System

Abstract

A system is provided.SOLUTION: A system comprising: means for a user to upload digital content created by a production AI; means for a server to receive and temporarily store the digital content; means for the server to analyze the digital content and resolve production processes; means for the server to identify learning sources based on the digital content; means for the server to identify which portions of a production depend on which learning sources based on the learning sources; and means for the server to provide the identified information to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] With the recent advancement of generative AI, it has become increasingly difficult to clarify the original learning source and information used in generated digital content. For this reason, it is important to ensure traceability to determine whether the resulting content constitutes copyright infringement or other rights infringement. When companies or celebrities use generative AI to create content, they need a way to clarify the origin of the resulting content and minimize the reputational risk associated with its use. [Means for solving the problem]

[0005] This invention relates to a traceability system that clarifies the origin of digital content created by generative AI. The system includes a means for users to upload digital content created by generative AI, a server for temporarily storing and analyzing the received digital content, a means for breaking down the generation process to identify learning sources, a means for identifying which parts depend on which learning sources based on the identified learning sources, and a means for providing this information to users. This allows users to understand the details of the learning sources used by the generative AI and minimize the risk of copyright infringement.

[0006] "Digital content" is a general term for information expressed in digital form, such as images, text, audio, and video.

[0007] "Generative AI" refers to systems and algorithms that use artificial intelligence technology to automatically generate digital content.

[0008] "User" refers to an individual or corporation that provides digital content created from generative AI using this system.

[0009] "Server" refers to a computer system that receives digital content, analyzes, stores, searches, and provides information to users.

[0010] An "analysis module" refers to a software component that executes an analysis method according to the type of digital content.

[0011] "Learning Source" refers to the dataset or information source that the generative AI uses to generate the digital content.

[0012] "Learning source database" refers to a database that stores datasets and information sources collected for use by generative AI.

[0013] "Means of identification" refers to technical methods or devices for clearly identifying the target data or information from the analysis results or search results.

[0014] "Traceability" refers to the property or ability of generative AI to be able to trace the origin of the content it generates and the learning sources used.

[0015] "Traceability Report" refers to a report that compiles detailed information about the provenance of digital content created by generative AI. [Brief explanation of the drawings]

[0016] [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

[0017] 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.

[0018] First, the terms used in the following description will be explained.

[0019] 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).

[0020] 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.

[0021] 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.

[0022] 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.

[0023] 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."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 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.

[0027] 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).

[0028] 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.

[0029] 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.

[0030] 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.

[0031] 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.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] 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."

[0037] Overall system overview

[0038] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[0039] System configuration

[0040] The system includes the following major components:

[0041] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0042] 2. Server: A central system with the following functions:

[0043] Receiving digital content

[0044] Temporary storage of digital content

[0045] Digital content analysis

[0046] Searching the learning source database

[0047] Identifying the location of use

[0048] Generate and provide traceability reports

[0049] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0050] Program processing

[0051] The program of this system is executed in the following manner.

[0052] Providing and Receiving Products

[0053] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[0054] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0055] Digital content analysis

[0056] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0057] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[0058] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0059] Identifying learning sources

[0060] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[0061] Identifying the location of use

[0062] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0063] Generating and providing traceability information

[0064] The server generates a traceability report based on the identified information, which includes the URL of the learning source and the location of the identified part.

[0065] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[0066] Specific examples

[0067] Example 1: Image creation

[0068] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[0069] The server receives the images and adds them to the analysis queue.

[0070] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[0071] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0072] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[0073] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[0074] Example 2: Text production

[0075] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[0076] The server receives the article and adds it to the analysis queue.

[0077] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0078] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0079] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[0080] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[0081] The above is an embodiment of the present invention, which makes it possible to provide a specific system for clarifying the origin of digital content created using generative AI and minimizing the risk of copyright infringement.

[0082] The processing flow will be explained below.

[0083] Step 1:

[0084] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[0085] Step 2:

[0086] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[0087] Step 3:

[0088] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[0089] Step 4:

[0090] The server analyzes the digital content using the selected analysis module.

[0091] Image Analysis Module: Extracts key features from an image using techniques such as edge detection, color analysis, and shape recognition.

[0092] Text Analysis Module: Performs grammatical analysis, tokenization, keyword extraction, etc. to identify important keywords and phrases.

[0093] Step 5:

[0094] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on keywords.

[0095] Step 6:

[0096] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources. This analysis uses statistical and clustering techniques to perform a specific mapping, identifying which original information a particular image part or text phrase comes from.

[0097] Step 7:

[0098] The server generates a traceability report based on the results, including the URL of the learning source, the location of the citation, and related metadata. The report is generated in a detailed and easy-to-understand format.

[0099] Step 8:

[0100] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Based on the report, the user can clearly understand the origin and usage of the product.

[0101] These are the processing steps of the system, which allow users to check in detail the origin of digital content created by generative AI, reducing the risk of copyright infringement.

[0102] Example 1

[0103] 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."

[0104] Currently, it is difficult to clarify the origin of digital data created by an artificial intelligence model and provide appropriate traceability. Therefore, a system is needed that identifies the learning source from which digital data was generated and provides this information to the user. The present invention aims to solve this problem and provide an efficient and effective system for identifying the origin of generated digital data.

[0105] 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.

[0106] In this invention, the server includes: means for a user to upload digital data created by an AI model; means for the server to receive and temporarily store the digital data; means for the server to analyze the digital data and decompose the generation process; means for the server to identify learning sources based on the digital data; means for the server to perform a detailed analysis of which parts of a product depend on which learning sources based on the identified learning sources; and means for the server to generate a traceability report based on the identified information and provide it to the user. This makes it possible to clarify the origin of digital data created by an AI model and provide an appropriate traceability report to the user.

[0107] "User" means any person or entity that utilizes an artificial intelligence model to generate digital data and uploads that data to the System.

[0108] An "artificial intelligence model" refers to a program or system that uses machine learning or deep learning algorithms to automatically create artifacts from data.

[0109] "Digital data" refers to information that is stored, processed, and transmitted electronically, and this information can exist in the form of images, text, audio, video, etc.

[0110] "Server" refers to a computer system capable of processing, analyzing, storing, and serving digital data received from users.

[0111] "Analysis module" refers to a software component that analyzes digital data according to its type and extracts features and keywords.

[0112] "Learning source" refers to a dataset or information source used in the training process of an artificial intelligence model, providing the information that forms the basis of the generated digital data.

[0113] "Learning source database" refers to a database that stores the underlying data sets and information sources that identify the origin of generated digital data.

[0114] A "traceability report" is a report that clearly shows the relationship between the analyzed digital data and its learning source, including information indicating which original data a particular part is based on.

[0115] Overall system overview

[0116] The system of the present invention aims to clarify the origin of digital data created by a generative AI model and provide appropriate traceability. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[0117] System configuration

[0118] The system includes the following major components:

[0119] 1. User terminal: A device (PC, smartphone, etc.) through which a user uploads digital data created by a generative AI model.

[0120] 2. Server: A central system with the following functions:

[0121] Receiving digital data

[0122] Temporary storage of digital data

[0123] Digital Data Analysis

[0124] Searching the learning source database

[0125] Identifying the location of use

[0126] Generate and provide traceability reports

[0127] 3. Training source database: A database that stores the datasets and information sources used to train a generative AI model.

[0128] Program processing

[0129] The program of this system is processed as follows.

[0130] Providing and Receiving Products

[0131] Users upload digital data created by generative AI models to the system using a dedicated web interface or API, along with metadata such as content type and creation date and time.

[0132] The server receives the uploaded digital data, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0133] Digital Data Analysis

[0134] The server sequentially retrieves the digital data in the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0135] Image analysis module: Uses OpenCV etc. to extract key image features (edges, hue, shape, etc.).

[0136] Text analysis module: Uses NLTK etc. to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[0137] Identifying learning sources

[0138] The server searches the training source database based on the extracted features and keywords, and uses SQL or Elasticsearch to query the database and identify data with high similarity.

[0139] Identifying usage locations and generating traceability information

[0140] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources, and determines the correspondence between features and keywords and learning sources using statistical and clustering techniques.

[0141] The server generates a traceability report based on the analysis results, including the correspondence between specific parts of the artifact and the learning source.

[0142] Providing traceability reports

[0143] The server provides the generated traceability reports to the user, which can be downloaded via a web interface or API, in PDF or HTML format.

[0144] Specific examples

[0145] Example 1: Image creation

[0146] Users upload vibrant landscape paintings created by generative AI models to the system.

[0147] The server receives the images and adds them to the analysis queue.

[0148] The server selects an image analysis module to extract the main features of the image (edges, color histogram).

[0149] The server searches the training source database based on the extracted features to identify datasets of similar landscape images.

[0150] The server analyzes which original image a particular part of the landscape painting comes from and performs a specific mapping.

[0151] The server generates a traceability report and provides it to the user, including information such as "The mountain part is from original image X, and the river part is from original image Y."

[0152] Example 2: Text production

[0153] Users upload articles about the latest technology trends to the system, which are created using a generative AI model.

[0154] The server receives the article and adds it to the analysis queue.

[0155] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0156] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0157] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping.

[0158] The server generates a traceability report and provides it to the user, which indicates which original article a particular technical term or explanation is based on.

[0159] In this way, the system of the present invention can clarify the origin of digital data created using a generative AI model and provide users with an appropriate traceability report.

[0160] Examples of prompt statements

[0161] For images: "Please identify the dataset or training source from which this image originated."

[0162] For text: "Please identify the source or quoted passage from which this article was based."

[0163] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0164] Step 1:

[0165] A user uploads digital data (such as images or text) created by a generative AI model to the system using a dedicated web interface or API. The input is the digital data and its metadata (content type, creation date and time), and the output is an upload request sent to the server. Specifically, the user opens a browser or a dedicated application, selects a file, enters metadata, and clicks the upload button.

[0166] Step 2:

[0167] The server receives an upload request from a user and temporarily stores the digital data and metadata. The input is the digital data and metadata sent by the user, and the output is the data stored in the server's temporary storage database. Specifically, the server saves the uploaded data in a specified directory (such as / tmp / uploads) and simultaneously records the metadata in the database.

[0168] Step 3:

[0169] The server adds the information of the stored digital data to the analysis queue. The input is the path and metadata of the temporarily stored digital data, and the output is the task added to the analysis queue. Specifically, the server adds the file path and metadata to a queue system (e.g., RabbitMQ or Kafka).

[0170] Step 4:

[0171] The server retrieves digital data from the analysis queue and selects an appropriate analysis module based on its type (image, text, etc.). The input is the data and metadata retrieved from the analysis queue, and the output is the selection of an analysis module. Specifically, the server checks the metadata and selects an image analysis module if the analysis target is an image, or a text analysis module if the analysis target is text.

[0172] Step 5:

[0173] The server analyzes the digital data using the selected analysis module to extract features and keywords. The input is the digital data entered into the analysis module, and the output is the extracted features and keywords. Specifically, the image analysis module uses OpenCV to extract edges and color histograms, and the text analysis module uses NLTK to perform grammar analysis and tokenization.

[0174] Step 6:

[0175] The server searches the training source database based on the extracted features and keywords to identify similar data. The input is the extracted features and keywords, and the output is information about similar training source data. Specifically, the server uses SQL or Elasticsearch to search the database and identify data with high similarity scores.

[0176] Step 7:

[0177] The server analyzes the search results in detail and identifies which parts of the generated digital data depend on which learning sources. The input is the search results, and the output is information showing the correspondence between the generated data and the learning sources. Specifically, statistical methods and clustering techniques are used to map specific parts to the learning sources.

[0178] Step 8:

[0179] The server generates a traceability report based on the identified information. The input is information showing the correspondence between the product and the learning source, and the output is a traceability report. Specifically, the report includes the URL of the learning source and the location information of the identified part, and is generated in PDF or HTML format.

[0180] Step 9:

[0181] The server provides the generated traceability report to the user. The input is the traceability report, and the output is the report in a format that the user can download. Specifically, the user logs into the web interface and clicks on a link to view and download the report.

[0182] (Application example 1)

[0183] 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."

[0184] The origin of digital content generated by generative AI models is often unclear, potentially raising issues regarding the reliability and copyright of the content. It is also difficult to track which learning source user-generated content was based on. Furthermore, there is no established method for utilizing this information in real time. To solve these problems, a system is needed that can clarify the origin of digital content and analyze its generation process.

[0185] 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.

[0186] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and decompose the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of the product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for a smartphone or robot to analyze the digital content in real time; and a means for visually displaying the information identified by the server. This makes the origin of the digital content clear, enabling the use of highly reliable digital content. Furthermore, real-time origin verification is also possible, allowing users to quickly verify the reliability of the content.

[0187] "Digital Content" refers to information that is generated, stored, and transmitted by electronic means.

[0188] "Generative AI" refers to systems or technologies that use artificial intelligence to generate new digital content.

[0189] "Uploading" refers to the act of a user sending digital content from their own terminal to a server.

[0190] "Server" refers to a computer system that receives, stores, processes, and serves digital data.

[0191] "Analysis" refers to the technical process of breaking down the content of digital content and clarifying its characteristics and the process by which it was created.

[0192] "Learning source" refers to the dataset or information source that generative AI uses to generate new content.

[0193] "Real-time" refers to processing and results being provided immediately.

[0194] "Visual display" refers to displaying analysis results and traceability information on a screen or display in a format that is easy for users to understand.

[0195] "Feature extraction" refers to the technique of identifying and extracting important attributes and patterns from digital content.

[0196] "Traceability" refers to the ability to track and record the origin of a product or the provenance of its components.

[0197] Overall system overview

[0198] This invention is a system that provides traceability functions to clarify the origin of digital content created by generative AI. Users upload the creations, and the server analyzes them to identify the learning source and usage points, and provides that information to the user. This system can be applied to security services, particularly applications installed on smartphones or robots that perform real-time analysis.

[0199] System configuration

[0200] 1. User Device

[0201] A user terminal is a device that users use to upload digital content created by generative AI. This includes smartphones, tablets, and PCs.

[0202] 2. Server

[0203] The server is the central component with the following functions:

[0204] Receiving and temporarily storing digital content: Receives and temporarily stores digital content uploaded by users.

[0205] Digital content analysis: Analyze digital content and break down its characteristics and creation process.

[0206] Identifying learning sources: Based on the analysis results, the learning sources are searched from the database.

[0207] Identifying the parts of the product that are based on learning sources and identifying which learning sources they depend on.

[0208] Providing traceability information: The identified information is provided to the user as a traceability report.

[0209] 3. Learning Source Database

[0210] The training source database is a database that stores the datasets and information sources used to train the generative AI. This database contains a large amount of datasets (image data, text data).

[0211] 4. Smartphone Robot

[0212] This system can be applied as an application installed on smartphones or robots. These devices have the ability to analyze digital content and provide traceability information in real time.

[0213] Program processing overview

[0214] Uploading Digital Content

[0215] Users upload digital content using a dedicated interface (for example, a web browser or a dedicated application). When uploading, metadata such as the content type and creation date and time are also added.

[0216] Receiving and temporarily storing digital content

[0217] The server receives and temporarily stores the digital content uploaded by the user, and once the storage is complete, adds the digital content to an analysis queue.

[0218] Digital content analysis

[0219] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.). For example, for images, it identifies edges and color histograms, while for text, it performs grammar analysis and keyword extraction.

[0220] Identifying learning sources

[0221] The server uses the analysis results to search the training source database to identify similar datasets, using feature matching and clustering techniques.

[0222] Identifying the location of use

[0223] The server performs detailed analysis of specific parts of the production based on the learning source and clarifies which learning source it depends on, thereby generating highly reliable traceability information.

[0224] Providing traceability information

[0225] The server generates a traceability report based on the identified information, which includes the URL of the learning source and details of the specific part, and can be viewed and checked by the user.

[0226] Examples of concrete examples and prompts

[0227] For example, to verify the originality of a photo taken with a smartphone, the user uploads the photo, the server analyzes the photo's characteristics, and searches the learning source database. A traceability report is generated based on the search results and provided to the user.

[0228] Example prompt sentence:

[0229] Task: Trace provenance

[0230] Content Type: Image

[0231] File path: 'path / to / image.jpg'

[0232] This system allows users to quickly verify the authenticity of digital content and obtain traceability information in real time.

[0233] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0234] Step 1: Upload your digital content

[0235] Users upload digital content using their smartphones or computers. They use a dedicated web interface or application to send the digital content file and its metadata (content type, creation date, etc.). The input is the digital content file selected by the user, and the output is the data received by the server. The specific action is when the user clicks the "Upload" button.

[0236] Step 2: Receive and temporarily store digital content

[0237] The server receives digital content sent by the user and temporarily stores it in storage. The input is the digital content file sent by the user, and the output is a file stored in the server's temporary storage area. Specifically, the server analyzes the HTTP request and writes the file data to disk.

[0238] Step 3: Analyzing the digital content

[0239] The server retrieves the temporarily stored digital content and selects an analysis module depending on the type. For example, for image files, it performs edge detection and color histogram analysis. For text files, it performs grammatical analysis and keyword extraction. The input is the temporarily stored digital content file, and the output is the analyzed feature data. Specifically, the server loads the analysis module and analyzes the data.

[0240] Step 4: Identify learning sources

[0241] The server searches the training source database based on the analysis results. It uses the extracted feature data as input to identify similar datasets. The input is the analyzed feature data, and the output is similar datasets identified from the training source database. Specifically, the server inputs the feature values ​​into a clustering algorithm and calculates the similarity.

[0242] Step 5: Identify the usage

[0243] The server determines which learning source a particular part of the digital content depends on based on information obtained from the learning source database. The input is information on similar datasets obtained from the learning source database, and the output is mapping information between each part of the product and its corresponding learning source. Specifically, the server uses statistical methods to match each part of the digital content with the learning source.

[0244] Step 6: Generate traceability information

[0245] The server generates a traceability report based on the identified usage location information. The report includes the URL of the learning source and detailed information about the specific part. The input is the mapping information, and the output is the traceability report provided to the user. Specifically, the server creates a report format based on the mapping information and fills in the data.

[0246] Step 7: Provide traceability information

[0247] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The input is the traceability report, and the output is a visualized report that the user can access. Specifically, the server returns the report upon the user's request.

[0248] As described above, this system is composed of multiple steps to clarify the origin of digital content and guarantee its reliability.

[0249] 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.

[0250] Overall system overview

[0251] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI, and an emotion engine that recognizes user emotions and optimizes analysis results. In this system, a user provides a product, a server analyzes the product to identify learning sources and usage points, adjusts the analysis results using the emotion engine, and provides that information to the user.

