognition (OCR) technology on the image data to extract text information within the image.

[2010] The server analyzes the content of the image using image recognition technology.

[2011] 3. Fact-checking module

[2012] The server analyzes the preprocessed text data and the information in the images using natural language processing techniques.

[2013] The server checks the text data against reliable databases and other trusted sources to assess its veracity.

[2014] 4. Emotion Engine Module

[2015] The server analyzes the user's voice data and facial expression data acquired through the user's terminal and recognizes emotions.

[2016] The server adjusts the method of providing the judgment results and summary information based on the user's emotions recognized by the emotion engine.

[2017] 5. Judgment Result Generation Module

[2018] The server determines whether the news or image is true or false based on the results of the fact check.

[2019] The server generates factual summary information and presents it to the user in an easy-to-understand format.

[2020] 6. Result Delivery Module

[2021] The server transmits the generated judgment result and summary information to the user's terminal.

[2022] Users can view the judgment results and summary information on their device and confirm the authenticity of the news and images.

[2023] Program processing

[2024] The program processing of this system will be explained in natural language below.

[2025] 1. Receiving Data

[2026] A user uses a terminal to input the URL or image data of a news article into the system.

[2027] The terminal sends the input data to the server.

[2028] 2. Data Preprocessing

[2029] The server obtains the text data of the news article from the received URL.

[2030] The server removes unnecessary information such as HTML tags and advertisements from the text data.

[2031] The server analyzes the image data using OCR technology and extracts text information.

[2032] The server uses image recognition technology to analyze the content of the image and identify relevant information.

[2033] 3. Conduct fact-checks

[2034] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[2035] The server compares the analyzed data with reliable databases and news sources to assess the veracity of the information.

[2036] 4. Emotion Recognition by Emotion Engine

[2037] The server uses an emotion engine to analyze the user's voice data and facial expression data acquired through the user's device, and recognizes the user's current emotions.

[2038] For example, the emotional state of the user, such as whether they are surprised, angry, or relaxed, is grasped.

[2039] 5. Generating the judgment result

[2040] Based on the results of the fact check, the server generates a score that determines whether the news article or image is true or false.

[2041] The server generates a news summary based on reliable information, including judgments and facts from reliable sources.

[2042] 6. Tailored delivery of results

[2043] The server adjusts the display method of the judgment result and summary information based on the user's emotional state recognized by the emotion engine.

[2044] For example, if the user is surprised, provide more detailed information to make it easier for the user to understand.

[2045] 7. Providing Results

[2046] The server transmits the generated judgment result and summary information to the user's terminal.

[2047] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[2048] Specific examples

[2049] 1. News article example

[2050] A user enters the URL of a news article that says, "A politician committed fraud."

[2051] The server retrieves the news article from the URL and preprocesses the text data.

[2052] The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[2053] The server determines that this article contains some errors and generates a summary news based on accurate information.

[2054] The server recognizes the user's current emotion and, if they are surprised for example, provides a result with detailed supporting information.

[2055] Users can check the judgment results and summary on their device.

[2056] 2. Example images

[2057] A user uploads a "photo of a specific incident."

[2058] The server analyzes the image data using OCR technology and extracts text information.

[2059] The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[2060] The server determines that the image is misleading and generates a factual explanation.

[2061] The server recognizes the user's current emotions and, for example, if they are anxious, provides results that are thoughtfully reassuring.

[2062] Users can check the judgment results and summary explanation on their terminal.

[2063] As described above, the present invention is a system that utilizes natural language processing technology and an emotion engine to evaluate the veracity of news articles and images in real time and provide information that is individually tailored to the user's emotions. As a result, users can make decisions based on reliable information without being misled by potentially misleading information. Furthermore, by taking the user's emotions into consideration, a more user-friendly and easy-to-use system can be realized.

[2064] The processing flow will be explained below.

[2065] Step 1:

[2066] The user inputs the URL or image data of a news article into the device.

[2067] Step 2:

[2068] The device sends the entered URL and image data to the server.

[2069] Step 3:

[2070] The server scrapes text data from the URL of the news article received.

[2071] The server removes unnecessary data such as HTML tags and advertisements from the news article retrieved from the URL and extracts pure text data.

[2072] Step 4:

[2073] The server uses optical character recognition (OCR) technology on the image data it receives to extract text information from the image.

[2074] The server uses image recognition technology to analyze the content of the image and identify categories and objects (e.g., specific people or landscapes).

[2075] Step 5:

[2076] The server analyzes the preprocessed text data and information in the images using natural language processing technology (e.g., BERT, GPT-3) to understand the context and content of the article.