[0252] System configuration

[0253] The system includes the following major components:

[0254] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0255] 2. Server: A central system with the following functions:

[0256] Receiving digital content

[0257] Temporary storage of digital content

[0258] Digital content analysis

[0259] Searching the learning source database

[0260] Identifying the location of use

[0261] Generate and provide traceability reports

[0262] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0263] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[0264] Program processing

[0265] The processing of the system programs will be explained in detail below.

[0266] Providing and Receiving Products

[0267] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[0268] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0269] Digital content analysis

[0270] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on the type (image, text, etc.). For example, it launches an image analysis module for images and a natural language processing module for text.

[0271] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[0272] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0273] Identifying learning sources and where to use them

[0274] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[0275] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0276] Use of emotion engine

[0277] The server activates an emotion engine to recognize the user's emotion.

[0278] The emotion engine analyzes the user's facial expressions and voice to identify emotions, for example, by using a webcam or microphone to collect real-time emotion data.

[0279] The emotion engine adjusts the analysis results based on the identified emotion, for example providing more detailed traceability information if the user is confused, or a concise report if the user is happy.

[0280] Generating and providing traceability information

[0281] The server combines the identified information with the results of the sentiment engine to generate a traceability report, which includes the URL of the learning source, the location of the citation, and associated metadata.

[0282] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[0283] Specific examples

[0284] Example 1: Image creation

[0285] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[0286] The server receives the images and adds them to the analysis queue.

[0287] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[0288] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0289] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[0290] The server activates the emotion engine, which analyzes the user's facial expressions and voice to identify their emotion. For example, if the user is confused, a more detailed explanation is added to the report.

[0291] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[0292] Example 2: Text production

[0293] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[0294] The server receives the article and adds it to the analysis queue.

[0295] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0296] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0297] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[0298] The server starts the emotion engine, which analyzes the user's emotions. For example, if the user has a positive feeling, it generates a concise and positive report.

[0299] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[0300] The above is an embodiment of the present invention. The present invention makes it possible to clarify the origin of digital content created using generative AI and provide an optimal report that corresponds to the user's emotions.

[0301] The processing flow will be explained below.

[0302] Step 1:

[0303] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[0304] Step 2:

[0305] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[0306] Step 3:

[0307] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[0308] Step 4:

[0309] The server analyzes the digital content using the selected analysis module.

[0310] Image Analysis Module: Uses feature extraction algorithms to identify key features of an image (edges, color, shape, etc.).

[0311] Text analysis module: Using natural language processing technology, it analyzes the structure of text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0312] Step 5:

[0313] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on the extracted keywords.

[0314] Step 6:

[0315] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0316] Step 7:

[0317] The server activates an emotion engine to recognize the user's emotion.

[0318] The emotion engine collects facial and voice data from the user and identifies their emotions based on this data. The data is collected in real time using a webcam and microphone.

[0319] An emotion engine tailors the presentation of the analysis results based on the identified emotion.

[0320] Step 8:

[0321] The server uses the emotion engine to generate a traceability report based on the refined analysis, including the URL of the learning source, the location of the citation, and related metadata.

[0322] Step 9:

[0323] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Detailed or concise reports are provided based on emotions to help users understand.

[0324] Example: Image production

[0325] 1. A user uploads an image of a landscape painting to the system.

[0326] 2. The server receives the image and adds it to the analysis queue.

[0327] 3. The server selects an image analysis module and extracts key features.

[0328] 4. The server searches the learning source database based on the extracted features to identify the original image.

[0329] 5. The server maps specific parts of the image to the original image.

[0330] 6. The server launches the emotion engine, which analyzes the user's facial expressions and voice in real time to identify their emotions.

[0331] 7. The server generates a traceability report based on the identified emotions in a format that is understandable to the user.

[0332] 8. The server provides the report to the user, who can view it via a web interface.

[0333] Example: For text productions

[0334] 1. A user uploads a technical article to the system.

[0335] 2. The server receives the article and adds it to the analysis queue.

[0336] 3. The server selects a text analysis module and performs a structural analysis of the text.

[0337] 4. The server searches the learning source database based on the extracted keywords to identify the original article.

[0338] 5. The server maps specific parts of the technical article to the original article.

[0339] 6. The server launches the emotion engine and analyzes the user's emotions in real time.

[0340] 7. The server generates a traceability report in an appropriate format based on the identified emotion.

[0341] 8. The server provides the report to the user, who can view it via a web interface.

[0342] This will clarify the provenance of digital content using generative AI and provide optimal reporting based on user sentiment.

[0343] Example 2

[0344] 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."

[0345] The origin of digital content created by generative AI is unclear, making it impossible to identify which learning sources it relies on. Furthermore, there is a problem in that traceability information is difficult to understand because appropriate information is not provided according to user sentiment.

[0346] 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.

[0347] In this invention, the server includes a means for receiving and temporarily storing digital content created by the AI ​​by a user, a means for analyzing the digital content and breaking down the creation process, and a means for identifying learning sources based on the digital content. This makes it possible to clarify the origin of the digital content created by the AI ​​and provide appropriate information according to the user's emotions.

[0348] "User" means an individual or organization that uses the system to upload digital content created by generative AI.

[0349] "Generative AI" is a system that automatically generates digital content (images, text, etc.) using artificial intelligence technology.

[0350] "Digital content" refers to electronic data created by generative AI, including images, text, and audio.

[0351] "Uploading" is the act of a user sending digital content created by generative AI to a server.

[0352] "Server" means a centralized computer system that receives, analyzes, stores, and serves digital content.

[0353] A "temporary storage means" is a storage device or storage service that allows a server to temporarily store digital content.

[0354] "Means for analyzing" refers to software or hardware that the server uses to analyze digital content and extract its structure and characteristics.

[0355] "Means for decomposing the creation process" refers to analytical techniques that allow the server to clarify the elements and methods used to create digital content.

[0356] A "learning source" is a dataset or information source that a generative AI uses for learning or training.

[0357] "Means for identifying" refers to the technology or algorithm that the server uses to identify which learning source a piece of digital content relies on.

[0358] The "means of provision" refers to the interface or API that the server uses to communicate and display analysis results and traceability information to users.

[0359] The "means for analyzing emotions" refers to machine learning models and emotion recognition technologies that allow the server to analyze the user's facial expressions and voice and identify the user's emotions.

[0360] "Means for adjusting information" refers to technologies and algorithms for optimizing traceability information by taking into account the results of user sentiment analysis.

[0361] Overall system overview

[0362] This invention relates to a system that clarifies the origin of digital content created by generative AI, recognizes user emotions, and optimizes analysis results. By using this system, it is possible to clarify which dataset the generative AI used to generate the content, and provide information that meets the user's needs.

[0363] System configuration

[0364] The system includes the following major components:

[0365] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0366] 2. Server: A central system with the following functions:

[0367] Receiving digital content

[0368] Temporary storage of digital content

[0369] Digital content analysis

[0370] Searching the learning source database

[0371] Identifying the location of use

[0372] Generate and provide traceability reports

[0373] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0374] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[0375] Details of data processing and data calculation

[0376] The specific operation of the system will be described below.

[0377] Providing and Receiving Products

[0378] Users upload digital content created by generative AI through a dedicated web interface or API, and provide simple metadata (content type, creation date, etc.) at the time.

[0379] The server receives the uploaded digital content and temporarily stores it in a cloud storage location (e.g., Amazon S3). Once the storage is complete, the digital content is added to the analysis queue.

[0380] Digital content analysis

[0381] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0382] Image Analysis Module: Performs image feature extraction using the OpenCV library to identify key features (edges, hue, shape, etc.).

[0383] Text analysis module: Uses natural language processing libraries such as NLTK and spaCy to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[0384] Identifying learning sources and where to use them

[0385] The server searches the training source database based on the extracted features and keywords, for example, by using SQL queries or Elasticsearch to search the data.

[0386] The server then uses the search results to perform a detailed analysis of which learning sources specific parts of the artifacts depend on, applying statistical methods and machine learning algorithms such as K-means clustering and DBSCAN.

[0387] Use of emotion engine

[0388] The server runs an emotion engine, which analyzes the user's facial expressions and voice to identify emotions. It uses libraries such as OpenVINO and Dlib to identify emotions such as smiling, confused, or angry using data from the webcam and microphone.

[0389] The emotion engine adjusts the analysis results based on the identified emotion, for example adding more detailed information to the traceability report if the user is confused.

[0390] Generating and providing traceability information

[0391] The server combines the identified information (the URL of the learning source, the location of the quoted section, and related metadata) with the results of the user sentiment analysis and generates a traceability report using tools such as JasperReports and Crystal Reports.

[0392] The server provides the generated traceability report to the user, who can view and download it via the system's web interface or API.

[0393] Specific examples

[0394] Example 1: Image creation

[0395] Users upload vivid landscape images created by generative AI through the system.

[0396] The server receives the images and adds them to the analysis queue.

[0397] The server selects an image analysis module to extract key features such as edges and color histograms.

[0398] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0399] The server activates an emotion engine, which analyzes the user's facial expressions and voice to identify their emotions. For example, if the user is confused, more detailed information is added to the traceability report.

[0400] The server generates a traceability report, providing the user with details about the original source of the image.

[0401] Example 2: Text production

[0402] Users upload AI-generated articles about the latest technology trends through the system.

[0403] The server receives the article and adds it to the analysis queue.

[0404] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0405] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0406] The server will launch an emotion engine, which will analyze the user's emotion, for example, if the user is confused, it will add more detailed information to the traceability report.

[0407] The server generates a traceability report, providing the user with details about the original source of the technical article.

[0408] The above is a specific embodiment of the present invention. This system makes it possible to clarify the origin of digital content created by generative AI and provide optimal traceability information according to the user's feelings.

[0409] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0410] Step 1: Providing and Receiving Products

[0411] Users upload digital content (e.g., landscape images or technical articles) created by generative AI using the system's web interface or API. As input, the digital content and its metadata (content type, creation date, etc.) are provided.

[0412] The server receives the uploaded digital content, and the output is that the digital content and metadata are stored on the server.

[0413] The server temporarily stores the content in cloud storage (e.g., Amazon S3) and adds it to the analysis queue, so the content is ready for analysis.

[0414] Step 2: Analyzing the digital content

[0415] The server sequentially retrieves digital content from the analysis queue, and the input is the digital content added to the analysis queue.

[0416] The server selects an analysis module based on the type of digital content (image, text, etc.), for example, an image analysis module for images and a text analysis module for text.

[0417] The server launches the selected modules to analyze the digital content. Specifically, the image analysis module uses the OpenCV library to analyze edges, hues, shapes, etc., while the text analysis module uses NLTK and spaCy to perform grammar analysis and keyword extraction. The output is the extracted features and keywords.

[0418] Step 3: Identify learning sources and where to use them

[0419] The server uses the features and keywords obtained from the analysis results to search the learning source database. The input is the features and keywords from the analysis results.

[0420] The server uses SQL queries or Elasticsearch to identify relevant datasets from the training source database, and the output is the relevant training source data.

[0421] The server uses statistical methods and machine learning algorithms such as K-means clustering and DBSCAN to perform a detailed analysis of which learning sources specific parts of the product depend on, thereby obtaining mapping information between the product and the learning sources.

[0422] Step 4: Use the Emotion Engine

[0423] The server starts the emotion engine, and inputs are the user's facial expressions and voice data.

[0424] The emotion engine uses libraries such as OpenVINO and Dlib to analyze the user's facial expressions and voice in real time to identify emotions. The output is the user's emotional information.

[0425] The server adjusts the analysis results based on the emotion engine's results, for example adding more detailed information to the traceability report if the user is confused.

[0426] Step 5: Generate and provide traceability information

[0427] The server integrates the identified information (the URL of the learning source, the location of the quoted part, and related metadata) with the sentiment analysis results to generate a traceability report. The input is the learning source information and the sentiment analysis results.

[0428] The server generates the traceability report using tools such as JasperReports or Crystal Reports. The output is a comprehensive traceability report.

[0429] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The output is a user-accessible traceability report.

[0430] (Application example 2)

[0431] 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."

[0432] It is important to identify the origin of digital content created by generative AI and provide it to users in a transparent manner. However, conventional systems have difficulty in identifying in detail which learning sources a generated product relies on, which is insufficient to help users understand. Furthermore, they lack the ability to provide feedback tailored to the user's emotions, resulting in a suboptimal user experience.

[0433] 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.

[0434] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and break down the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of a product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for the server to analyze the user's emotions using an emotion recognition engine; and a means for the server to adjust the analysis results based on the user's emotions and provide optimal feedback. This not only makes it possible to identify the origin of the product in detail and provide it to the user transparently, but also to provide optimal feedback based on the user's emotions, improving the user experience.

[0435] "Generative AI" is a system that automatically generates content using artificial intelligence technology.

[0436] "Digital content" refers to information that is generated and stored electronically, such as images, videos, and text.

[0437] "Upload" is an operation in which a user sends data from their device to a server.

[0438] A "server" is a central processing unit that has functions such as receiving, storing, analyzing, and providing data.

[0439] "Temporarily storing" refers to storing the received digital content in a server for a certain period of time.

[0440] "Analysis" is the process of examining received digital content to identify its components and characteristics.

[0441] "Decomposing the generation process" refers to breaking down the creation steps and original data of content created by generative AI into their component parts and revealing them.

[0442] A "learning source" is a dataset or information source that the generative AI references when generating content.

[0443] "Dependency" refers to the state in which one element depends on another element for its existence or function.

[0444] "Identifying" refers to the act of clarifying related information and elements based on the analysis results.

[0445] "Providing" means showing or making available to the user the analysis results or other information.

[0446] An "emotion recognition engine" is a system that analyzes a user's facial expressions and voice data to identify their emotions.

[0447] "Feedback" refers to information or advice returned to the user based on the analysis results and evaluation.

[0448] MODE FOR CARRYING OUT THE INVENTION

[0449] The present invention relates to a system that identifies the origin of digital content created by generative AI and provides appropriate feedback to users. The system of the present invention mainly includes the following components:

[0450] Overall system overview

[0451] The system of the present invention is composed of a user terminal, a server, a learning source database, and an emotion recognition engine.

[0452] Component Details

[0453] User terminal

[0454] A user device is a device such as a smartphone or PC that allows users to upload digital content created by generative AI. The user device has the function of sending content to a server through a dedicated application.

[0455] server

[0456] The server is a central system with multiple functions:

[0457] 1. Receiving and storing content:

[0458] It has the function of receiving and temporarily storing digital content uploaded by users.

[0459] 2. Content Analysis:

[0460] It analyzes the content uploaded by users and breaks down the generation process, using an image analysis module for images and a natural language processing module for text.

[0461] 3. Identifying learning sources:

[0462] Based on the results of the analysis, similar data is identified by searching the training source database, which stores the datasets and information used to train the generative AI.

[0463] 4. Emotion recognition and feedback regulation:

[0464] It has the ability to analyze the user's emotions using an emotion recognition engine and adjust the content of the feedback provided based on the results.

[0465] Emotion Recognition Engine

[0466] An emotion recognition engine is a system that analyzes a user's facial expressions and voice to identify their emotions. It uses a webcam or microphone to collect emotional data in real time and recognizes emotions based on that data. Specific emotion recognition technologies used include EmotionRecognizer.

[0467] Specific examples of operations

[0468] Content upload and analysis

[0469] For example, consider the case where a user wants to upload a landscape image created using a generative AI model. The user launches an application on a smartphone or PC, selects a landscape image, and uploads it.

[0470] Obtaining a traceability report

[0471] Uploaded images are sent to a server for temporary storage. The server uses an image analysis module to extract key features from the image and uses that data to search a database of training sources. Once similar data is identified, a traceability report is generated, including the URL of the original image and details of the citation.

[0472] Emotion recognition and feedback provision

[0473] When a user receives a report, an emotion recognition engine analyzes the user's facial expressions and voice. For example, if the user is confused, a more detailed explanation can be added to the report. Conversely, if the user is satisfied, a concise and positive report can be provided. This improves the user experience.

[0474] Examples of prompt statements

[0475] For example, you can generate a landscape image by inputting the following prompt sentence into a generative AI model:

[0476] "Generate the following landscape image. It's a beautiful scene featuring a bright blue sky, lush green mountains, and a flowing river."

[0477] Using this prompt, users can generate digital content based on the specified content, and through a subsequent traceability and feedback process, they can obtain more detailed information and appropriate feedback.

[0478] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0479] Step 1:

[0480] User device uploads generated content

[0481] Users upload digital content (e.g., images or text) created using generative AI models to a server via their device (smartphone or PC), and enter the content's metadata (e.g., content type, creation date, etc.).

[0482] Input: Generated content and metadata

[0483] Output: Sending content to the server

[0484] Specific operation: The user launches the dedicated application, clicks the upload button, selects the content file and metadata, and sends them.

[0485] Step 2:

[0486] The server receives the content and temporarily stores it

[0487] The server receives and temporarily stores digital content uploaded by users, which then adds the content to an analysis queue.

[0488] Input: Generated content and metadata sent from the user device

[0489] Output: Save content to temporary storage area on the server

[0490] Specific operation: The server receives the HTTP request, analyzes the attachment, and stores it in a temporary storage area.

[0491] Step 3:

[0492] The server analyzes the digital content

[0493] The server sequentially retrieves stored digital content from the analysis queue and selects the appropriate analysis module based on the content type (e.g., image, text): for images, it uses the image analysis module, and for text, it uses the natural language processing module.

[0494] Input: Digital content stored in temporary storage

[0495] Output: Extracted features and keywords

[0496] What it does: The server's internal analysis engine loads the content and performs feature extraction processing. For example, edge detection and color histogram analysis are used to extract image features.

[0497] Step 4:

[0498] The server searches the learning source database to identify similar data

[0499] The server searches the learning source database based on the extracted features and keywords to identify similar data, thereby clarifying which learning source the content relies on.

[0500] Input: Extracted features and keywords

[0501] Output: Similar data and related information

[0502] What happens: The server searches the database using SQL queries and search algorithms to identify the most similar data.

[0503] Step 5:

[0504] The server generates a traceability report and provides it to the user.

[0505] The server generates a traceability report based on the identified learning sources, including the URL of the original data and the location of specific parts, and provides the report to users via a web interface or API.

[0506] Input: Similar data and related information

[0507] Output: Traceability report

[0508] What it does: The server runs a report generation script to create a report detailing the identified learning sources, which is then saved in a user-accessible format.

[0509] Step 6:

[0510] The server uses an emotion recognition engine to analyze the user's emotions.

[0511] When a user receives a traceability report, the server uses an emotion recognition engine to analyze the user's face and voice to collect emotional data, using a library called EmotionRecognizer.

[0512] Input: User facial and voice data

[0513] Output: Recognized emotion data

[0514] What it does: The emotion recognition engine collects data in real time from your webcam and microphone, then runs emotion analysis algorithms to identify emotions.

[0515] Step 7:

[0516] The server adjusts the analysis results based on the user's emotions and provides feedback.

[0517] The server adjusts the feedback it provides based on the user's recognized emotional data: providing detailed explanations if the user is confused, and concise feedback if the user is satisfied.

[0518] Input: Recognized emotion data

[0519] Output: Regulated Feedback

[0520] Specific operation: The server dynamically generates feedback content based on emotion data and displays it to the user, allowing the user to receive optimized feedback.

[0521] 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.

[0522] 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.

[0523] 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.

[0524] [Second embodiment]

[0525] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0526] 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.

[0527] 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).

[0528] 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.

[0529] 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.

[0530] 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).

[0531] 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.

[0532] 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.

[0533] 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.

[0534] 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.

[0535] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0536] 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."

[0537] Overall system overview

[0538] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[0539] System configuration

[0540] The system includes the following major components:

[0541] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0542] 2. Server: A central system with the following functions:

[0543] Receiving digital content

[0544] Temporary storage of digital content

[0545] Digital content analysis

[0546] Searching the learning source database

[0547] Identifying the location of use

[0548] Generate and provide traceability reports

[0549] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0550] Program processing

[0551] The program of this system is executed in the following manner.

[0552] Providing and Receiving Products

[0553] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[0554] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0555] Digital content analysis

[0556] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0557] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[0558] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0559] Identifying learning sources

[0560] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[0561] Identifying the location of use

[0562] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0563] Generating and providing traceability information

[0564] The server generates a traceability report based on the identified information, which includes the URL of the learning source and the location of the identified part.

[0565] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[0566] Specific examples

[0567] Example 1: Image creation

[0568] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[0569] The server receives the images and adds them to the analysis queue.

[0570] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[0571] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0572] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[0573] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[0574] Example 2: Text production

[0575] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[0576] The server receives the article and adds it to the analysis queue.

[0577] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0578] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0579] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[0580] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[0581] The above is an embodiment of the present invention, which makes it possible to provide a specific system for clarifying the origin of digital content created using generative AI and minimizing the risk of copyright infringement.

[0582] The processing flow will be explained below.

[0583] Step 1:

[0584] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[0585] Step 2:

[0586] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[0587] Step 3:

[0588] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[0589] Step 4:

[0590] The server analyzes the digital content using the selected analysis module.

[0591] Image Analysis Module: Extracts key features from an image using techniques such as edge detection, color analysis, and shape recognition.

[0592] Text Analysis Module: Performs grammatical analysis, tokenization, keyword extraction, etc. to identify important keywords and phrases.

[0593] Step 5:

[0594] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on keywords.

[0595] Step 6:

[0596] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources. This analysis uses statistical and clustering techniques to perform a specific mapping, identifying which original information a particular image part or text phrase comes from.

[0597] Step 7:

[0598] The server generates a traceability report based on the results, including the URL of the learning source, the location of the citation, and related metadata. The report is generated in a detailed and easy-to-understand format.

[0599] Step 8:

[0600] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Based on the report, the user can clearly understand the origin and usage of the product.

[0601] These are the processing steps of the system, which allow users to check in detail the origin of digital content created by generative AI, reducing the risk of copyright infringement.

[0602] Example 1

[0603] 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."

[0604] Currently, it is difficult to clarify the origin of digital data created by an artificial intelligence model and provide appropriate traceability. Therefore, a system is needed that identifies the learning source from which digital data was generated and provides this information to the user. The present invention aims to solve this problem and provide an efficient and effective system for identifying the origin of generated digital data.

[0605] 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.

[0606] In this invention, the server includes: means for a user to upload digital data created by an AI model; means for the server to receive and temporarily store the digital data; means for the server to analyze the digital data and decompose the generation process; means for the server to identify learning sources based on the digital data; means for the server to perform a detailed analysis of which parts of a product depend on which learning sources based on the identified learning sources; and means for the server to generate a traceability report based on the identified information and provide it to the user. This makes it possible to clarify the origin of digital data created by an AI model and provide an appropriate traceability report to the user.

[0607] "User" means any person or entity that utilizes an artificial intelligence model to generate digital data and uploads that data to the System.

[0608] An "artificial intelligence model" refers to a program or system that uses machine learning or deep learning algorithms to automatically create artifacts from data.

[0609] "Digital data" refers to information that is stored, processed, and transmitted electronically, and this information can exist in the form of images, text, audio, video, etc.

[0610] "Server" refers to a computer system capable of processing, analyzing, storing, and serving digital data received from users.

[0611] "Analysis module" refers to a software component that analyzes digital data according to its type and extracts features and keywords.

[0612] "Learning source" refers to a dataset or information source used in the training process of an artificial intelligence model, providing the information that forms the basis of the generated digital data.

[0613] "Learning source database" refers to a database that stores the underlying data sets and information sources that identify the origin of generated digital data.

[0614] A "traceability report" is a report that clearly shows the relationship between the analyzed digital data and its learning source, including information indicating which original data a particular part is based on.

[0615] Overall system overview

[0616] The system of the present invention aims to clarify the origin of digital data created by a generative AI model and provide appropriate traceability. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[0617] System configuration

[0618] The system includes the following major components:

[0619] 1. User terminal: A device (PC, smartphone, etc.) through which a user uploads digital data created by a generative AI model.

[0620] 2. Server: A central system with the following functions:

[0621] Receiving digital data

[0622] Temporary storage of digital data

[0623] Digital Data Analysis

[0624] Searching the learning source database

[0625] Identifying the location of use

[0626] Generate and provide traceability reports

[0627] 3. Training source database: A database that stores the datasets and information sources used to train a generative AI model.

[0628] Program processing

[0629] The program of this system is processed as follows.

[0630] Providing and Receiving Products

[0631] Users upload digital data created by generative AI models to the system using a dedicated web interface or API, along with metadata such as content type and creation date and time.

[0632] The server receives the uploaded digital data, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0633] Digital Data Analysis

[0634] The server sequentially retrieves the digital data in the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0635] Image analysis module: Uses OpenCV etc. to extract key image features (edges, hue, shape, etc.).

[0636] Text analysis module: Uses NLTK etc. to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[0637] Identifying learning sources

[0638] The server searches the training source database based on the extracted features and keywords, and uses SQL or Elasticsearch to query the database and identify data with high similarity.

[0639] Identifying usage locations and generating traceability information

[0640] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources, and determines the correspondence between features and keywords and learning sources using statistical and clustering techniques.

[0641] The server generates a traceability report based on the analysis results, including the correspondence between specific parts of the artifact and the learning source.

[0642] Providing traceability reports

[0643] The server provides the generated traceability reports to the user, which can be downloaded via a web interface or API, in PDF or HTML format.

[0644] Specific examples

[0645] Example 1: Image creation

[0646] Users upload vibrant landscape paintings created by generative AI models to the system.

[0647] The server receives the images and adds them to the analysis queue.

[0648] The server selects an image analysis module to extract the main features of the image (edges, color histogram).

[0649] The server searches the training source database based on the extracted features to identify datasets of similar landscape images.

[0650] The server analyzes which original image a particular part of the landscape painting comes from and performs a specific mapping.

[0651] The server generates a traceability report and provides it to the user, including information such as "The mountain part is from original image X, and the river part is from original image Y."

[0652] Example 2: Text production

[0653] Users upload articles about the latest technology trends to the system, which are created using a generative AI model.

[0654] The server receives the article and adds it to the analysis queue.

[0655] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0656] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0657] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping.

[0658] The server generates a traceability report and provides it to the user, which indicates which original article a particular technical term or explanation is based on.

[0659] In this way, the system of the present invention can clarify the origin of digital data created using a generative AI model and provide users with an appropriate traceability report.

[0660] Examples of prompt statements

[0661] For images: "Please identify the dataset or training source from which this image originated."

[0662] For text: "Please identify the source or quoted passage from which this article was based."

[0663] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0664] Step 1:

[0665] A user uploads digital data (such as images or text) created by a generative AI model to the system using a dedicated web interface or API. The input is the digital data and its metadata (content type, creation date and time), and the output is an upload request sent to the server. Specifically, the user opens a browser or a dedicated application, selects a file, enters metadata, and clicks the upload button.

[0666] Step 2:

[0667] The server receives an upload request from a user and temporarily stores the digital data and metadata. The input is the digital data and metadata sent by the user, and the output is the data stored in the server's temporary storage database. Specifically, the server saves the uploaded data in a specified directory (such as / tmp / uploads) and simultaneously records the metadata in the database.

[0668] Step 3:

[0669] The server adds the information of the stored digital data to the analysis queue. The input is the path and metadata of the temporarily stored digital data, and the output is the task added to the analysis queue. Specifically, the server adds the file path and metadata to a queue system (e.g., RabbitMQ or Kafka).

[0670] Step 4:

[0671] The server retrieves digital data from the analysis queue and selects an appropriate analysis module based on its type (image, text, etc.). The input is the data and metadata retrieved from the analysis queue, and the output is the selection of an analysis module. Specifically, the server checks the metadata and selects an image analysis module if the analysis target is an image, or a text analysis module if the analysis target is text.

[0672] Step 5:

[0673] The server analyzes the digital data using the selected analysis module to extract features and keywords. The input is the digital data entered into the analysis module, and the output is the extracted features and keywords. Specifically, the image analysis module uses OpenCV to extract edges and color histograms, and the text analysis module uses NLTK to perform grammar analysis and tokenization.

[0674] Step 6:

[0675] The server searches the training source database based on the extracted features and keywords to identify similar data. The input is the extracted features and keywords, and the output is information about similar training source data. Specifically, the server uses SQL or Elasticsearch to search the database and identify data with high similarity scores.

[0676] Step 7:

[0677] The server analyzes the search results in detail and identifies which parts of the generated digital data depend on which learning sources. The input is the search results, and the output is information showing the correspondence between the generated data and the learning sources. Specifically, statistical methods and clustering techniques are used to map specific parts to the learning sources.

[0678] Step 8:

[0679] The server generates a traceability report based on the identified information. The input is information showing the correspondence between the product and the learning source, and the output is a traceability report. Specifically, the report includes the URL of the learning source and the location information of the identified part, and is generated in PDF or HTML format.

[0680] Step 9:

[0681] The server provides the generated traceability report to the user. The input is the traceability report, and the output is the report in a format that the user can download. Specifically, the user logs into the web interface and clicks on a link to view and download the report.

[0682] (Application example 1)

[0683] 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."

[0684] The origin of digital content generated by generative AI models is often unclear, potentially raising issues regarding the reliability and copyright of the content. It is also difficult to track which learning source user-generated content was based on. Furthermore, there is no established method for utilizing this information in real time. To solve these problems, a system is needed that can clarify the origin of digital content and analyze its generation process.

[0685] 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.

[0686] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and decompose the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of the product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for a smartphone or robot to analyze the digital content in real time; and a means for visually displaying the information identified by the server. This makes the origin of the digital content clear, enabling the use of highly reliable digital content. Furthermore, real-time origin verification is also possible, allowing users to quickly verify the reliability of the content.

[0687] "Digital Content" refers to information that is generated, stored, and transmitted by electronic means.

[0688] "Generative AI" refers to systems or technologies that use artificial intelligence to generate new digital content.

[0689] "Uploading" refers to the act of a user sending digital content from their own terminal to a server.

[0690] "Server" refers to a computer system that receives, stores, processes, and serves digital data.

[0691] "Analysis" refers to the technical process of breaking down the content of digital content and clarifying its characteristics and the process by which it was created.

[0692] "Learning source" refers to the dataset or information source that generative AI uses to generate new content.

[0693] "Real-time" refers to processing and results being provided immediately.

[0694] "Visual display" refers to displaying analysis results and traceability information on a screen or display in a format that is easy for users to understand.

[0695] "Feature extraction" refers to the technique of identifying and extracting important attributes and patterns from digital content.

[0696] "Traceability" refers to the ability to track and record the origin of a product or the provenance of its components.

[0697] Overall system overview

[0698] This invention is a system that provides traceability functions to clarify the origin of digital content created by generative AI. Users upload the creations, and the server analyzes them to identify the learning source and usage points, and provides that information to the user. This system can be applied to security services, particularly applications installed on smartphones or robots that perform real-time analysis.

[0699] System configuration

[0700] 1. User Device

[0701] A user terminal is a device that users use to upload digital content created by generative AI. This includes smartphones, tablets, and PCs.

[0702] 2. Server

[0703] The server is the central component with the following functions:

[0704] Receiving and temporarily storing digital content: Receives and temporarily stores digital content uploaded by users.

[0705] Digital content analysis: Analyze digital content and break down its characteristics and creation process.

[0706] Identifying learning sources: Based on the analysis results, the learning sources are searched from the database.

[0707] Identifying the parts of the product that are based on learning sources and identifying which learning sources they depend on.

[0708] Providing traceability information: The identified information is provided to the user as a traceability report.

[0709] 3. Learning Source Database

[0710] The training source database is a database that stores the datasets and information sources used to train the generative AI. This database contains a large amount of datasets (image data, text data).

[0711] 4. Smartphone Robot

[0712] This system can be applied as an application installed on smartphones or robots. These devices have the ability to analyze digital content and provide traceability information in real time.

[0713] Program processing overview

[0714] Uploading Digital Content

[0715] Users upload digital content using a dedicated interface (for example, a web browser or a dedicated application). When uploading, metadata such as the content type and creation date and time are also added.

[0716] Receiving and temporarily storing digital content

[0717] The server receives and temporarily stores the digital content uploaded by the user, and once the storage is complete, adds the digital content to an analysis queue.

[0718] Digital content analysis

[0719] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.). For example, for images, it identifies edges and color histograms, while for text, it performs grammar analysis and keyword extraction.

[0720] Identifying learning sources

[0721] The server uses the analysis results to search the training source database to identify similar datasets, using feature matching and clustering techniques.

[0722] Identifying the location of use

[0723] The server performs detailed analysis of specific parts of the production based on the learning source and clarifies which learning source it depends on, thereby generating highly reliable traceability information.

[0724] Providing traceability information

[0725] The server generates a traceability report based on the identified information, which includes the URL of the learning source and details of the specific part, and can be viewed and checked by the user.

[0726] Examples of concrete examples and prompts

[0727] For example, to verify the originality of a photo taken with a smartphone, the user uploads the photo, the server analyzes the photo's characteristics, and searches the learning source database. A traceability report is generated based on the search results and provided to the user.

[0728] Example prompt sentence:

[0729] Task: Trace provenance

[0730] Content Type: Image

[0731] File path: 'path / to / image.jpg'

[0732] This system allows users to quickly verify the authenticity of digital content and obtain traceability information in real time.

[0733] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0734] Step 1: Upload your digital content

[0735] Users upload digital content using their smartphones or computers. They use a dedicated web interface or application to send the digital content file and its metadata (content type, creation date, etc.). The input is the digital content file selected by the user, and the output is the data received by the server. The specific action is when the user clicks the "Upload" button.

[0736] Step 2: Receive and temporarily store digital content

[0737] The server receives digital content sent by the user and temporarily stores it in storage. The input is the digital content file sent by the user, and the output is a file stored in the server's temporary storage area. Specifically, the server analyzes the HTTP request and writes the file data to disk.

[0738] Step 3: Analyzing the digital content

[0739] The server retrieves the temporarily stored digital content and selects an analysis module depending on the type. For example, for image files, it performs edge detection and color histogram analysis. For text files, it performs grammatical analysis and keyword extraction. The input is the temporarily stored digital content file, and the output is the analyzed feature data. Specifically, the server loads the analysis module and analyzes the data.

[0740] Step 4: Identify learning sources

[0741] The server searches the training source database based on the analysis results. It uses the extracted feature data as input to identify similar datasets. The input is the analyzed feature data, and the output is similar datasets identified from the training source database. Specifically, the server inputs the feature values ​​into a clustering algorithm and calculates the similarity.

[0742] Step 5: Identify the usage

[0743] The server determines which learning source a particular part of the digital content depends on based on information obtained from the learning source database. The input is information on similar datasets obtained from the learning source database, and the output is mapping information between each part of the product and its corresponding learning source. Specifically, the server uses statistical methods to match each part of the digital content with the learning source.

[0744] Step 6: Generate traceability information

[0745] The server generates a traceability report based on the identified usage location information. The report includes the URL of the learning source and detailed information about the specific part. The input is the mapping information, and the output is the traceability report provided to the user. Specifically, the server creates a report format based on the mapping information and fills in the data.

[0746] Step 7: Provide traceability information

[0747] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The input is the traceability report, and the output is a visualized report that the user can access. Specifically, the server returns the report upon the user's request.

[0748] As described above, this system is composed of multiple steps to clarify the origin of digital content and guarantee its reliability.

[0749] 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.

[0750] Overall system overview

[0751] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI, and an emotion engine that recognizes user emotions and optimizes analysis results. In this system, a user provides a product, a server analyzes the product to identify learning sources and usage points, adjusts the analysis results using the emotion engine, and provides that information to the user.

[0752] System configuration

[0753] The system includes the following major components:

[0754] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0755] 2. Server: A central system with the following functions:

[0756] Receiving digital content

[0757] Temporary storage of digital content

[0758] Digital content analysis

[0759] Searching the learning source database

[0760] Identifying the location of use

[0761] Generate and provide traceability reports

[0762] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0763] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[0764] Program processing

[0765] The processing of the system programs will be explained in detail below.

[0766] Providing and Receiving Products

[0767] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[0768] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[0769] Digital content analysis

[0770] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on the type (image, text, etc.). For example, it launches an image analysis module for images and a natural language processing module for text.

[0771] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[0772] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0773] Identifying learning sources and where to use them

[0774] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[0775] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0776] Use of emotion engine

[0777] The server activates an emotion engine to recognize the user's emotion.

[0778] The emotion engine analyzes the user's facial expressions and voice to identify emotions, for example, by using a webcam or microphone to collect real-time emotion data.

[0779] The emotion engine adjusts the analysis results based on the identified emotion, for example providing more detailed traceability information if the user is confused, or a concise report if the user is happy.

[0780] Generating and providing traceability information

[0781] The server combines the identified information with the results of the sentiment engine to generate a traceability report, which includes the URL of the learning source, the location of the citation, and associated metadata.

[0782] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[0783] Specific examples

[0784] Example 1: Image creation

[0785] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[0786] The server receives the images and adds them to the analysis queue.

[0787] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[0788] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0789] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[0790] The server activates the emotion engine, which analyzes the user's facial expressions and voice to identify their emotion. For example, if the user is confused, a more detailed explanation is added to the report.