[2077] Step 6:

[2078] The server cross-checks the parsed data against reliable databases and other trusted sources (e.g., public fact-checking sites) to assess the veracity of the information.

[2079] Step 7:

[2080] Based on the results of the fact check, the server scores the news article or image as true or false.

[2081] Step 8:

[2082] The server uses an emotion engine to analyze the voice and facial expression data it receives from the user's device and recognizes the user's current emotions.

[2083] Step 9:

[2084] The server adjusts the method of providing the judgment results and summary information based on the user's emotional state.

[2085] For example, if the user is surprised, add more information in an easy-to-understand format.

[2086] Step 10:

[2087] The server generates an accurate news summary, including judgments and facts from reliable sources.

[2088] Step 11:

[2089] The server transmits the generated judgment result and summary information to the user's terminal.

[2090] Step 12:

[2091] Users can view the judgment results and summary information on their device and confirm the authenticity of news articles and images.

[2092] Example 2

[2093] 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."

[2094] In modern society, quickly and accurately assessing the veracity of news articles and images has become an important issue. The inability to instantly determine the reliability of such information can lead to the spread of false information and misleading images, which can have a significant impact on society. Furthermore, as the amount of information increases, it is becoming increasingly difficult for users to quickly verify the authenticity of information. Furthermore, information provided does not take into account the user's feelings, which can lead to misunderstandings and anxiety.

[2095] 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.

[2096] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their authenticity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data to recognize emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts and transmitting it to the terminal in a form that corresponds to the user's emotional state. This makes it possible to quickly and accurately evaluate the authenticity of news articles and images and provide information that corresponds to the user's emotions.

[2097] A "news article" is text data containing current information and news reports that are primarily published on websites and digital media.

[2098] "Image data" refers to file formats (e.g., JPEG, PNG) that contain visual information stored in a digital format.

[2099] A "terminal" is an electronic device such as a computer or smartphone that is used to input news articles and image data and view the results.

[2100] "Server" is a central control device that provides resources for data processing and analysis and manages the operation of the entire system.

[2101] "Preprocessing" refers to the process of removing unnecessary information from news articles or image data and converting them into a format suitable for analysis.

[2102] "Text data" refers to purely textual information extracted from news articles or image data.

[2103] "Optical character recognition technology" refers to technology (e.g., OCR) used to convert text in an image into digital characters.

[2104] "Natural language processing technology" refers to computer science technology for analyzing the meaning of text data and understanding its context and content.

[2105] A "trusted database" is a data repository containing verified information sources that can be used to assess veracity.

[2106] "Emotion recognition technology" is a technology that analyzes a user's voice data and facial expression data to understand their emotional state.

[2107] "Decision result" refers to the conclusion reached after evaluating the veracity of news articles or image data.

[2108] "Summary information" refers to information that succinctly summarizes the important points of a news article or image data.

[2109] A "reliable source" is a source or database that is widely recognized as providing truthful information.

[2110] This invention relates to a real-time fact-checking system that utilizes natural language processing and emotion recognition technologies. This system evaluates the veracity of news articles and image data and provides information according to the user's emotions, thereby realizing the provision of highly reliable information.

[2111] System configuration and operation

[2112] The system includes the following major components:

[2113] server

[2114] Device (PC or smartphone)

[2115] Receiving data

[2116] The user inputs the URL and image data of the news article into the terminal, which then sends it to the server.

[2117] Data Preprocessing

[2118] The server downloads the HTML data from the URL of the received news article and extracts the text data using an HTML parser (e.g., BeautifulSoup). It then removes unnecessary information, such as HTML tags and advertisements, from the text data. For image data, it extracts the text information from the image using an OCR engine (e.g., Tesseract OCR) and analyzes the image content using an image recognition model (e.g., ResNet, YOLO).

[2119] Data analysis and evaluation

[2120] The server inputs the preprocessed text data and information in the images into natural language processing models (e.g., BERT, GPT-3) to analyze the context and content, and then compares the information with reliable databases (e.g., Wikidata, FactCheck.org) and news sources via API to assess the veracity of the information.

[2121] Emotion recognition

[2122] The server inputs the voice data and facial expression images sent from the device into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to understand the user's emotional state. For example, it recognizes emotions such as surprise, anger, or relaxation.

[2123] Generation of judgment results and summary information

[2124] The server determines whether a news article or image is true or false based on the fact-check results and generates a truthfulness score. The server then uses natural language generation technology (e.g., GPT-3) to create a news summary, which includes the results of the judgment and facts based on reliable sources.