[0791] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[0792] Example 2: Text production

[0793] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[0794] The server receives the article and adds it to the analysis queue.

[0795] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0796] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0797] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[0798] The server starts the emotion engine, which analyzes the user's emotions. For example, if the user has a positive feeling, it generates a concise and positive report.

[0799] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[0800] The above is an embodiment of the present invention. The present invention makes it possible to clarify the origin of digital content created using generative AI and provide an optimal report that corresponds to the user's emotions.

[0801] The processing flow will be explained below.

[0802] Step 1:

[0803] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[0804] Step 2:

[0805] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[0806] Step 3:

[0807] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[0808] Step 4:

[0809] The server analyzes the digital content using the selected analysis module.

[0810] Image Analysis Module: Uses feature extraction algorithms to identify key features of an image (edges, color, shape, etc.).

[0811] Text analysis module: Using natural language processing technology, it analyzes the structure of text (grammar analysis, tokenization) and extracts important keywords and phrases.

[0812] Step 5:

[0813] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on the extracted keywords.

[0814] Step 6:

[0815] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[0816] Step 7:

[0817] The server activates an emotion engine to recognize the user's emotion.

[0818] The emotion engine collects facial and voice data from the user and identifies their emotions based on this data. The data is collected in real time using a webcam and microphone.

[0819] An emotion engine tailors the presentation of the analysis results based on the identified emotion.

[0820] Step 8:

[0821] The server uses the emotion engine to generate a traceability report based on the refined analysis, including the URL of the learning source, the location of the citation, and related metadata.

[0822] Step 9:

[0823] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Detailed or concise reports are provided based on emotions to help users understand.

[0824] Example: Image production

[0825] 1. A user uploads an image of a landscape painting to the system.

[0826] 2. The server receives the image and adds it to the analysis queue.

[0827] 3. The server selects an image analysis module and extracts key features.

[0828] 4. The server searches the learning source database based on the extracted features to identify the original image.

[0829] 5. The server maps specific parts of the image to the original image.

[0830] 6. The server launches the emotion engine, which analyzes the user's facial expressions and voice in real time to identify their emotions.

[0831] 7. The server generates a traceability report based on the identified emotions in a format that is understandable to the user.

[0832] 8. The server provides the report to the user, who can view it via a web interface.

[0833] Example: For text productions

[0834] 1. A user uploads a technical article to the system.

[0835] 2. The server receives the article and adds it to the analysis queue.

[0836] 3. The server selects a text analysis module and performs a structural analysis of the text.

[0837] 4. The server searches the learning source database based on the extracted keywords to identify the original article.

[0838] 5. The server maps specific parts of the technical article to the original article.

[0839] 6. The server launches the emotion engine and analyzes the user's emotions in real time.

[0840] 7. The server generates a traceability report in an appropriate format based on the identified emotion.

[0841] 8. The server provides the report to the user, who can view it via a web interface.

[0842] This will clarify the provenance of digital content using generative AI and provide optimal reporting based on user sentiment.

[0843] Example 2

[0844] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0845] The origin of digital content created by generative AI is unclear, making it impossible to identify which learning sources it relies on. Furthermore, there is a problem in that traceability information is difficult to understand because appropriate information is not provided according to user sentiment.

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

[0847] In this invention, the server includes a means for receiving and temporarily storing digital content created by the AI ​​by a user, a means for analyzing the digital content and breaking down the creation process, and a means for identifying learning sources based on the digital content. This makes it possible to clarify the origin of the digital content created by the AI ​​and provide appropriate information according to the user's emotions.

[0848] "User" means an individual or organization that uses the system to upload digital content created by generative AI.

[0849] "Generative AI" is a system that automatically generates digital content (images, text, etc.) using artificial intelligence technology.

[0850] "Digital content" refers to electronic data created by generative AI, including images, text, and audio.

[0851] "Uploading" is the act of a user sending digital content created by generative AI to a server.

[0852] "Server" means a centralized computer system that receives, analyzes, stores, and serves digital content.

[0853] A "temporary storage means" is a storage device or storage service that allows a server to temporarily store digital content.

[0854] "Means for analyzing" refers to software or hardware that the server uses to analyze digital content and extract its structure and characteristics.

[0855] "Means for decomposing the creation process" refers to analytical techniques that allow the server to clarify the elements and methods used to create digital content.

[0856] A "learning source" is a dataset or information source that a generative AI uses for learning or training.

[0857] "Means for identifying" refers to the technology or algorithm that the server uses to identify which learning source a piece of digital content relies on.

[0858] The "means of provision" refers to the interface or API that the server uses to communicate and display analysis results and traceability information to users.

[0859] The "means for analyzing emotions" refers to machine learning models and emotion recognition technologies that allow the server to analyze the user's facial expressions and voice and identify the user's emotions.

[0860] "Means for adjusting information" refers to technologies and algorithms for optimizing traceability information by taking into account the results of user sentiment analysis.

[0861] Overall system overview

[0862] This invention relates to a system that clarifies the origin of digital content created by generative AI, recognizes user emotions, and optimizes analysis results. By using this system, it is possible to clarify which dataset the generative AI used to generate the content, and provide information that meets the user's needs.

[0863] System configuration

[0864] The system includes the following major components:

[0865] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[0866] 2. Server: A central system with the following functions:

[0867] Receiving digital content

[0868] Temporary storage of digital content

[0869] Digital content analysis

[0870] Searching the learning source database

[0871] Identifying the location of use

[0872] Generate and provide traceability reports

[0873] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[0874] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[0875] Details of data processing and data calculation

[0876] The specific operation of the system will be described below.

[0877] Providing and Receiving Products

[0878] Users upload digital content created by generative AI through a dedicated web interface or API, and provide simple metadata (content type, creation date, etc.) at the time.

[0879] The server receives the uploaded digital content and temporarily stores it in a cloud storage location (e.g., Amazon S3). Once the storage is complete, the digital content is added to the analysis queue.

[0880] Digital content analysis

[0881] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[0882] Image Analysis Module: Performs image feature extraction using the OpenCV library to identify key features (edges, hue, shape, etc.).

[0883] Text analysis module: Uses natural language processing libraries such as NLTK and spaCy to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[0884] Identifying learning sources and where to use them

[0885] The server searches the training source database based on the extracted features and keywords, for example, by using SQL queries or Elasticsearch to search the data.

[0886] The server then uses the search results to perform a detailed analysis of which learning sources specific parts of the artifacts depend on, applying statistical methods and machine learning algorithms such as K-means clustering and DBSCAN.

[0887] Use of emotion engine

[0888] The server runs an emotion engine, which analyzes the user's facial expressions and voice to identify emotions. It uses libraries such as OpenVINO and Dlib to identify emotions such as smiling, confused, or angry using data from the webcam and microphone.

[0889] The emotion engine adjusts the analysis results based on the identified emotion, for example adding more detailed information to the traceability report if the user is confused.

[0890] Generating and providing traceability information

[0891] The server combines the identified information (the URL of the learning source, the location of the quoted section, and related metadata) with the results of the user sentiment analysis and generates a traceability report using tools such as JasperReports and Crystal Reports.

[0892] The server provides the generated traceability report to the user, who can view and download it via the system's web interface or API.

[0893] Specific examples

[0894] Example 1: Image creation

[0895] Users upload vivid landscape images created by generative AI through the system.

[0896] The server receives the images and adds them to the analysis queue.

[0897] The server selects an image analysis module to extract key features such as edges and color histograms.

[0898] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[0899] The server activates an emotion engine, which analyzes the user's facial expressions and voice to identify their emotions. For example, if the user is confused, more detailed information is added to the traceability report.

[0900] The server generates a traceability report, providing the user with details about the original source of the image.

[0901] Example 2: Text production

[0902] Users upload AI-generated articles about the latest technology trends through the system.

[0903] The server receives the article and adds it to the analysis queue.

[0904] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[0905] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[0906] The server will launch an emotion engine, which will analyze the user's emotion, for example, if the user is confused, it will add more detailed information to the traceability report.

[0907] The server generates a traceability report, providing the user with details about the original source of the technical article.

[0908] The above is a specific embodiment of the present invention. This system makes it possible to clarify the origin of digital content created by generative AI and provide optimal traceability information according to the user's feelings.

[0909] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0910] Step 1: Providing and Receiving Products

[0911] Users upload digital content (e.g., landscape images or technical articles) created by generative AI using the system's web interface or API. As input, the digital content and its metadata (content type, creation date, etc.) are provided.

[0912] The server receives the uploaded digital content, and the output is that the digital content and metadata are stored on the server.

[0913] The server temporarily stores the content in cloud storage (e.g., Amazon S3) and adds it to the analysis queue, so the content is ready for analysis.

[0914] Step 2: Analyzing the digital content

[0915] The server sequentially retrieves digital content from the analysis queue, and the input is the digital content added to the analysis queue.

[0916] The server selects an analysis module based on the type of digital content (image, text, etc.), for example, an image analysis module for images and a text analysis module for text.

[0917] The server launches the selected modules to analyze the digital content. Specifically, the image analysis module uses the OpenCV library to analyze edges, hues, shapes, etc., while the text analysis module uses NLTK and spaCy to perform grammar analysis and keyword extraction. The output is the extracted features and keywords.

[0918] Step 3: Identify learning sources and where to use them

[0919] The server uses the features and keywords obtained from the analysis results to search the learning source database. The input is the features and keywords from the analysis results.

[0920] The server uses SQL queries or Elasticsearch to identify relevant datasets from the training source database, and the output is the relevant training source data.

[0921] The server uses statistical methods and machine learning algorithms such as K-means clustering and DBSCAN to perform a detailed analysis of which learning sources specific parts of the product depend on, thereby obtaining mapping information between the product and the learning sources.

[0922] Step 4: Use the Emotion Engine

[0923] The server starts the emotion engine, and inputs are the user's facial expressions and voice data.

[0924] The emotion engine uses libraries such as OpenVINO and Dlib to analyze the user's facial expressions and voice in real time to identify emotions. The output is the user's emotional information.

[0925] The server adjusts the analysis results based on the emotion engine's results, for example adding more detailed information to the traceability report if the user is confused.

[0926] Step 5: Generate and provide traceability information

[0927] The server integrates the identified information (the URL of the learning source, the location of the quoted part, and related metadata) with the sentiment analysis results to generate a traceability report. The input is the learning source information and the sentiment analysis results.

[0928] The server generates the traceability report using tools such as JasperReports or Crystal Reports. The output is a comprehensive traceability report.

[0929] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The output is a user-accessible traceability report.

[0930] (Application example 2)

[0931] 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."

[0932] It is important to identify the origin of digital content created by generative AI and provide it to users in a transparent manner. However, conventional systems have difficulty in identifying in detail which learning sources a generated product relies on, which is insufficient to help users understand. Furthermore, they lack the ability to provide feedback tailored to the user's emotions, resulting in a suboptimal user experience.

[0933] 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.

[0934] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and break down the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of a product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for the server to analyze the user's emotions using an emotion recognition engine; and a means for the server to adjust the analysis results based on the user's emotions and provide optimal feedback. This not only makes it possible to identify the origin of the product in detail and provide it to the user transparently, but also to provide optimal feedback based on the user's emotions, improving the user experience.

[0935] "Generative AI" is a system that automatically generates content using artificial intelligence technology.

[0936] "Digital content" refers to information that is generated and stored electronically, such as images, videos, and text.

[0937] "Upload" is an operation in which a user sends data from their device to a server.

[0938] A "server" is a central processing unit that has functions such as receiving, storing, analyzing, and providing data.

[0939] "Temporarily storing" refers to storing the received digital content in a server for a certain period of time.

[0940] "Analysis" is the process of examining received digital content to identify its components and characteristics.

[0941] "Decomposing the generation process" refers to breaking down the creation steps and original data of content created by generative AI into their component parts and revealing them.

[0942] A "learning source" is a dataset or information source that the generative AI references when generating content.

[0943] "Dependency" refers to the state in which one element depends on another element for its existence or function.

[0944] "Identifying" refers to the act of clarifying related information and elements based on the analysis results.

[0945] "Providing" means showing or making available to the user the analysis results or other information.

[0946] An "emotion recognition engine" is a system that analyzes a user's facial expressions and voice data to identify their emotions.

[0947] "Feedback" refers to information or advice returned to the user based on the analysis results and evaluation.

[0948] MODE FOR CARRYING OUT THE INVENTION

[0949] The present invention relates to a system that identifies the origin of digital content created by generative AI and provides appropriate feedback to users. The system of the present invention mainly includes the following components:

[0950] Overall system overview

[0951] The system of the present invention is composed of a user terminal, a server, a learning source database, and an emotion recognition engine.

[0952] Component Details

[0953] User terminal

[0954] A user device is a device such as a smartphone or PC that allows users to upload digital content created by generative AI. The user device has the function of sending content to a server through a dedicated application.

[0955] server

[0956] The server is a central system with multiple functions:

[0957] 1. Receiving and storing content:

[0958] It has the function of receiving and temporarily storing digital content uploaded by users.

[0959] 2. Content Analysis:

[0960] It analyzes the content uploaded by users and breaks down the generation process, using an image analysis module for images and a natural language processing module for text.

[0961] 3. Identifying learning sources:

[0962] Based on the results of the analysis, similar data is identified by searching the training source database, which stores the datasets and information used to train the generative AI.

[0963] 4. Emotion recognition and feedback regulation:

[0964] It has the ability to analyze the user's emotions using an emotion recognition engine and adjust the content of the feedback provided based on the results.

[0965] Emotion Recognition Engine

[0966] An emotion recognition engine is a system that analyzes a user's facial expressions and voice to identify their emotions. It uses a webcam or microphone to collect emotional data in real time and recognizes emotions based on that data. Specific emotion recognition technologies used include EmotionRecognizer.

[0967] Specific examples of operations

[0968] Content upload and analysis

[0969] For example, consider the case where a user wants to upload a landscape image created using a generative AI model. The user launches an application on a smartphone or PC, selects a landscape image, and uploads it.

[0970] Obtaining a traceability report

[0971] Uploaded images are sent to a server for temporary storage. The server uses an image analysis module to extract key features from the image and uses that data to search a database of training sources. Once similar data is identified, a traceability report is generated, including the URL of the original image and details of the citation.

[0972] Emotion recognition and feedback provision

[0973] When a user receives a report, an emotion recognition engine analyzes the user's facial expressions and voice. For example, if the user is confused, a more detailed explanation can be added to the report. Conversely, if the user is satisfied, a concise and positive report can be provided. This improves the user experience.

[0974] Examples of prompt statements

[0975] For example, you can generate a landscape image by inputting the following prompt sentence into a generative AI model:

[0976] "Generate the following landscape image. It's a beautiful scene featuring a bright blue sky, lush green mountains, and a flowing river."

[0977] Using this prompt, users can generate digital content based on the specified content, and through a subsequent traceability and feedback process, they can obtain more detailed information and appropriate feedback.

[0978] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0979] Step 1:

[0980] User device uploads generated content

[0981] Users upload digital content (e.g., images or text) created using generative AI models to a server via their device (smartphone or PC), and enter the content's metadata (e.g., content type, creation date, etc.).

[0982] Input: Generated content and metadata

[0983] Output: Sending content to the server

[0984] Specific operation: The user launches the dedicated application, clicks the upload button, selects the content file and metadata, and sends them.

[0985] Step 2:

[0986] The server receives the content and temporarily stores it

[0987] The server receives and temporarily stores digital content uploaded by users, which then adds the content to an analysis queue.

[0988] Input: Generated content and metadata sent from the user device

[0989] Output: Save content to temporary storage area on the server

[0990] Specific operation: The server receives the HTTP request, analyzes the attachment, and stores it in a temporary storage area.

[0991] Step 3:

[0992] The server analyzes the digital content

[0993] The server sequentially retrieves stored digital content from the analysis queue and selects the appropriate analysis module based on the content type (e.g., image, text): for images, it uses the image analysis module, and for text, it uses the natural language processing module.

[0994] Input: Digital content stored in temporary storage

[0995] Output: Extracted features and keywords

[0996] What it does: The server's internal analysis engine loads the content and performs feature extraction processing. For example, edge detection and color histogram analysis are used to extract image features.

[0997] Step 4:

[0998] The server searches the learning source database to identify similar data

[0999] The server searches the learning source database based on the extracted features and keywords to identify similar data, thereby clarifying which learning source the content relies on.

[1000] Input: Extracted features and keywords

[1001] Output: Similar data and related information

[1002] What happens: The server searches the database using SQL queries and search algorithms to identify the most similar data.

[1003] Step 5:

[1004] The server generates a traceability report and provides it to the user.

[1005] The server generates a traceability report based on the identified learning sources, including the URL of the original data and the location of specific parts, and provides the report to users via a web interface or API.

[1006] Input: Similar data and related information

[1007] Output: Traceability report

[1008] What it does: The server runs a report generation script to create a report detailing the identified learning sources, which is then saved in a user-accessible format.

[1009] Step 6:

[1010] The server uses an emotion recognition engine to analyze the user's emotions.

[1011] When a user receives a traceability report, the server uses an emotion recognition engine to analyze the user's face and voice to collect emotional data, using a library called EmotionRecognizer.

[1012] Input: User facial and voice data

[1013] Output: Recognized emotion data

[1014] What it does: The emotion recognition engine collects data in real time from your webcam and microphone, then runs emotion analysis algorithms to identify emotions.

[1015] Step 7:

[1016] The server adjusts the analysis results based on the user's emotions and provides feedback.

[1017] The server adjusts the feedback it provides based on the user's recognized emotional data: providing detailed explanations if the user is confused, and concise feedback if the user is satisfied.

[1018] Input: Recognized emotion data

[1019] Output: Regulated Feedback

[1020] Specific operation: The server dynamically generates feedback content based on emotion data and displays it to the user, allowing the user to receive optimized feedback.

[1021] 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.

[1022] 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.

[1023] 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.

[1024] [Third embodiment]

[1025] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1026] 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.

[1027] 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).

[1028] 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.

[1029] 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.

[1030] 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).

[1031] 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.

[1032] 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.

[1033] 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.

[1034] 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.

[1035] 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.

[1036] 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."

[1037] Overall system overview

[1038] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[1039] System configuration

[1040] The system includes the following major components:

[1041] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1042] 2. Server: A central system with the following functions:

[1043] Receiving digital content

[1044] Temporary storage of digital content

[1045] Digital content analysis

[1046] Searching the learning source database

[1047] Identifying the location of use

[1048] Generate and provide traceability reports

[1049] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1050] Program processing

[1051] The program of this system is executed in the following manner.

[1052] Providing and Receiving Products

[1053] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[1054] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1055] Digital content analysis

[1056] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1057] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[1058] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1059] Identifying learning sources

[1060] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[1061] Identifying the location of use

[1062] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1063] Generating and providing traceability information

[1064] The server generates a traceability report based on the identified information, which includes the URL of the learning source and the location of the identified part.