[2125] Providing information

[2126] The server adjusts the display method of the judgment results and summary information according to the user's emotional state. For example, if the user is surprised, it provides more detailed information, and if the user is relaxed, it provides simple information. The generated information is sent to the terminal, where the user can view it.

[2127] Specific examples

[2128] News article example

[2129] 1. The user enters the URL of a news article that says, "A politician committed fraud."

[2130] 2. The server retrieves the news article from the URL and preprocesses the text data.

[2131] 3. The server analyzes the text data using natural language processing technology and compares it with a reliable database.

[2132] 4. The server determines that this article contains some errors and generates a summary news article based on accurate information.

[2133] 5. The server recognizes the user's emotion and provides a result with detailed supporting information if, for example, they are surprised.

[2134] 6. The user checks the judgment results and summary on the terminal.

[2135] Image example

[2136] 1. A user uploads a photo of a specific incident.

[2137] 2. The server analyzes the image data using OCR technology and extracts text information.

[2138] 3. The server uses image recognition technology to analyze the content of the image and compare it with a reliable database.

[2139] 4. The server determines that the image is potentially misleading and generates a factual explanation.

[2140] 5. The server recognizes the user's emotions and provides reassuring information if, for example, the user is worried.

[2141] 6. The user checks the judgment result and summary explanation on the terminal.

[2142] Example prompt sentence:

[2143] The user is prompted to enter information in a format such as "Please enter the URL of a news article" or "Please upload image data."

[2144] As described above, the present invention makes it possible to evaluate the veracity of news articles and images in real time and provide information adjusted according to the user's emotions.

[2145] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2146] Program processing steps

[2147] Step 1: Receiving Data

[2148] A user uses a device to input a URL or image data of a news article.

[2149] Specific actions: Paste or upload the URL of a news article or image file into the input form displayed on the device.

[2150] Input: News article URL or image data.

[2151] Output: The received URL or image data.

[2152] Step 2: Sending data

[2153] The terminal transmits the input data to the server.

[2154] Specific operation: The device sends an HTTP POST request to the server, including the received URL or image data.

[2155] Input: Received URL or image data.

[2156] Output: URL or image data received by the server.

[2157] Step 3: Acquiring and preprocessing news articles

[2158] The server retrieves the text data of the news article from the received URL and removes unnecessary information.

[2159] What happens: The server uses an HTML parser (e.g. BeautifulSoup) to extract the text data and remove HTML tags and advertisements.

[2160] Input: The received URL.

[2161] Output: Pure text data.

[2162] Step 4: OCR analysis of image data

[2163] The image data received by the server is analyzed using an OCR engine to extract text information.

[2164] What happens: The server uses an OCR engine (e.g., Tesseract OCR) to extract text information from the image.

[2165] Input: Received image data.

[2166] Output: The extracted text information.

[2167] Step 5: Image content analysis

[2168] The server analyzes the content of the image using an image recognition model.

[2169] Specific operation: The server inputs image data into an image recognition model (e.g., ResNet, YOLO) and identifies the content of the image (people, objects, scenes, etc.).

[2170] Input: Received image data.

[2171] Output: Analyzed image content.

[2172] Step 6: Analyzing text and image data

[2173] The server analyzes the preprocessed text data and the extracted text in the image using natural language processing technology.

[2174] Specific operation: The server inputs text data into a natural language processing model (e.g., BERT, GPT-3) and analyzes the context and content.

[2175] Input: Pure text data and extracted text information.

[2176] Output: Semantic information of the parsed text and image data.

[2177] Step 7: Veracity Assessment

[2178] The server checks the parsed data against a reliable database to assess its veracity.

[2179] Specific operation: The server collates the parsed data via the database API and evaluates its authenticity.

[2180] Input: Parsed text and image data.

[2181] Output: A veracity rating score.

[2182] Step 8: Recognize emotions

[2183] The server analyzes the user's voice data and facial expression data to recognize the user's emotional state.

[2184] Specific operation: The server inputs voice data and facial expression data into an emotion recognition engine (e.g., OpenFace, IBM Watson Emotion Recognition) to recognize emotions.

[2185] Input: User's voice and facial expression data.

[2186] Output: The perceived emotional state of the user.

[2187] Step 9: Generate decision results and summary information

[2188] The server generates a judgment result based on the truthfulness evaluation and creates a summarized news.

[2189] Specific operation: The server generates a news summary using natural language generation technology (e.g., GPT-3). The server creates a news summary based on the judgment results.

[2190] Input: Truthfulness rating score.

[2191] Output: Judgment results and summary news.

[2192] Step 10: Coordinating input

[2193] The server provides the results in a coordinated manner.

[2194] Specific behavior: The server adjusts the way information is displayed depending on the user's emotional state, choosing whether to provide detailed or concise information.

[2195] Input: The perceived emotional state of the user and the judgment result.

[2196] Output: Adjusted judgement results and summary news.

[2197] Step 11: Sending the results

[2198] The server transmits the generated judgment result and summary information to the user's terminal.

[2199] Specific behavior: The server sends the result using push notification or HTTP response.

[2200] Input: Adjusted judgement results and summary news.

[2201] Output: The results sent to the terminal.

[2202] Step 12: View the results

[2203] The user checks the judgment results and summary information on the terminal.

[2204] Specific operations: The user uses the device to view the received judgment results and summary information.

[2205] Input: The result sent to the terminal.

[2206] Output: User understanding and judgment.

[2207] This concludes the program processing flow for this system, including the specific operations performed at each step. This enables the system to quickly and accurately evaluate the veracity of news articles and images, and provide information tailored to the user's emotions.

[2208] (Application example 2)

[2209] 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."

[2210] In modern society, the proliferation of fake news and fabricated images has led to the problem of many people making decisions based on incorrect information. This problem is a factor that causes social confusion and misunderstanding. Furthermore, while user emotions have a significant impact on determining the authenticity of information, current systems provide uniform information without taking user emotions into consideration, which reduces user understanding and satisfaction.

[2211] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[2212] In this invention, the server includes means for receiving news articles or image data from a terminal, means for preprocessing the received data and extracting text data or text information from the images, means for analyzing the extracted text data or information from the images and evaluating their veracity by comparing it with a reliable database or other reliable information sources, means for analyzing the user's voice data and facial expression data and recognizing emotions, means for determining whether the news or images are true or false based on the evaluation results, and means for generating summary information based on the evaluation results and facts, adjusting the display method of the result based on the user's emotional state, and transmitting the summary information to the terminal. This makes it easier for users to obtain individually tailored, reliable information without being misled by erroneous information.

[2213] A "news article" is text data and image data containing news information distributed online or offline.

[2214] "Image data" means files containing visual information stored and transmitted in digital format.

[2215] "Natural language processing technology" is a technology that uses computers to understand, interpret, and generate human language.

[2216] A "real-time fact-checking system" is a system that instantly evaluates the veracity of information and provides the results immediately.

[2217] "Data preprocessing" is the process of removing unnecessary information from received data and converting it into a format suitable for analysis.

[2218] "Text data" is a data format that includes information about sentences and characters.

[2219] "Extract" refers to the operation of extracting necessary parts from data.

[2220] A "database" is a collection of information that systematically stores large amounts of information and makes it searchable and usable.

[2221] A "reliable source" is an entity or medium that provides information that is highly accurate, reliable, and has been assessed.

[2222] "Emotion recognition" is a technology that analyzes voice data and facial expression data to determine a user's emotional state.

[2223] "Evaluation results" are the conclusions or grades reached after determining the truth or falsity of the information.

[2224] "Summary information" is the result of summarizing a large amount of information in a concise manner.

[2225] A "terminal" is a device such as a computer, smartphone, or tablet that is connected to the Internet or a network.

[2226] "Result presentation method" refers to the means or format in which information is presented to the user.

[2227] The system that realizes this application example evaluates the veracity of news articles and images in real time and provides results according to the user's emotions. Below is a natural language explanation of the program processing of this system.

[2228] The server first receives the URL or image data of the news article from the user's device. The device then sends the input data to the server. This process requires an internet connection, and the user can use a device such as a smartphone or tablet.

[2229] The server retrieves the text data of the news article from the received URL, and then uses a data preprocessing module to remove unnecessary information such as HTML tags and advertisements, converting it into pure text data. This process uses BeautifulSoup as a scraping library and Tesseract OCR as optical character recognition (OCR) technology.

[2230] The server then analyzes the extracted text data and textual information within the images and uses natural language processing techniques (using Hugging Face's Transformer library for its NLP model) to assess their veracity by matching them with trusted databases and other reliable sources, such as public databases and established news sources.

[2231] The server also includes an emotion recognition module to analyze the user's voice and facial expression data to recognize the user's emotional state. For this purpose, it uses OpenCV and the DeepFace library, which allows it to understand the user's emotions, such as surprised, angry, relaxed, etc.

[2232] The server determines whether a news item or image is true or false based on the fact-check results and generates a score. It then generates a news summary based on reliable information, including the results of the fact-check and facts from reliable sources. The server then adjusts the display of the results and summary information based on the user's emotional state, as recognized by the emotion engine. For example, if the user is surprised, the server provides more detailed information to make it easier for the user to understand.