[1065] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[1066] Specific examples

[1067] Example 1: Image creation

[1068] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[1069] The server receives the images and adds them to the analysis queue.

[1070] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[1071] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1072] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[1073] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[1074] Example 2: Text production

[1075] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[1076] The server receives the article and adds it to the analysis queue.

[1077] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1078] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1079] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[1080] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[1081] The above is an embodiment of the present invention, which makes it possible to provide a specific system for clarifying the origin of digital content created using generative AI and minimizing the risk of copyright infringement.

[1082] The processing flow will be explained below.

[1083] Step 1:

[1084] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[1085] Step 2:

[1086] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[1087] Step 3:

[1088] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[1089] Step 4:

[1090] The server analyzes the digital content using the selected analysis module.

[1091] Image Analysis Module: Extracts key features from an image using techniques such as edge detection, color analysis, and shape recognition.

[1092] Text Analysis Module: Performs grammatical analysis, tokenization, keyword extraction, etc. to identify important keywords and phrases.

[1093] Step 5:

[1094] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on keywords.

[1095] Step 6:

[1096] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources. This analysis uses statistical and clustering techniques to perform a specific mapping, identifying which original information a particular image part or text phrase comes from.

[1097] Step 7:

[1098] The server generates a traceability report based on the results, including the URL of the learning source, the location of the citation, and related metadata. The report is generated in a detailed and easy-to-understand format.

[1099] Step 8:

[1100] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Based on the report, the user can clearly understand the origin and usage of the product.

[1101] These are the processing steps of the system, which allow users to check in detail the origin of digital content created by generative AI, reducing the risk of copyright infringement.

[1102] Example 1

[1103] 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."

[1104] Currently, it is difficult to clarify the origin of digital data created by an artificial intelligence model and provide appropriate traceability. Therefore, a system is needed that identifies the learning source from which digital data was generated and provides this information to the user. The present invention aims to solve this problem and provide an efficient and effective system for identifying the origin of generated digital data.

[1105] 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.

[1106] In this invention, the server includes: means for a user to upload digital data created by an AI model; means for the server to receive and temporarily store the digital data; means for the server to analyze the digital data and decompose the generation process; means for the server to identify learning sources based on the digital data; means for the server to perform a detailed analysis of which parts of a product depend on which learning sources based on the identified learning sources; and means for the server to generate a traceability report based on the identified information and provide it to the user. This makes it possible to clarify the origin of digital data created by an AI model and provide an appropriate traceability report to the user.

[1107] "User" means any person or entity that utilizes an artificial intelligence model to generate digital data and uploads that data to the System.

[1108] An "artificial intelligence model" refers to a program or system that uses machine learning or deep learning algorithms to automatically create artifacts from data.

[1109] "Digital data" refers to information that is stored, processed, and transmitted electronically, and this information can exist in the form of images, text, audio, video, etc.

[1110] "Server" refers to a computer system capable of processing, analyzing, storing, and serving digital data received from users.

[1111] "Analysis module" refers to a software component that analyzes digital data according to its type and extracts features and keywords.

[1112] "Learning source" refers to a dataset or information source used in the training process of an artificial intelligence model, providing the information that forms the basis of the generated digital data.

[1113] "Learning source database" refers to a database that stores the underlying data sets and information sources that identify the origin of generated digital data.

[1114] A "traceability report" is a report that clearly shows the relationship between the analyzed digital data and its learning source, including information indicating which original data a particular part is based on.

[1115] Overall system overview

[1116] The system of the present invention aims to clarify the origin of digital data created by a generative AI model and provide appropriate traceability. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[1117] System configuration

[1118] The system includes the following major components:

[1119] 1. User terminal: A device (PC, smartphone, etc.) through which a user uploads digital data created by a generative AI model.

[1120] 2. Server: A central system with the following functions:

[1121] Receiving digital data

[1122] Temporary storage of digital data

[1123] Digital Data Analysis

[1124] Searching the learning source database

[1125] Identifying the location of use

[1126] Generate and provide traceability reports

[1127] 3. Training source database: A database that stores the datasets and information sources used to train a generative AI model.

[1128] Program processing

[1129] The program of this system is processed as follows.

[1130] Providing and Receiving Products

[1131] Users upload digital data created by generative AI models to the system using a dedicated web interface or API, along with metadata such as content type and creation date and time.

[1132] The server receives the uploaded digital data, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1133] Digital Data Analysis

[1134] The server sequentially retrieves the digital data in the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1135] Image analysis module: Uses OpenCV etc. to extract key image features (edges, hue, shape, etc.).

[1136] Text analysis module: Uses NLTK etc. to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[1137] Identifying learning sources

[1138] The server searches the training source database based on the extracted features and keywords, and uses SQL or Elasticsearch to query the database and identify data with high similarity.

[1139] Identifying usage locations and generating traceability information

[1140] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources, and determines the correspondence between features and keywords and learning sources using statistical and clustering techniques.

[1141] The server generates a traceability report based on the analysis results, including the correspondence between specific parts of the artifact and the learning source.

[1142] Providing traceability reports

[1143] The server provides the generated traceability reports to the user, which can be downloaded via a web interface or API, in PDF or HTML format.

[1144] Specific examples

[1145] Example 1: Image creation

[1146] Users upload vibrant landscape paintings created by generative AI models to the system.

[1147] The server receives the images and adds them to the analysis queue.

[1148] The server selects an image analysis module to extract the main features of the image (edges, color histogram).

[1149] The server searches the training source database based on the extracted features to identify datasets of similar landscape images.

[1150] The server analyzes which original image a particular part of the landscape painting comes from and performs a specific mapping.

[1151] The server generates a traceability report and provides it to the user, including information such as "The mountain part is from original image X, and the river part is from original image Y."

[1152] Example 2: Text production

[1153] Users upload articles about the latest technology trends to the system, which are created using a generative AI model.

[1154] The server receives the article and adds it to the analysis queue.

[1155] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1156] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1157] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping.

[1158] The server generates a traceability report and provides it to the user, which indicates which original article a particular technical term or explanation is based on.

[1159] In this way, the system of the present invention can clarify the origin of digital data created using a generative AI model and provide users with an appropriate traceability report.

[1160] Examples of prompt statements

[1161] For images: "Please identify the dataset or training source from which this image originated."

[1162] For text: "Please identify the source or quoted passage from which this article was based."

[1163] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1164] Step 1:

[1165] A user uploads digital data (such as images or text) created by a generative AI model to the system using a dedicated web interface or API. The input is the digital data and its metadata (content type, creation date and time), and the output is an upload request sent to the server. Specifically, the user opens a browser or a dedicated application, selects a file, enters metadata, and clicks the upload button.

[1166] Step 2:

[1167] The server receives an upload request from a user and temporarily stores the digital data and metadata. The input is the digital data and metadata sent by the user, and the output is the data stored in the server's temporary storage database. Specifically, the server saves the uploaded data in a specified directory (such as / tmp / uploads) and simultaneously records the metadata in the database.

[1168] Step 3:

[1169] The server adds the information of the stored digital data to the analysis queue. The input is the path and metadata of the temporarily stored digital data, and the output is the task added to the analysis queue. Specifically, the server adds the file path and metadata to a queue system (e.g., RabbitMQ or Kafka).

[1170] Step 4:

[1171] The server retrieves digital data from the analysis queue and selects an appropriate analysis module based on its type (image, text, etc.). The input is the data and metadata retrieved from the analysis queue, and the output is the selection of an analysis module. Specifically, the server checks the metadata and selects an image analysis module if the analysis target is an image, or a text analysis module if the analysis target is text.

[1172] Step 5:

[1173] The server analyzes the digital data using the selected analysis module to extract features and keywords. The input is the digital data entered into the analysis module, and the output is the extracted features and keywords. Specifically, the image analysis module uses OpenCV to extract edges and color histograms, and the text analysis module uses NLTK to perform grammar analysis and tokenization.

[1174] Step 6:

[1175] The server searches the training source database based on the extracted features and keywords to identify similar data. The input is the extracted features and keywords, and the output is information about similar training source data. Specifically, the server uses SQL or Elasticsearch to search the database and identify data with high similarity scores.

[1176] Step 7:

[1177] The server analyzes the search results in detail and identifies which parts of the generated digital data depend on which learning sources. The input is the search results, and the output is information showing the correspondence between the generated data and the learning sources. Specifically, statistical methods and clustering techniques are used to map specific parts to the learning sources.

[1178] Step 8:

[1179] The server generates a traceability report based on the identified information. The input is information showing the correspondence between the product and the learning source, and the output is a traceability report. Specifically, the report includes the URL of the learning source and the location information of the identified part, and is generated in PDF or HTML format.

[1180] Step 9:

[1181] The server provides the generated traceability report to the user. The input is the traceability report, and the output is the report in a format that the user can download. Specifically, the user logs into the web interface and clicks on a link to view and download the report.

[1182] (Application example 1)

[1183] 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."

[1184] The origin of digital content generated by generative AI models is often unclear, potentially raising issues regarding the reliability and copyright of the content. It is also difficult to track which learning source user-generated content was based on. Furthermore, there is no established method for utilizing this information in real time. To solve these problems, a system is needed that can clarify the origin of digital content and analyze its generation process.

[1185] 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.

[1186] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and decompose the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of the product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for a smartphone or robot to analyze the digital content in real time; and a means for visually displaying the information identified by the server. This makes the origin of the digital content clear, enabling the use of highly reliable digital content. Furthermore, real-time origin verification is also possible, allowing users to quickly verify the reliability of the content.

[1187] "Digital Content" refers to information that is generated, stored, and transmitted by electronic means.

[1188] "Generative AI" refers to systems or technologies that use artificial intelligence to generate new digital content.

[1189] "Uploading" refers to the act of a user sending digital content from their own terminal to a server.

[1190] "Server" refers to a computer system that receives, stores, processes, and serves digital data.

[1191] "Analysis" refers to the technical process of breaking down the content of digital content and clarifying its characteristics and the process by which it was created.

[1192] "Learning source" refers to the dataset or information source that generative AI uses to generate new content.

[1193] "Real-time" refers to processing and results being provided immediately.

[1194] "Visual display" refers to displaying analysis results and traceability information on a screen or display in a format that is easy for users to understand.

[1195] "Feature extraction" refers to the technique of identifying and extracting important attributes and patterns from digital content.

[1196] "Traceability" refers to the ability to track and record the origin of a product or the provenance of its components.

[1197] Overall system overview

[1198] This invention is a system that provides traceability functions to clarify the origin of digital content created by generative AI. Users upload the creations, and the server analyzes them to identify the learning source and usage points, and provides that information to the user. This system can be applied to security services, particularly applications installed on smartphones or robots that perform real-time analysis.

[1199] System configuration

[1200] 1. User Device

[1201] A user terminal is a device that users use to upload digital content created by generative AI. This includes smartphones, tablets, and PCs.

[1202] 2. Server

[1203] The server is the central component with the following functions:

[1204] Receiving and temporarily storing digital content: Receives and temporarily stores digital content uploaded by users.

[1205] Digital content analysis: Analyze digital content and break down its characteristics and creation process.

[1206] Identifying learning sources: Based on the analysis results, the learning sources are searched from the database.

[1207] Identifying the parts of the product that are based on learning sources and identifying which learning sources they depend on.

[1208] Providing traceability information: The identified information is provided to the user as a traceability report.

[1209] 3. Learning Source Database

[1210] The training source database is a database that stores the datasets and information sources used to train the generative AI. This database contains a large amount of datasets (image data, text data).

[1211] 4. Smartphone Robot

[1212] This system can be applied as an application installed on smartphones or robots. These devices have the ability to analyze digital content and provide traceability information in real time.

[1213] Program processing overview

[1214] Uploading Digital Content

[1215] Users upload digital content using a dedicated interface (for example, a web browser or a dedicated application). When uploading, metadata such as the content type and creation date and time are also added.

[1216] Receiving and temporarily storing digital content

[1217] The server receives and temporarily stores the digital content uploaded by the user, and once the storage is complete, adds the digital content to an analysis queue.

[1218] Digital content analysis

[1219] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.). For example, for images, it identifies edges and color histograms, while for text, it performs grammar analysis and keyword extraction.

[1220] Identifying learning sources

[1221] The server uses the analysis results to search the training source database to identify similar datasets, using feature matching and clustering techniques.

[1222] Identifying the location of use

[1223] The server performs detailed analysis of specific parts of the production based on the learning source and clarifies which learning source it depends on, thereby generating highly reliable traceability information.

[1224] Providing traceability information

[1225] The server generates a traceability report based on the identified information, which includes the URL of the learning source and details of the specific part, and can be viewed and checked by the user.

[1226] Examples of concrete examples and prompts

[1227] For example, to verify the originality of a photo taken with a smartphone, the user uploads the photo, the server analyzes the photo's characteristics, and searches the learning source database. A traceability report is generated based on the search results and provided to the user.

[1228] Example prompt sentence:

[1229] Task: Trace provenance

[1230] Content Type: Image

[1231] File path: 'path / to / image.jpg'

[1232] This system allows users to quickly verify the authenticity of digital content and obtain traceability information in real time.

[1233] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1234] Step 1: Upload your digital content

[1235] Users upload digital content using their smartphones or computers. They use a dedicated web interface or application to send the digital content file and its metadata (content type, creation date, etc.). The input is the digital content file selected by the user, and the output is the data received by the server. The specific action is when the user clicks the "Upload" button.

[1236] Step 2: Receive and temporarily store digital content

[1237] The server receives digital content sent by the user and temporarily stores it in storage. The input is the digital content file sent by the user, and the output is a file stored in the server's temporary storage area. Specifically, the server analyzes the HTTP request and writes the file data to disk.

[1238] Step 3: Analyzing the digital content

[1239] The server retrieves the temporarily stored digital content and selects an analysis module depending on the type. For example, for image files, it performs edge detection and color histogram analysis. For text files, it performs grammatical analysis and keyword extraction. The input is the temporarily stored digital content file, and the output is the analyzed feature data. Specifically, the server loads the analysis module and analyzes the data.

[1240] Step 4: Identify learning sources

[1241] The server searches the training source database based on the analysis results. It uses the extracted feature data as input to identify similar datasets. The input is the analyzed feature data, and the output is similar datasets identified from the training source database. Specifically, the server inputs the feature values ​​into a clustering algorithm and calculates the similarity.

[1242] Step 5: Identify the usage

[1243] The server determines which learning source a particular part of the digital content depends on based on information obtained from the learning source database. The input is information on similar datasets obtained from the learning source database, and the output is mapping information between each part of the product and its corresponding learning source. Specifically, the server uses statistical methods to match each part of the digital content with the learning source.

[1244] Step 6: Generate traceability information

[1245] The server generates a traceability report based on the identified usage location information. The report includes the URL of the learning source and detailed information about the specific part. The input is the mapping information, and the output is the traceability report provided to the user. Specifically, the server creates a report format based on the mapping information and fills in the data.

[1246] Step 7: Provide traceability information

[1247] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The input is the traceability report, and the output is a visualized report that the user can access. Specifically, the server returns the report upon the user's request.

[1248] As described above, this system is composed of multiple steps to clarify the origin of digital content and guarantee its reliability.

[1249] 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.

[1250] Overall system overview

[1251] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI, and an emotion engine that recognizes user emotions and optimizes analysis results. In this system, a user provides a product, a server analyzes the product to identify learning sources and usage points, adjusts the analysis results using the emotion engine, and provides that information to the user.

[1252] System configuration

[1253] The system includes the following major components:

[1254] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1255] 2. Server: A central system with the following functions:

[1256] Receiving digital content

[1257] Temporary storage of digital content

[1258] Digital content analysis

[1259] Searching the learning source database

[1260] Identifying the location of use

[1261] Generate and provide traceability reports

[1262] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1263] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[1264] Program processing

[1265] The processing of the system programs will be explained in detail below.

[1266] Providing and Receiving Products

[1267] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[1268] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1269] Digital content analysis

[1270] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on the type (image, text, etc.). For example, it launches an image analysis module for images and a natural language processing module for text.

[1271] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[1272] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1273] Identifying learning sources and where to use them

[1274] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[1275] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1276] Use of emotion engine

[1277] The server activates an emotion engine to recognize the user's emotion.

[1278] The emotion engine analyzes the user's facial expressions and voice to identify emotions, for example, by using a webcam or microphone to collect real-time emotion data.

[1279] The emotion engine adjusts the analysis results based on the identified emotion, for example providing more detailed traceability information if the user is confused, or a concise report if the user is happy.

[1280] Generating and providing traceability information

[1281] The server combines the identified information with the results of the sentiment engine to generate a traceability report, which includes the URL of the learning source, the location of the citation, and associated metadata.

[1282] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[1283] Specific examples

[1284] Example 1: Image creation

[1285] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[1286] The server receives the images and adds them to the analysis queue.

[1287] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[1288] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1289] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[1290] The server activates the emotion engine, which analyzes the user's facial expressions and voice to identify their emotion. For example, if the user is confused, a more detailed explanation is added to the report.

[1291] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[1292] Example 2: Text production

[1293] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[1294] The server receives the article and adds it to the analysis queue.

[1295] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1296] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1297] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[1298] The server starts the emotion engine, which analyzes the user's emotions. For example, if the user has a positive feeling, it generates a concise and positive report.

[1299] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[1300] The above is an embodiment of the present invention. The present invention makes it possible to clarify the origin of digital content created using generative AI and provide an optimal report that corresponds to the user's emotions.

[1301] The processing flow will be explained below.

[1302] Step 1:

[1303] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[1304] Step 2:

[1305] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[1306] Step 3:

[1307] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[1308] Step 4:

[1309] The server analyzes the digital content using the selected analysis module.

[1310] Image Analysis Module: Uses feature extraction algorithms to identify key features of an image (edges, color, shape, etc.).

[1311] Text analysis module: Using natural language processing technology, it analyzes the structure of text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1312] Step 5:

[1313] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on the extracted keywords.

[1314] Step 6:

[1315] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1316] Step 7:

[1317] The server activates an emotion engine to recognize the user's emotion.

[1318] The emotion engine collects facial and voice data from the user and identifies their emotions based on this data. The data is collected in real time using a webcam and microphone.

[1319] An emotion engine tailors the presentation of the analysis results based on the identified emotion.

[1320] Step 8:

[1321] The server uses the emotion engine to generate a traceability report based on the refined analysis, including the URL of the learning source, the location of the citation, and related metadata.

[1322] Step 9:

[1323] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Detailed or concise reports are provided based on emotions to help users understand.