[2233] The results are then sent from the server to the user's device, where the user can view the results and summary information to verify the authenticity of the news article or image.

[2234] Specific examples

[2235] As a concrete example, the following scenario can be considered.

[2236] A user enters the URL of a news article about a politician committing fraud. The server retrieves the news article from the URL and preprocesses the text data. Next, the server analyzes the text using natural language processing technology and compares it with a reliable database. The server determines that the article contains some errors and generates a summary news based on accurate information. Furthermore, the server recognizes the user's current emotions and provides results with detailed supplementary information if the user is surprised. The user can check the results and summary on their device.

[2237] Prompt Sentence Examples

[2238] The user inputs the URL or image data of a news article. The system receives the URL or image data and performs preprocessing. It then analyzes the data using natural language processing techniques and compares it with a reliable database. It recognizes the user's emotion (e.g., joy) and generates and displays results according to that emotion.

[2239] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2240] Step 1:

[2241] The user inputs the URL and image data of the news article via the terminal.

[2242] Input: News article URL or image data

[2243] How it works: A user provides data through an input form on a device such as a smartphone or tablet.

[2244] Output: The entered data is sent from the terminal to the server.

[2245] Step 2:

[2246] The server receives the URL or image data of the news article.

[2247] Input: URL of news article or image data sent from the device

[2248] How it works: The server receives the HTTP request and stores the data locally.

[2249] Output: The received data is stored on the server and handed over to the next process.

[2250] Step 3:

[2251] The server preprocesses the data received.

[2252] Input: URL or image data of a saved news article

[2253] How it works: For news articles, it uses BeautifulSoup to remove HTML tags and ads and extract pure text data. For image data, it uses Tesseract OCR to extract text information from images.

[2254] Output: Preprocessed text data or text information in images.

[2255] Step 4:

[2256] The server analyzes the preprocessed data.

[2257] Input: Preprocessed text data or text information in images

[2258] How it works: Uses Hugging Face's Transformer library for natural language processing to understand the context and content of data.

[2259] Output: Parsed data.

[2260] Step 5:

[2261] The server checks the parsed data against a reliable database to assess its veracity.

[2262] Input: Parsed data

[2263] How it works: The parsed data is compared to public databases and established news sources to generate a credibility score.

[2264] Output: A truthfulness score.

[2265] Step 6:

[2266] The server analyzes the user's voice data and facial expression data to recognize their emotional state.

[2267] Input: User's voice data and facial expression data

[2268] How it works: It uses OpenCV and DeepFace libraries to analyze user emotions and identify their emotional state.

[2269] Output: The user's emotional state.

[2270] Step 7:

[2271] The server makes a decision based on the results of the fact check.

[2272] Input: Truthfulness score, user emotional state

[2273] How it works: Based on the evaluation results, it determines whether a news item or image is true or false, generates an appropriate score, and adjusts how the results are displayed based on the user's emotional state.

[2274] Output: Judgment results and how they are displayed.

[2275] Step 8:

[2276] The server generates the judgment result and summary information and transmits them to the terminal.

[2277] Input: Judgment result, display method

[2278] How it works: The server generates news summaries based on reliable information, presents them in a format that is easy for users to understand, and sends the adjusted results to the user's device.

[2279] Output: The judgment result and summary information sent to the user's terminal.

[2280] Step 9:

[2281] The user views the provided results and summary.

[2282] Input: Judgment result and summary information sent from the server

[2283] How it works: Users check the information displayed on their smartphones, tablets, or other devices to verify the authenticity of news articles and images.

[2284] Output: User understanding and judgment.

[2285] 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.

[2286] 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.

[2287] 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 robot 414.

[2288] 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.

[2289] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2290] 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.

[2291] 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).

[2292] 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.

[2293] 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."

[2294] 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.

[2295] 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).

[2296] 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.

[2297] 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.

[2298] 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.

[2299] 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.

[2300] 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.

[2301] 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.

[2302] 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 har...

Claims

1. A real-time fact-checking system that uses natural language processing technology to assess the veracity of news articles and images. means for receiving news articles or image data from a terminal; means for preprocessing the received data to extract text data and text information within images; A means of analyzing the extracted text data and information in the images and assessing their veracity by comparing them with reliable databases and other reliable sources; and A means for determining whether the news or image is true or false based on the evaluation result; means for generating summary information based on the judgment results and facts and transmitting the summary information to a terminal; A system including:

2. 2. The system of claim 1, wherein the preprocessing means removes HTML tags and unnecessary data from the news articles to extract pure text data.

3. 2. The system of claim 1, wherein the preprocessing means applies optical character recognition techniques to the image data to extract text information within the image.

Images