[1324] Example: Image production

[1325] 1. A user uploads an image of a landscape painting to the system.

[1326] 2. The server receives the image and adds it to the analysis queue.

[1327] 3. The server selects an image analysis module and extracts key features.

[1328] 4. The server searches the learning source database based on the extracted features to identify the original image.

[1329] 5. The server maps specific parts of the image to the original image.

[1330] 6. The server launches the emotion engine, which analyzes the user's facial expressions and voice in real time to identify their emotions.

[1331] 7. The server generates a traceability report based on the identified emotions in a format that is understandable to the user.

[1332] 8. The server provides the report to the user, who can view it via a web interface.

[1333] Example: For text productions

[1334] 1. A user uploads a technical article to the system.

[1335] 2. The server receives the article and adds it to the analysis queue.

[1336] 3. The server selects a text analysis module and performs a structural analysis of the text.

[1337] 4. The server searches the learning source database based on the extracted keywords to identify the original article.

[1338] 5. The server maps specific parts of the technical article to the original article.

[1339] 6. The server launches the emotion engine and analyzes the user's emotions in real time.

[1340] 7. The server generates a traceability report in an appropriate format based on the identified emotion.

[1341] 8. The server provides the report to the user, who can view it via a web interface.

[1342] This will clarify the provenance of digital content using generative AI and provide optimal reporting based on user sentiment.

[1343] Example 2

[1344] 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."

[1345] The origin of digital content created by generative AI is unclear, making it impossible to identify which learning sources it relies on. Furthermore, there is a problem in that traceability information is difficult to understand because appropriate information is not provided according to user sentiment.

[1346] 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.

[1347] In this invention, the server includes a means for receiving and temporarily storing digital content created by the AI ​​by a user, a means for analyzing the digital content and breaking down the creation process, and a means for identifying learning sources based on the digital content. This makes it possible to clarify the origin of the digital content created by the AI ​​and provide appropriate information according to the user's emotions.

[1348] "User" means an individual or organization that uses the system to upload digital content created by generative AI.

[1349] "Generative AI" is a system that automatically generates digital content (images, text, etc.) using artificial intelligence technology.

[1350] "Digital content" refers to electronic data created by generative AI, including images, text, and audio.

[1351] "Uploading" is the act of a user sending digital content created by generative AI to a server.

[1352] "Server" means a centralized computer system that receives, analyzes, stores, and serves digital content.

[1353] A "temporary storage means" is a storage device or storage service that allows a server to temporarily store digital content.

[1354] "Means for analyzing" refers to software or hardware that the server uses to analyze digital content and extract its structure and characteristics.

[1355] "Means for decomposing the creation process" refers to analytical techniques that allow the server to clarify the elements and methods used to create digital content.

[1356] A "learning source" is a dataset or information source that a generative AI uses for learning or training.

[1357] "Means for identifying" refers to the technology or algorithm that the server uses to identify which learning source a piece of digital content relies on.

[1358] The "means of provision" refers to the interface or API that the server uses to communicate and display analysis results and traceability information to users.

[1359] The "means for analyzing emotions" refers to machine learning models and emotion recognition technologies that allow the server to analyze the user's facial expressions and voice and identify the user's emotions.

[1360] "Means for adjusting information" refers to technologies and algorithms for optimizing traceability information by taking into account the results of user sentiment analysis.

[1361] Overall system overview

[1362] This invention relates to a system that clarifies the origin of digital content created by generative AI, recognizes user emotions, and optimizes analysis results. By using this system, it is possible to clarify which dataset the generative AI used to generate the content, and provide information that meets the user's needs.

[1363] System configuration

[1364] The system includes the following major components:

[1365] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1366] 2. Server: A central system with the following functions:

[1367] Receiving digital content

[1368] Temporary storage of digital content

[1369] Digital content analysis

[1370] Searching the learning source database

[1371] Identifying the location of use

[1372] Generate and provide traceability reports

[1373] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1374] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[1375] Details of data processing and data calculation

[1376] The specific operation of the system will be described below.

[1377] Providing and Receiving Products

[1378] Users upload digital content created by generative AI through a dedicated web interface or API, and provide simple metadata (content type, creation date, etc.) at the time.

[1379] The server receives the uploaded digital content and temporarily stores it in a cloud storage location (e.g., Amazon S3). Once the storage is complete, the digital content is added to the analysis queue.

[1380] Digital content analysis

[1381] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1382] Image Analysis Module: Performs image feature extraction using the OpenCV library to identify key features (edges, hue, shape, etc.).

[1383] Text analysis module: Uses natural language processing libraries such as NLTK and spaCy to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[1384] Identifying learning sources and where to use them

[1385] The server searches the training source database based on the extracted features and keywords, for example, by using SQL queries or Elasticsearch to search the data.

[1386] The server then uses the search results to perform a detailed analysis of which learning sources specific parts of the artifacts depend on, applying statistical methods and machine learning algorithms such as K-means clustering and DBSCAN.

[1387] Use of emotion engine

[1388] The server runs an emotion engine, which analyzes the user's facial expressions and voice to identify emotions. It uses libraries such as OpenVINO and Dlib to identify emotions such as smiling, confused, or angry using data from the webcam and microphone.

[1389] The emotion engine adjusts the analysis results based on the identified emotion, for example adding more detailed information to the traceability report if the user is confused.

[1390] Generating and providing traceability information

[1391] The server combines the identified information (the URL of the learning source, the location of the quoted section, and related metadata) with the results of the user sentiment analysis and generates a traceability report using tools such as JasperReports and Crystal Reports.

[1392] The server provides the generated traceability report to the user, who can view and download it via the system's web interface or API.

[1393] Specific examples

[1394] Example 1: Image creation

[1395] Users upload vivid landscape images created by generative AI through the system.

[1396] The server receives the images and adds them to the analysis queue.

[1397] The server selects an image analysis module to extract key features such as edges and color histograms.

[1398] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1399] The server activates an emotion engine, which analyzes the user's facial expressions and voice to identify their emotions. For example, if the user is confused, more detailed information is added to the traceability report.

[1400] The server generates a traceability report, providing the user with details about the original source of the image.

[1401] Example 2: Text production

[1402] Users upload AI-generated articles about the latest technology trends through the system.

[1403] The server receives the article and adds it to the analysis queue.

[1404] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1405] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1406] The server will launch an emotion engine, which will analyze the user's emotion, for example, if the user is confused, it will add more detailed information to the traceability report.

[1407] The server generates a traceability report, providing the user with details about the original source of the technical article.

[1408] The above is a specific embodiment of the present invention. This system makes it possible to clarify the origin of digital content created by generative AI and provide optimal traceability information according to the user's feelings.

[1409] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1410] Step 1: Providing and Receiving Products

[1411] Users upload digital content (e.g., landscape images or technical articles) created by generative AI using the system's web interface or API. As input, the digital content and its metadata (content type, creation date, etc.) are provided.

[1412] The server receives the uploaded digital content, and the output is that the digital content and metadata are stored on the server.

[1413] The server temporarily stores the content in cloud storage (e.g., Amazon S3) and adds it to the analysis queue, so the content is ready for analysis.

[1414] Step 2: Analyzing the digital content

[1415] The server sequentially retrieves digital content from the analysis queue, and the input is the digital content added to the analysis queue.

[1416] The server selects an analysis module based on the type of digital content (image, text, etc.), for example, an image analysis module for images and a text analysis module for text.

[1417] The server launches the selected modules to analyze the digital content. Specifically, the image analysis module uses the OpenCV library to analyze edges, hues, shapes, etc., while the text analysis module uses NLTK and spaCy to perform grammar analysis and keyword extraction. The output is the extracted features and keywords.

[1418] Step 3: Identify learning sources and where to use them

[1419] The server uses the features and keywords obtained from the analysis results to search the learning source database. The input is the features and keywords from the analysis results.

[1420] The server uses SQL queries or Elasticsearch to identify relevant datasets from the training source database, and the output is the relevant training source data.

[1421] The server uses statistical methods and machine learning algorithms such as K-means clustering and DBSCAN to perform a detailed analysis of which learning sources specific parts of the product depend on, thereby obtaining mapping information between the product and the learning sources.

[1422] Step 4: Use the Emotion Engine

[1423] The server starts the emotion engine, and inputs are the user's facial expressions and voice data.

[1424] The emotion engine uses libraries such as OpenVINO and Dlib to analyze the user's facial expressions and voice in real time to identify emotions. The output is the user's emotional information.

[1425] The server adjusts the analysis results based on the emotion engine's results, for example adding more detailed information to the traceability report if the user is confused.

[1426] Step 5: Generate and provide traceability information

[1427] The server integrates the identified information (the URL of the learning source, the location of the quoted part, and related metadata) with the sentiment analysis results to generate a traceability report. The input is the learning source information and the sentiment analysis results.

[1428] The server generates the traceability report using tools such as JasperReports or Crystal Reports. The output is a comprehensive traceability report.

[1429] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The output is a user-accessible traceability report.

[1430] (Application example 2)

[1431] 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."

[1432] It is important to identify the origin of digital content created by generative AI and provide it to users in a transparent manner. However, conventional systems have difficulty in identifying in detail which learning sources a generated product relies on, which is insufficient to help users understand. Furthermore, they lack the ability to provide feedback tailored to the user's emotions, resulting in a suboptimal user experience.

[1433] 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.

[1434] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and break down the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of a product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for the server to analyze the user's emotions using an emotion recognition engine; and a means for the server to adjust the analysis results based on the user's emotions and provide optimal feedback. This not only makes it possible to identify the origin of the product in detail and provide it to the user transparently, but also to provide optimal feedback based on the user's emotions, improving the user experience.

[1435] "Generative AI" is a system that automatically generates content using artificial intelligence technology.

[1436] "Digital content" refers to information that is generated and stored electronically, such as images, videos, and text.

[1437] "Upload" is an operation in which a user sends data from their device to a server.

[1438] A "server" is a central processing unit that has functions such as receiving, storing, analyzing, and providing data.

[1439] "Temporarily storing" refers to storing the received digital content in a server for a certain period of time.

[1440] "Analysis" is the process of examining received digital content to identify its components and characteristics.

[1441] "Decomposing the generation process" refers to breaking down the creation steps and original data of content created by generative AI into their component parts and revealing them.

[1442] A "learning source" is a dataset or information source that the generative AI references when generating content.

[1443] "Dependency" refers to the state in which one element depends on another element for its existence or function.

[1444] "Identifying" refers to the act of clarifying related information and elements based on the analysis results.

[1445] "Providing" means showing or making available to the user the analysis results or other information.

[1446] An "emotion recognition engine" is a system that analyzes a user's facial expressions and voice data to identify their emotions.

[1447] "Feedback" refers to information or advice returned to the user based on the analysis results and evaluation.

[1448] MODE FOR CARRYING OUT THE INVENTION

[1449] The present invention relates to a system that identifies the origin of digital content created by generative AI and provides appropriate feedback to users. The system of the present invention mainly includes the following components:

[1450] Overall system overview

[1451] The system of the present invention is composed of a user terminal, a server, a learning source database, and an emotion recognition engine.

[1452] Component Details

[1453] User terminal

[1454] A user device is a device such as a smartphone or PC that allows users to upload digital content created by generative AI. The user device has the function of sending content to a server through a dedicated application.

[1455] server

[1456] The server is a central system with multiple functions:

[1457] 1. Receiving and storing content:

[1458] It has the function of receiving and temporarily storing digital content uploaded by users.

[1459] 2. Content Analysis:

[1460] It analyzes the content uploaded by users and breaks down the generation process, using an image analysis module for images and a natural language processing module for text.

[1461] 3. Identifying learning sources:

[1462] Based on the results of the analysis, similar data is identified by searching the training source database, which stores the datasets and information used to train the generative AI.

[1463] 4. Emotion recognition and feedback regulation:

[1464] It has the ability to analyze the user's emotions using an emotion recognition engine and adjust the content of the feedback provided based on the results.

[1465] Emotion Recognition Engine

[1466] An emotion recognition engine is a system that analyzes a user's facial expressions and voice to identify their emotions. It uses a webcam or microphone to collect emotional data in real time and recognizes emotions based on that data. Specific emotion recognition technologies used include EmotionRecognizer.

[1467] Specific examples of operations

[1468] Content upload and analysis

[1469] For example, consider the case where a user wants to upload a landscape image created using a generative AI model. The user launches an application on a smartphone or PC, selects a landscape image, and uploads it.

[1470] Obtaining a traceability report

[1471] Uploaded images are sent to a server for temporary storage. The server uses an image analysis module to extract key features from the image and uses that data to search a database of training sources. Once similar data is identified, a traceability report is generated, including the URL of the original image and details of the citation.

[1472] Emotion recognition and feedback provision

[1473] When a user receives a report, an emotion recognition engine analyzes the user's facial expressions and voice. For example, if the user is confused, a more detailed explanation can be added to the report. Conversely, if the user is satisfied, a concise and positive report can be provided. This improves the user experience.

[1474] Examples of prompt statements

[1475] For example, you can generate a landscape image by inputting the following prompt sentence into a generative AI model:

[1476] "Generate the following landscape image. It's a beautiful scene featuring a bright blue sky, lush green mountains, and a flowing river."

[1477] Using this prompt, users can generate digital content based on the specified content, and through a subsequent traceability and feedback process, they can obtain more detailed information and appropriate feedback.

[1478] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1479] Step 1:

[1480] User device uploads generated content

[1481] Users upload digital content (e.g., images or text) created using generative AI models to a server via their device (smartphone or PC), and enter the content's metadata (e.g., content type, creation date, etc.).

[1482] Input: Generated content and metadata

[1483] Output: Sending content to the server

[1484] Specific operation: The user launches the dedicated application, clicks the upload button, selects the content file and metadata, and sends them.

[1485] Step 2:

[1486] The server receives the content and temporarily stores it

[1487] The server receives and temporarily stores digital content uploaded by users, which then adds the content to an analysis queue.

[1488] Input: Generated content and metadata sent from the user device

[1489] Output: Save content to temporary storage area on the server

[1490] Specific operation: The server receives the HTTP request, analyzes the attachment, and stores it in a temporary storage area.

[1491] Step 3:

[1492] The server analyzes the digital content

[1493] The server sequentially retrieves stored digital content from the analysis queue and selects the appropriate analysis module based on the content type (e.g., image, text): for images, it uses the image analysis module, and for text, it uses the natural language processing module.

[1494] Input: Digital content stored in temporary storage

[1495] Output: Extracted features and keywords

[1496] What it does: The server's internal analysis engine loads the content and performs feature extraction processing. For example, edge detection and color histogram analysis are used to extract image features.

[1497] Step 4:

[1498] The server searches the learning source database to identify similar data

[1499] The server searches the learning source database based on the extracted features and keywords to identify similar data, thereby clarifying which learning source the content relies on.

[1500] Input: Extracted features and keywords

[1501] Output: Similar data and related information

[1502] What happens: The server searches the database using SQL queries and search algorithms to identify the most similar data.

[1503] Step 5:

[1504] The server generates a traceability report and provides it to the user.

[1505] The server generates a traceability report based on the identified learning sources, including the URL of the original data and the location of specific parts, and provides the report to users via a web interface or API.

[1506] Input: Similar data and related information

[1507] Output: Traceability report

[1508] What it does: The server runs a report generation script to create a report detailing the identified learning sources, which is then saved in a user-accessible format.

[1509] Step 6:

[1510] The server uses an emotion recognition engine to analyze the user's emotions.

[1511] When a user receives a traceability report, the server uses an emotion recognition engine to analyze the user's face and voice to collect emotional data, using a library called EmotionRecognizer.

[1512] Input: User facial and voice data

[1513] Output: Recognized emotion data

[1514] What it does: The emotion recognition engine collects data in real time from your webcam and microphone, then runs emotion analysis algorithms to identify emotions.

[1515] Step 7:

[1516] The server adjusts the analysis results based on the user's emotions and provides feedback.

[1517] The server adjusts the feedback it provides based on the user's recognized emotional data: providing detailed explanations if the user is confused, and concise feedback if the user is satisfied.

[1518] Input: Recognized emotion data

[1519] Output: Regulated Feedback

[1520] Specific operation: The server dynamically generates feedback content based on emotion data and displays it to the user, allowing the user to receive optimized feedback.

[1521] 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.

[1522] 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.

[1523] 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.

[1524] [Fourth embodiment]

[1525] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1526] 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.

[1527] 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).

[1528] 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.

[1529] 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.

[1530] 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).

[1531] 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.

[1532] 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.

[1533] 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.

[1534] 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.

[1535] 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.

[1536] 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.

[1537] 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."

[1538] Overall system overview

[1539] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[1540] System configuration

[1541] The system includes the following major components:

[1542] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1543] 2. Server: A central system with the following functions:

[1544] Receiving digital content

[1545] Temporary storage of digital content

[1546] Digital content analysis

[1547] Searching the learning source database

[1548] Identifying the location of use

[1549] Generate and provide traceability reports

[1550] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1551] Program processing

[1552] The program of this system is executed in the following manner.

[1553] Providing and Receiving Products

[1554] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[1555] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1556] Digital content analysis

[1557] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1558] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[1559] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1560] Identifying learning sources

[1561] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[1562] Identifying the location of use

[1563] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1564] Generating and providing traceability information

[1565] The server generates a traceability report based on the identified information, which includes the URL of the learning source and the location of the identified part.

[1566] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[1567] Specific examples

[1568] Example 1: Image creation

[1569] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[1570] The server receives the images and adds them to the analysis queue.

[1571] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[1572] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1573] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[1574] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[1575] Example 2: Text production

[1576] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[1577] The server receives the article and adds it to the analysis queue.

[1578] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1579] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1580] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[1581] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[1582] The above is an embodiment of the present invention, which makes it possible to provide a specific system for clarifying the origin of digital content created using generative AI and minimizing the risk of copyright infringement.

[1583] The processing flow will be explained below.

[1584] Step 1:

[1585] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[1586] Step 2:

[1587] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[1588] Step 3:

[1589] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[1590] Step 4:

[1591] The server analyzes the digital content using the selected analysis module.

[1592] Image Analysis Module: Extracts key features from an image using techniques such as edge detection, color analysis, and shape recognition.

[1593] Text Analysis Module: Performs grammatical analysis, tokenization, keyword extraction, etc. to identify important keywords and phrases.

[1594] Step 5:

[1595] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on keywords.

[1596] Step 6:

[1597] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources. This analysis uses statistical and clustering techniques to perform a specific mapping, identifying which original information a particular image part or text phrase comes from.

[1598] Step 7:

[1599] The server generates a traceability report based on the results, including the URL of the learning source, the location of the citation, and related metadata. The report is generated in a detailed and easy-to-understand format.

[1600] Step 8:

[1601] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Based on the report, the user can clearly understand the origin and usage of the product.

[1602] These are the processing steps of the system, which allow users to check in detail the origin of digital content created by generative AI, reducing the risk of copyright infringement.

[1603] Example 1

[1604] 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."

[1605] Currently, it is difficult to clarify the origin of digital data created by an artificial intelligence model and provide appropriate traceability. Therefore, a system is needed that identifies the learning source from which digital data was generated and provides this information to the user. The present invention aims to solve this problem and provide an efficient and effective system for identifying the origin of generated digital data.

[1606] 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.

[1607] In this invention, the server includes: means for a user to upload digital data created by an AI model; means for the server to receive and temporarily store the digital data; means for the server to analyze the digital data and decompose the generation process; means for the server to identify learning sources based on the digital data; means for the server to perform a detailed analysis of which parts of a product depend on which learning sources based on the identified learning sources; and means for the server to generate a traceability report based on the identified information and provide it to the user. This makes it possible to clarify the origin of digital data created by an AI model and provide an appropriate traceability report to the user.

[1608] "User" means any person or entity that utilizes an artificial intelligence model to generate digital data and uploads that data to the System.

[1609] An "artificial intelligence model" refers to a program or system that uses machine learning or deep learning algorithms to automatically create artifacts from data.

[1610] "Digital data" refers to information that is stored, processed, and transmitted electronically, and this information can exist in the form of images, text, audio, video, etc.

[1611] "Server" refers to a computer system capable of processing, analyzing, storing, and serving digital data received from users.

[1612] "Analysis module" refers to a software component that analyzes digital data according to its type and extracts features and keywords.

[1613] "Learning source" refers to a dataset or information source used in the training process of an artificial intelligence model, providing the information that forms the basis of the generated digital data.

[1614] "Learning source database" refers to a database that stores the underlying data sets and information sources that identify the origin of generated digital data.

[1615] A "traceability report" is a report that clearly shows the relationship between the analyzed digital data and its learning source, including information indicating which original data a particular part is based on.

[1616] Overall system overview

[1617] The system of the present invention aims to clarify the origin of digital data created by a generative AI model and provide appropriate traceability. In this system, a user provides a product, and a server analyzes the product to identify the learning source and usage point, and provides that information to the user.

[1618] System configuration

[1619] The system includes the following major components:

[1620] 1. User terminal: A device (PC, smartphone, etc.) through which a user uploads digital data created by a generative AI model.

[1621] 2. Server: A central system with the following functions:

[1622] Receiving digital data

[1623] Temporary storage of digital data

[1624] Digital Data Analysis

[1625] Searching the learning source database

[1626] Identifying the location of use

[1627] Generate and provide traceability reports

[1628] 3. Training source database: A database that stores the datasets and information sources used to train a generative AI model.

[1629] Program processing

[1630] The program of this system is processed as follows.

[1631] Providing and Receiving Products

[1632] Users upload digital data created by generative AI models to the system using a dedicated web interface or API, along with metadata such as content type and creation date and time.

[1633] The server receives the uploaded digital data, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1634] Digital Data Analysis

[1635] The server sequentially retrieves the digital data in the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1636] Image analysis module: Uses OpenCV etc. to extract key image features (edges, hue, shape, etc.).

[1637] Text analysis module: Uses NLTK etc. to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[1638] Identifying learning sources

[1639] The server searches the training source database based on the extracted features and keywords, and uses SQL or Elasticsearch to query the database and identify data with high similarity.

[1640] Identifying usage locations and generating traceability information

[1641] The server performs a detailed analysis of the search results to determine which parts of the generated results depend on which learning sources, and determines the correspondence between features and keywords and learning sources using statistical and clustering techniques.

[1642] The server generates a traceability report based on the analysis results, including the correspondence between specific parts of the artifact and the learning source.

[1643] Providing traceability reports

[1644] The server provides the generated traceability reports to the user, which can be downloaded via a web interface or API, in PDF or HTML format.

[1645] Specific examples

[1646] Example 1: Image creation

[1647] Users upload vibrant landscape paintings created by generative AI models to the system.

[1648] The server receives the images and adds them to the analysis queue.

[1649] The server selects an image analysis module to extract the main features of the image (edges, color histogram).

[1650] The server searches the training source database based on the extracted features to identify datasets of similar landscape images.

[1651] The server analyzes which original image a particular part of the landscape painting comes from and performs a specific mapping.

[1652] The server generates a traceability report and provides it to the user, including information such as "The mountain part is from original image X, and the river part is from original image Y."

[1653] Example 2: Text production

[1654] Users upload articles about the latest technology trends to the system, which are created using a generative AI model.

[1655] The server receives the article and adds it to the analysis queue.

[1656] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1657] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1658] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping.

[1659] The server generates a traceability report and provides it to the user, which indicates which original article a particular technical term or explanation is based on.

[1660] In this way, the system of the present invention can clarify the origin of digital data created using a generative AI model and provide users with an appropriate traceability report.

[1661] Examples of prompt statements

[1662] For images: "Please identify the dataset or training source from which this image originated."

[1663] For text: "Please identify the source or quoted passage from which this article was based."

[1664] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1665] Step 1:

[1666] A user uploads digital data (such as images or text) created by a generative AI model to the system using a dedicated web interface or API. The input is the digital data and its metadata (content type, creation date and time), and the output is an upload request sent to the server. Specifically, the user opens a browser or a dedicated application, selects a file, enters metadata, and clicks the upload button.

[1667] Step 2:

[1668] The server receives an upload request from a user and temporarily stores the digital data and metadata. The input is the digital data and metadata sent by the user, and the output is the data stored in the server's temporary storage database. Specifically, the server saves the uploaded data in a specified directory (such as / tmp / uploads) and simultaneously records the metadata in the database.

[1669] Step 3:

[1670] The server adds the information of the stored digital data to the analysis queue. The input is the path and metadata of the temporarily stored digital data, and the output is the task added to the analysis queue. Specifically, the server adds the file path and metadata to a queue system (e.g., RabbitMQ or Kafka).

[1671] Step 4:

[1672] The server retrieves digital data from the analysis queue and selects an appropriate analysis module based on its type (image, text, etc.). The input is the data and metadata retrieved from the analysis queue, and the output is the selection of an analysis module. Specifically, the server checks the metadata and selects an image analysis module if the analysis target is an image, or a text analysis module if the analysis target is text.

[1673] Step 5:

[1674] The server analyzes the digital data using the selected analysis module to extract features and keywords. The input is the digital data entered into the analysis module, and the output is the extracted features and keywords. Specifically, the image analysis module uses OpenCV to extract edges and color histograms, and the text analysis module uses NLTK to perform grammar analysis and tokenization.

[1675] Step 6:

[1676] The server searches the training source database based on the extracted features and keywords to identify similar data. The input is the extracted features and keywords, and the output is information about similar training source data. Specifically, the server uses SQL or Elasticsearch to search the database and identify data with high similarity scores.

[1677] Step 7:

[1678] The server analyzes the search results in detail and identifies which parts of the generated digital data depend on which learning sources. The input is the search results, and the output is information showing the correspondence between the generated data and the learning sources. Specifically, statistical methods and clustering techniques are used to map specific parts to the learning sources.

[1679] Step 8:

[1680] The server generates a traceability report based on the identified information. The input is information showing the correspondence between the product and the learning source, and the output is a traceability report. Specifically, the report includes the URL of the learning source and the location information of the identified part, and is generated in PDF or HTML format.

[1681] Step 9:

[1682] The server provides the generated traceability report to the user. The input is the traceability report, and the output is the report in a format that the user can download. Specifically, the user logs into the web interface and clicks on a link to view and download the report.

[1683] (Application example 1)

[1684] 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."

[1685] The origin of digital content generated by generative AI models is often unclear, potentially raising issues regarding the reliability and copyright of the content. It is also difficult to track which learning source user-generated content was based on. Furthermore, there is no established method for utilizing this information in real time. To solve these problems, a system is needed that can clarify the origin of digital content and analyze its generation process.

[1686] 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.

[1687] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and decompose the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of the product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for a smartphone or robot to analyze the digital content in real time; and a means for visually displaying the information identified by the server. This makes the origin of the digital content clear, enabling the use of highly reliable digital content. Furthermore, real-time origin verification is also possible, allowing users to quickly verify the reliability of the content.

[1688] "Digital Content" refers to information that is generated, stored, and transmitted by electronic means.

[1689] "Generative AI" refers to systems or technologies that use artificial intelligence to generate new digital content.

[1690] "Uploading" refers to the act of a user sending digital content from their own terminal to a server.

[1691] "Server" refers to a computer system that receives, stores, processes, and serves digital data.

[1692] "Analysis" refers to the technical process of breaking down the content of digital content and clarifying its characteristics and the process by which it was created.

[1693] "Learning source" refers to the dataset or information source that generative AI uses to generate new content.

[1694] "Real-time" refers to processing and results being provided immediately.

[1695] "Visual display" refers to displaying analysis results and traceability information on a screen or display in a format that is easy for users to understand.

[1696] "Feature extraction" refers to the technique of identifying and extracting important attributes and patterns from digital content.

[1697] "Traceability" refers to the ability to track and record the origin of a product or the provenance of its components.

[1698] Overall system overview

[1699] This invention is a system that provides traceability functions to clarify the origin of digital content created by generative AI. Users upload the creations, and the server analyzes them to identify the learning source and usage points, and provides that information to the user. This system can be applied to security services, particularly applications installed on smartphones or robots that perform real-time analysis.

[1700] System configuration

[1701] 1. User Device

[1702] A user terminal is a device that users use to upload digital content created by generative AI. This includes smartphones, tablets, and PCs.

[1703] 2. Server

[1704] The server is the central component with the following functions:

[1705] Receiving and temporarily storing digital content: Receives and temporarily stores digital content uploaded by users.

[1706] Digital content analysis: Analyze digital content and break down its characteristics and creation process.

[1707] Identifying learning sources: Based on the analysis results, the learning sources are searched from the database.

[1708] Identifying the parts of the product that are based on learning sources and identifying which learning sources they depend on.

[1709] Providing traceability information: The identified information is provided to the user as a traceability report.

[1710] 3. Learning Source Database

[1711] The training source database is a database that stores the datasets and information sources used to train the generative AI. This database contains a large amount of datasets (image data, text data).

[1712] 4. Smartphone Robot

[1713] This system can be applied as an application installed on smartphones or robots. These devices have the ability to analyze digital content and provide traceability information in real time.

[1714] Program processing overview

[1715] Uploading Digital Content

[1716] Users upload digital content using a dedicated interface (for example, a web browser or a dedicated application). When uploading, metadata such as the content type and creation date and time are also added.

[1717] Receiving and temporarily storing digital content

[1718] The server receives and temporarily stores the digital content uploaded by the user, and once the storage is complete, adds the digital content to an analysis queue.

[1719] Digital content analysis

[1720] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.). For example, for images, it identifies edges and color histograms, while for text, it performs grammar analysis and keyword extraction.

[1721] Identifying learning sources

[1722] The server uses the analysis results to search the training source database to identify similar datasets, using feature matching and clustering techniques.

[1723] Identifying the location of use

[1724] The server performs detailed analysis of specific parts of the production based on the learning source and clarifies which learning source it depends on, thereby generating highly reliable traceability information.

[1725] Providing traceability information

[1726] The server generates a traceability report based on the identified information, which includes the URL of the learning source and details of the specific part, and can be viewed and checked by the user.

[1727] Examples of concrete examples and prompts

[1728] For example, to verify the originality of a photo taken with a smartphone, the user uploads the photo, the server analyzes the photo's characteristics, and searches the learning source database. A traceability report is generated based on the search results and provided to the user.

[1729] Example prompt sentence:

[1730] Task: Trace provenance

[1731] Content Type: Image

[1732] File path: 'path / to / image.jpg'

[1733] This system allows users to quickly verify the authenticity of digital content and obtain traceability information in real time.

[1734] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1735] Step 1: Upload your digital content

[1736] Users upload digital content using their smartphones or computers. They use a dedicated web interface or application to send the digital content file and its metadata (content type, creation date, etc.). The input is the digital content file selected by the user, and the output is the data received by the server. The specific action is when the user clicks the "Upload" button.

[1737] Step 2: Receive and temporarily store digital content

[1738] The server receives digital content sent by the user and temporarily stores it in storage. The input is the digital content file sent by the user, and the output is a file stored in the server's temporary storage area. Specifically, the server analyzes the HTTP request and writes the file data to disk.

[1739] Step 3: Analyzing the digital content

[1740] The server retrieves the temporarily stored digital content and selects an analysis module depending on the type. For example, for image files, it performs edge detection and color histogram analysis. For text files, it performs grammatical analysis and keyword extraction. The input is the temporarily stored digital content file, and the output is the analyzed feature data. Specifically, the server loads the analysis module and analyzes the data.

[1741] Step 4: Identify learning sources

[1742] The server searches the training source database based on the analysis results. It uses the extracted feature data as input to identify similar datasets. The input is the analyzed feature data, and the output is similar datasets identified from the training source database. Specifically, the server inputs the feature values ​​into a clustering algorithm and calculates the similarity.

[1743] Step 5: Identify the usage

[1744] The server determines which learning source a particular part of the digital content depends on based on information obtained from the learning source database. The input is information on similar datasets obtained from the learning source database, and the output is mapping information between each part of the product and its corresponding learning source. Specifically, the server uses statistical methods to match each part of the digital content with the learning source.

[1745] Step 6: Generate traceability information

[1746] The server generates a traceability report based on the identified usage location information. The report includes the URL of the learning source and detailed information about the specific part. The input is the mapping information, and the output is the traceability report provided to the user. Specifically, the server creates a report format based on the mapping information and fills in the data.

[1747] Step 7: Provide traceability information

[1748] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The input is the traceability report, and the output is a visualized report that the user can access. Specifically, the server returns the report upon the user's request.

[1749] As described above, this system is composed of multiple steps to clarify the origin of digital content and guarantee its reliability.

[1750] 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.

[1751] Overall system overview

[1752] The system of the present invention provides a traceability function to clarify the origin of digital content created by generative AI, and an emotion engine that recognizes user emotions and optimizes analysis results. In this system, a user provides a product, a server analyzes the product to identify learning sources and usage points, adjusts the analysis results using the emotion engine, and provides that information to the user.

[1753] System configuration

[1754] The system includes the following major components:

[1755] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1756] 2. Server: A central system with the following functions:

[1757] Receiving digital content

[1758] Temporary storage of digital content

[1759] Digital content analysis

[1760] Searching the learning source database

[1761] Identifying the location of use

[1762] Generate and provide traceability reports

[1763] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1764] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[1765] Program processing

[1766] The processing of the system programs will be explained in detail below.

[1767] Providing and Receiving Products

[1768] Users upload digital content created by generative AI to the system using a dedicated web interface or API, along with simple metadata (content type, creation date, etc.).

[1769] The server receives the uploaded digital content, stores it in a temporary storage location, and once it is saved, adds it to the analysis queue.

[1770] Digital content analysis

[1771] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on the type (image, text, etc.). For example, it launches an image analysis module for images and a natural language processing module for text.

[1772] Image Analysis Module: Performs image feature extraction and identifies key features (edges, color, shape, etc.).

[1773] Text analysis module: Analyzes the structure of the text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1774] Identifying learning sources and where to use them

[1775] The server searches the learning source database based on the extracted features and keywords. The learning source database contains a large number of data sets (image data, text data), and the server searches through these to identify similar data.

[1776] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1777] Use of emotion engine

[1778] The server activates an emotion engine to recognize the user's emotion.

[1779] The emotion engine analyzes the user's facial expressions and voice to identify emotions, for example, by using a webcam or microphone to collect real-time emotion data.

[1780] The emotion engine adjusts the analysis results based on the identified emotion, for example providing more detailed traceability information if the user is confused, or a concise report if the user is happy.

[1781] Generating and providing traceability information

[1782] The server combines the identified information with the results of the sentiment engine to generate a traceability report, which includes the URL of the learning source, the location of the citation, and associated metadata.

[1783] The server provides the generated traceability report to the user, who can view and download it via a web interface or API.

[1784] Specific examples

[1785] Example 1: Image creation

[1786] Users upload images created by generative AI to the system, which are vibrant landscape paintings.

[1787] The server receives the images and adds them to the analysis queue.

[1788] The server selects an image analysis module to extract key features of the image (e.g., edges and color histograms).

[1789] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1790] The server analyzes which original image a particular part of the landscape comes from and performs a specific mapping (for example, which original image a particular mountain or river part is based on).

[1791] The server activates the emotion engine, which analyzes the user's facial expressions and voice to identify their emotion. For example, if the user is confused, a more detailed explanation is added to the report.

[1792] The server generates a traceability report and provides it to the user, who can check the URL of the original image and details of the quoted part.

[1793] Example 2: Text production

[1794] Users upload articles created by generative AI to the system, which discuss the latest technology trends.

[1795] The server receives the article and adds it to the analysis queue.

[1796] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1797] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1798] The server analyzes which original article a particular part of a technical article comes from and performs a specific mapping (for example, which original article a particular technical term or explanation is based on).

[1799] The server starts the emotion engine, which analyzes the user's emotions. For example, if the user has a positive feeling, it generates a concise and positive report.

[1800] The server generates a traceability report and provides it to the user, who can check the URL of the original article and details of the quoted phrase.

[1801] The above is an embodiment of the present invention. The present invention makes it possible to clarify the origin of digital content created using generative AI and provide an optimal report that corresponds to the user's emotions.

[1802] The processing flow will be explained below.

[1803] Step 1:

[1804] Users upload digital content created by generative AI to the system. They use a dedicated web interface or API to send digital content such as images, text, audio, and video. When sending, it is recommended to add metadata such as the content type and the date and time of creation.

[1805] Step 2:

[1806] The server receives the digital content uploaded by the user, temporarily stores it in a database, and once stored, adds it to a queue for analysis.

[1807] Step 3:

[1808] The server sequentially retrieves the digital content added to the analysis queue and selects an appropriate analysis module based on its type, for example, launching an image analysis module for an image and a text analysis module for a text.

[1809] Step 4:

[1810] The server analyzes the digital content using the selected analysis module.

[1811] Image Analysis Module: Uses feature extraction algorithms to identify key features of an image (edges, color, shape, etc.).

[1812] Text analysis module: Using natural language processing technology, it analyzes the structure of text (grammar analysis, tokenization) and extracts important keywords and phrases.

[1813] Step 5:

[1814] The server searches the learning source database based on the extracted features and keywords. For example, in the case of images, the extracted feature vector is used as input to search for similar images, and in the case of text, similar text is searched for based on the extracted keywords.

[1815] Step 6:

[1816] The server performs a detailed analysis of the search results to determine which parts of the product depend on which learning source, using statistical and clustering techniques to map specific parts of the product to specific parts of the learning source.

[1817] Step 7:

[1818] The server activates an emotion engine to recognize the user's emotion.

[1819] The emotion engine collects facial and voice data from the user and identifies their emotions based on this data. The data is collected in real time using a webcam and microphone.

[1820] An emotion engine tailors the presentation of the analysis results based on the identified emotion.

[1821] Step 8:

[1822] The server uses the emotion engine to generate a traceability report based on the refined analysis, including the URL of the learning source, the location of the citation, and related metadata.

[1823] Step 9:

[1824] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. Detailed or concise reports are provided based on emotions to help users understand.

[1825] Example: Image production

[1826] 1. A user uploads an image of a landscape painting to the system.

[1827] 2. The server receives the image and adds it to the analysis queue.

[1828] 3. The server selects an image analysis module and extracts key features.

[1829] 4. The server searches the learning source database based on the extracted features to identify the original image.

[1830] 5. The server maps specific parts of the image to the original image.

[1831] 6. The server launches the emotion engine, which analyzes the user's facial expressions and voice in real time to identify their emotions.

[1832] 7. The server generates a traceability report based on the identified emotions in a format that is understandable to the user.

[1833] 8. The server provides the report to the user, who can view it via a web interface.

[1834] Example: For text productions

[1835] 1. A user uploads a technical article to the system.

[1836] 2. The server receives the article and adds it to the analysis queue.

[1837] 3. The server selects a text analysis module and performs a structural analysis of the text.

[1838] 4. The server searches the learning source database based on the extracted keywords to identify the original article.

[1839] 5. The server maps specific parts of the technical article to the original article.

[1840] 6. The server launches the emotion engine and analyzes the user's emotions in real time.

[1841] 7. The server generates a traceability report in an appropriate format based on the identified emotion.

[1842] 8. The server provides the report to the user, who can view it via a web interface.

[1843] This will clarify the provenance of digital content using generative AI and provide optimal reporting based on user sentiment.

[1844] Example 2

[1845] 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."

[1846] The origin of digital content created by generative AI is unclear, making it impossible to identify which learning sources it relies on. Furthermore, there is a problem in that traceability information is difficult to understand because appropriate information is not provided according to user sentiment.

[1847] 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.

[1848] In this invention, the server includes a means for receiving and temporarily storing digital content created by the AI ​​by a user, a means for analyzing the digital content and breaking down the creation process, and a means for identifying learning sources based on the digital content. This makes it possible to clarify the origin of the digital content created by the AI ​​and provide appropriate information according to the user's emotions.

[1849] "User" means an individual or organization that uses the system to upload digital content created by generative AI.

[1850] "Generative AI" is a system that automatically generates digital content (images, text, etc.) using artificial intelligence technology.

[1851] "Digital content" refers to electronic data created by generative AI, including images, text, and audio.

[1852] "Uploading" is the act of a user sending digital content created by generative AI to a server.

[1853] "Server" means a centralized computer system that receives, analyzes, stores, and serves digital content.

[1854] A "temporary storage means" is a storage device or storage service that allows a server to temporarily store digital content.

[1855] "Means for analyzing" refers to software or hardware that the server uses to analyze digital content and extract its structure and characteristics.

[1856] "Means for decomposing the creation process" refers to analytical techniques that allow the server to clarify the elements and methods used to create digital content.

[1857] A "learning source" is a dataset or information source that a generative AI uses for learning or training.

[1858] "Means for identifying" refers to the technology or algorithm that the server uses to identify which learning source a piece of digital content relies on.

[1859] The "means of provision" refers to the interface or API that the server uses to communicate and display analysis results and traceability information to users.

[1860] The "means for analyzing emotions" refers to machine learning models and emotion recognition technologies that allow the server to analyze the user's facial expressions and voice and identify the user's emotions.

[1861] "Means for adjusting information" refers to technologies and algorithms for optimizing traceability information by taking into account the results of user sentiment analysis.

[1862] Overall system overview

[1863] This invention relates to a system that clarifies the origin of digital content created by generative AI, recognizes user emotions, and optimizes analysis results. By using this system, it is possible to clarify which dataset the generative AI used to generate the content, and provide information that meets the user's needs.

[1864] System configuration

[1865] The system includes the following major components:

[1866] 1. User device: A device (PC, smartphone, etc.) that users use to upload digital content created by generative AI.

[1867] 2. Server: A central system with the following functions:

[1868] Receiving digital content

[1869] Temporary storage of digital content

[1870] Digital content analysis

[1871] Searching the learning source database

[1872] Identifying the location of use

[1873] Generate and provide traceability reports

[1874] 3. Training source database: A database that stores the datasets and information sources used to train the generative AI.

[1875] 4. Emotion engine: A system that recognizes the user's emotions and adjusts the analysis results accordingly.

[1876] Details of data processing and data calculation

[1877] The specific operation of the system will be described below.

[1878] Providing and Receiving Products

[1879] Users upload digital content created by generative AI through a dedicated web interface or API, and provide simple metadata (content type, creation date, etc.) at the time.

[1880] The server receives the uploaded digital content and temporarily stores it in a cloud storage location (e.g., Amazon S3). Once the storage is complete, the digital content is added to the analysis queue.

[1881] Digital content analysis

[1882] The server sequentially retrieves digital content from the analysis queue and selects the appropriate analysis module based on its type (image, text, etc.).

[1883] Image Analysis Module: Performs image feature extraction using the OpenCV library to identify key features (edges, hue, shape, etc.).

[1884] Text analysis module: Uses natural language processing libraries such as NLTK and spaCy to analyze the structure of text (grammar analysis, tokenization) and extract important keywords and phrases.

[1885] Identifying learning sources and where to use them

[1886] The server searches the training source database based on the extracted features and keywords, for example, by using SQL queries or Elasticsearch to search the data.

[1887] The server then uses the search results to perform a detailed analysis of which learning sources specific parts of the artifacts depend on, applying statistical methods and machine learning algorithms such as K-means clustering and DBSCAN.

[1888] Use of emotion engine

[1889] The server runs an emotion engine, which analyzes the user's facial expressions and voice to identify emotions. It uses libraries such as OpenVINO and Dlib to identify emotions such as smiling, confused, or angry using data from the webcam and microphone.

[1890] The emotion engine adjusts the analysis results based on the identified emotion, for example adding more detailed information to the traceability report if the user is confused.

[1891] Generating and providing traceability information

[1892] The server combines the identified information (the URL of the learning source, the location of the quoted section, and related metadata) with the results of the user sentiment analysis and generates a traceability report using tools such as JasperReports and Crystal Reports.

[1893] The server provides the generated traceability report to the user, who can view and download it via the system's web interface or API.

[1894] Specific examples

[1895] Example 1: Image creation

[1896] Users upload vivid landscape images created by generative AI through the system.

[1897] The server receives the images and adds them to the analysis queue.

[1898] The server selects an image analysis module to extract key features such as edges and color histograms.

[1899] The server searches the learning source database based on the extracted features to identify a dataset of similar landscape paintings.

[1900] The server activates an emotion engine, which analyzes the user's facial expressions and voice to identify their emotions. For example, if the user is confused, more detailed information is added to the traceability report.

[1901] The server generates a traceability report, providing the user with details about the original source of the image.

[1902] Example 2: Text production

[1903] Users upload AI-generated articles about the latest technology trends through the system.

[1904] The server receives the article and adds it to the analysis queue.

[1905] The server selects a text analysis module, analyzes the structure of the text, and extracts important keywords and phrases.

[1906] The server searches the learning source database based on the extracted keywords to identify a dataset of similar technical articles.

[1907] The server will launch an emotion engine, which will analyze the user's emotion, for example, if the user is confused, it will add more detailed information to the traceability report.

[1908] The server generates a traceability report, providing the user with details about the original source of the technical article.

[1909] The above is a specific embodiment of the present invention. This system makes it possible to clarify the origin of digital content created by generative AI and provide optimal traceability information according to the user's feelings.

[1910] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1911] Step 1: Providing and Receiving Products

[1912] Users upload digital content (e.g., landscape images or technical articles) created by generative AI using the system's web interface or API. As input, the digital content and its metadata (content type, creation date, etc.) are provided.

[1913] The server receives the uploaded digital content, and the output is that the digital content and metadata are stored on the server.

[1914] The server temporarily stores the content in cloud storage (e.g., Amazon S3) and adds it to the analysis queue, so the content is ready for analysis.

[1915] Step 2: Analyzing the digital content

[1916] The server sequentially retrieves digital content from the analysis queue, and the input is the digital content added to the analysis queue.

[1917] The server selects an analysis module based on the type of digital content (image, text, etc.), for example, an image analysis module for images and a text analysis module for text.

[1918] The server launches the selected modules to analyze the digital content. Specifically, the image analysis module uses the OpenCV library to analyze edges, hues, shapes, etc., while the text analysis module uses NLTK and spaCy to perform grammar analysis and keyword extraction. The output is the extracted features and keywords.

[1919] Step 3: Identify learning sources and where to use them

[1920] The server uses the features and keywords obtained from the analysis results to search the learning source database. The input is the features and keywords from the analysis results.

[1921] The server uses SQL queries or Elasticsearch to identify relevant datasets from the training source database, and the output is the relevant training source data.

[1922] The server uses statistical methods and machine learning algorithms such as K-means clustering and DBSCAN to perform a detailed analysis of which learning sources specific parts of the product depend on, thereby obtaining mapping information between the product and the learning sources.

[1923] Step 4: Use the Emotion Engine

[1924] The server starts the emotion engine, and inputs are the user's facial expressions and voice data.

[1925] The emotion engine uses libraries such as OpenVINO and Dlib to analyze the user's facial expressions and voice in real time to identify emotions. The output is the user's emotional information.

[1926] The server adjusts the analysis results based on the emotion engine's results, for example adding more detailed information to the traceability report if the user is confused.

[1927] Step 5: Generate and provide traceability information

[1928] The server integrates the identified information (the URL of the learning source, the location of the quoted part, and related metadata) with the sentiment analysis results to generate a traceability report. The input is the learning source information and the sentiment analysis results.

[1929] The server generates the traceability report using tools such as JasperReports or Crystal Reports. The output is a comprehensive traceability report.

[1930] The server provides the generated traceability report to the user, who can view and download it via a web interface or API. The output is a user-accessible traceability report.

[1931] (Application example 2)

[1932] 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."

[1933] It is important to identify the origin of digital content created by generative AI and provide it to users in a transparent manner. However, conventional systems have difficulty in identifying in detail which learning sources a generated product relies on, which is insufficient to help users understand. Furthermore, they lack the ability to provide feedback tailored to the user's emotions, resulting in a suboptimal user experience.

[1934] 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.

[1935] In this invention, the server includes: a means for a user to upload digital content created by a generative AI; a means for the server to receive and temporarily store the digital content; a means for the server to analyze the digital content and break down the generation process; a means for the server to identify learning sources based on the digital content; a means for the server to identify which parts of a product depend on which learning sources based on the learning sources; a means for the server to provide the identified information to the user; a means for the server to analyze the user's emotions using an emotion recognition engine; and a means for the server to adjust the analysis results based on the user's emotions and provide optimal feedback. This not only makes it possible to identify the origin of the product in detail and provide it to the user transparently, but also to provide optimal feedback based on the user's emotions, improving the user experience.

[1936] "Generative AI" is a system that automatically generates content using artificial intelligence technology.

[1937] "Digital content" refers to information that is generated and stored electronically, such as images, videos, and text.

[1938] "Upload" is an operation in which a user sends data from their device to a server.

[1939] A "server" is a central processing unit that has functions such as receiving, storing, analyzing, and providing data.

[1940] "Temporarily storing" refers to storing the received digital content in a server for a certain period of time.

[1941] "Analysis" is the process of examining received digital content to identify its components and characteristics.

[1942] "Decomposing the generation process" refers to breaking down the creation steps and original data of content created by generative AI into their component parts and revealing them.

[1943] A "learning source" is a dataset or information source that the generative AI references when generating content.

[1944] "Dependency" refers to the state in which one element depends on another element for its existence or function.

[1945] "Identifying" refers to the act of clarifying related information and elements based on the analysis results.

[1946] "Providing" means showing or making available to the user the analysis results or other information.

[1947] An "emotion recognition engine" is a system that analyzes a user's facial expressions and voice data to identify their emotions.

[1948] "Feedback" refers to information or advice returned to the user based on the analysis results and evaluation.

[1949] MODE FOR CARRYING OUT THE INVENTION

[1950] The present invention relates to a system that identifies the origin of digital content created by generative AI and provides appropriate feedback to users. The system of the present invention mainly includes the following components:

[1951] Overall system overview

[1952] The system of the present invention is composed of a user terminal, a server, a learning source database, and an emotion recognition engine.

[1953] Component Details

[1954] User terminal

[1955] A user device is a device such as a smartphone or PC that allows users to upload digital content created by generative AI. The user device has the function of sending content to a server through a dedicated application.

[1956] server

[1957] The server is a central system with multiple functions:

[1958] 1. Receiving and storing content:

[1959] It has the function of receiving and temporarily storing digital content uploaded by users.

[1960] 2. Content Analysis:

[1961] It analyzes the content uploaded by users and breaks down the generation process, using an image analysis module for images and a natural language processing module for text.

[1962] 3. Identifying learning sources:

[1963] Based on the results of the analysis, similar data is identified by searching the training source database, which stores the datasets and information used to train the generative AI.

[1964] 4. Emotion recognition and feedback regulation:

[1965] It has the ability to analyze the user's emotions using an emotion recognition engine and adjust the content of the feedback provided based on the results.

[1966] Emotion Recognition Engine

[1967] An emotion recognition engine is a system that analyzes a user's facial expressions and voice to identify their emotions. It uses a webcam or microphone to collect emotional data in real time and recognizes emotions based on that data. Specific emotion recognition technologies used include EmotionRecognizer.

[1968] Specific examples of operations

[1969] Content upload and analysis

[1970] For example, consider the case where a user wants to upload a landscape image created using a generative AI model. The user launches an application on a smartphone or PC, selects a landscape image, and uploads it.

[1971] Obtaining a traceability report

[1972] Uploaded images are sent to a server for temporary storage. The server uses an image analysis module to extract key features from the image and uses that data to search a database of training sources. Once similar data is identified, a traceability report is generated, including the URL of the original image and details of the citation.

[1973] Emotion recognition and feedback provision

[1974] When a user receives a report, an emotion recognition engine analyzes the user's facial expressions and voice. For example, if the user is confused, a more detailed explanation can be added to the report. Conversely, if the user is satisfied, a concise and positive report can be provided. This improves the user experience.

[1975] Examples of prompt statements

[1976] For example, you can generate a landscape image by inputting the following prompt sentence into a generative AI model:

[1977] "Generate the following landscape image. It's a beautiful scene featuring a bright blue sky, lush green mountains, and a flowing river."

[1978] Using this prompt, users can generate digital content based on the specified content, and through a subsequent traceability and feedback process, they can obtain more detailed information and appropriate feedback.

[1979] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1980] Step 1:

[1981] User device uploads generated content

[1982] Users upload digital content (e.g., images or text) created using generative AI models to a server via their device (smartphone or PC), and enter the content's metadata (e.g., content type, creation date, etc.).

[1983] Input: Generated content and metadata

[1984] Output: Sending content to the server

[1985] Specific operation: The user launches the dedicated application, clicks the upload button, selects the content file and metadata, and sends them.

[1986] Step 2:

[1987] The server receives the content and temporarily stores it

[1988] The server receives and temporarily stores digital content uploaded by users, which then adds the content to an analysis queue.

[1989] Input: Generated content and metadata sent from the user device

[1990] Output: Save content to temporary storage area on the server

[1991] Specific operation: The server receives the HTTP request, analyzes the attachment, and stores it in a temporary storage area.

[1992] Step 3:

[1993] The server analyzes the digital content

[1994] The server sequentially retrieves stored digital content from the analysis queue and selects the appropriate analysis module based on the content type (e.g., image, text): for images, it uses the image analysis module, and for text, it uses the natural language processing module.

[1995] Input: Digital content stored in temporary storage

[1996] Output: Extracted features and keywords

[1997] What it does: The server's internal analysis engine loads the content and performs feature extraction processing. For example, edge detection and color histogram analysis are used to extract image features.

[1998] Step 4:

[1999] The server searches the learning source database to identify similar data

[2000] The server searches the learning source database based on the extracted features and keywords to identify similar data, thereby clarifying which learning source the content relies on.

[2001] Input: Extracted features and keywords

[2002] Output: Similar data and related information

[2003] What happens: The server searches the database using SQL queries and search algorithms to identify the most similar data.

[2004] Step 5:

[2005] The server generates a traceability report and provides it to the user.

[2006] The server generates a traceability report based on the identified learning sources, including the URL of the original data and the location of specific parts, and provides the report to users via a web interface or API.

[2007] Input: Similar data and related information

[2008] Output: Traceability report

[2009] What it does: The server runs a report generation script to create a report detailing the identified learning sources, which is then saved in a user-accessible format.

[2010] Step 6:

[2011] The server uses an emotion recognition engine to analyze the user's emotions.

[2012] When a user receives a traceability report, the server uses an emotion recognition engine to analyze the user's face and voice to collect emotional data, using a library called EmotionRecognizer.

[2013] Input: User facial and voice data

[2014] Output: Recognized emotion data

[2015] What it does: The emotion recognition engine collects data in real time from your webcam and microphone, then runs emotion analysis algorithms to identify emotions.

[2016] Step 7:

[2017] The server adjusts the analysis results based on the user's emotions and provides feedback.

[2018] The server adjusts the feedback it provides based on the user's recognized emotional data: providing detailed explanations if the user is confused, and concise feedback if the user is satisfied.

[2019] Input: Recognized emotion data

[2020] Output: Regulated Feedback

[2021] Specific operation: The server dynamically generates feedback content based on emotion data and displays it to the user, allowing the user to receive optimized feedback.

[2022] 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.

[2023] 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.

[2024] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2025] 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.

[2026] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mappe...

Claims

1. A means for users to upload digital content created by the generative AI; a server receiving and temporarily storing the digital content; A server analyzes the digital content and analyzes the creation process; a server means for identifying a learning source based on the digital content; means for the server to identify, based on said learning sources, which parts of the production depend on which learning sources; a means for the server to provide the specified information to a user; A system including:

2. The system of claim 1 , wherein the server includes means for selecting an analysis module based on the type of the digital content.

3. 2. The system according to claim 1, wherein the server includes means for searching a learning source database based on the analysis result of the digital content.

Images