system

Abstract

Provide a system. A means for receiving a creative work; a means for assigning unique identification information to the creative work and generating metadata; means for generating a QR code or a URL based on the generated metadata; A means for embedding the generated QR code or URL into a creative work; A means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and A means for measuring the similarity of the new content generated with existing creations in the database; A system including means for reporting the identification information and similarity of highly similar creative works.

Description

[Technical Field]

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

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

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

[0004] In recent years, advances in generative AI have increased the risk of unauthorized use of creative works such as source code, novels, and illustrations. As a result, the rights of authors of creative works are infringed, resulting in situations where they are unable to receive fair compensation. Furthermore, it is difficult to track the degree to which generated content is similar to existing creative works, making it difficult to monitor copyright infringement. Therefore, there is a need for a system that can reject learning uses of creative works and track the similarity of generated content. [Means for solving the problem]

[0005] The present invention provides the following means to solve the above problems. The system includes a means for receiving a creative work and a means for assigning unique identification information to the creative work and generating metadata. The system also provides a means for generating a QR code (registered trademark) or URL based on the generated metadata and embedding it in the creative work. The system also includes a means for checking the metadata of the loaded dataset and excluding creative works with specific flags from learning. Furthermore, the system provides a means for measuring the similarity of generated new content to creative works in an existing database and a means for reporting the identification information and similarity of highly similar creative works, thereby making it possible to refuse the use of the creative work for learning and prevent copyright infringement of the generated content.

[0006] "Creative works" are works created by humans, such as source code, novels, and illustrations.

[0007] "Identifying Information" is unique data generated to represent the unique characteristics of a creative work, such as a SHA-256 hash value.

[0008] "Metadata" is additional information about a creative work, including identification information and a flag indicating whether it can be used for educational purposes.

[0009] A "QR code" is a two-dimensional code that visually represents information and is used, for example, to embed metadata in creative works.

[0010] "URL" means a unique web address providing access to a creative work.

[0011] A "flag" is a signal intended to indicate a particular state or setting associated with a creative work, including, for example, a "Do-Not-Learn" setting.

[0012] A "dataset" is a collection of creative works compiled for study or analysis.

[0013] "Similarity" is a measure that expresses the degree of commonality or similarity between two or more creative works, and for example, cosine similarity is used.

[0014] "Generated content" is content such as text, images, and code newly created by generative AI.

[0015] "Learning" is the process by which generative AI trains a model based on a dataset and generates an output.

[0016] "Reporting" is the act of notifying the user or administrator of the evaluation results of the degree of coincidence or similarity. [Brief explanation of the drawings]

[0017] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0018] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0020] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0021] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0022] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0023] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0024] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0025] [First embodiment]

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

[0027] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0028] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0029] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0030] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0031] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0032] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0034] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0035] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0036] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0037] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0038] This invention relates to a system that protects the copyright of creative works and prevents their unauthorized use by AI. Below, we will generate a program for this system and explain its processing in natural language.

[0039] Overall system overview

[0040] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities.

[0041] Program processing

[0042] Upload your creations and generate metadata

[0043] Users upload their creations to the server, which then generates identification information for the received creations using the SHA-256 hash algorithm and includes that information in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata.

[0044] Generate a QR code or URL

[0045] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0046] Processing the training dataset

[0047] When a device (including an AI model) loads a new dataset, it checks the metadata of each piece of data. If the "Do-Not-Learn" flag is set to "True," the data is excluded from the learning process. This prevents specific creative works from being used for learning without permission.

[0048] Measuring and reporting similarity

[0049] A user uploads newly generated content to the server. The server receives the content and matches it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and notified to the user or administrator. The report includes the identity and similarity of the matching creative works.

[0050] Specific examples

[0051] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, saves it as metadata with a flag "Do-Not-Learn: True," and then uses this metadata to generate a QR code, which is then embedded in the novel.

[0052] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0053] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0054] The system of the present invention makes it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0055] The processing flow will be explained below.

[0056] Step 1:

[0057] Users upload their own creations to the server, and can also indicate their preference that they do not want their creations to be used for study purposes.

[0058] Step 2:

[0059] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0060] Step 3:

[0061] The server generates identification information and adds a flag "Do-Not-Learn: True" to the creative work as metadata based on the user's request that the work not be used for learning purposes.

[0062] Step 4:

[0063] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0064] Step 5:

[0065] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0066] Step 6:

[0067] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0068] Step 7:

[0069] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0070] Step 8:

[0071] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0072] Step 9:

[0073] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0074] Step 10:

[0075] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0076] Step 11:

[0077] The server generates a report containing the identification information of highly similar creative works and the results of the similarity, and notifies the user or administrator. This notification enables a prompt response when there is a risk of copyright infringement.

[0078] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works and protect copyrights.

[0079] Example 1

[0080] 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."

[0081] There is a need to prevent copyright infringement by preventing the unauthorized use of creative works in training datasets. Furthermore, a system is needed to quickly report information when newly generated content by a generative AI model is similar to existing creative works. However, current systems are sometimes insufficient to protect creative works or detect unauthorized use.

[0082] 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.

[0083] In this invention, the server includes means for receiving creative works, means for generating identification information for the received creative works using the SHA-256 hash algorithm and creating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded dataset and excluding creative works for which the "Do-Not-Learn" flag is true, means for measuring the similarity of the generated new content with creative works in an existing database using TF-IDF vectorization, and means for reporting the identification information and similarity of highly similar creative works, thereby preventing unauthorized use of creative works and enabling rapid detection of copyright infringement.

[0084] "Creative works" refers to all works such as documents, images, audio, and video created by users themselves.

[0085] The "SHA-256 hash algorithm" is a cryptographic hash function that converts input data into a fixed-length 256-bit hash value.

[0086] "Identifying Information" means a hash value or other unique data generated to uniquely identify a Creative Work.

[0087] "Metadata" is auxiliary data containing information about the creative work, such as identifying information and "Do-Not-Learn" flags.

[0088] A "QR code" is a type of two-dimensional barcode, a data representation format for visually encoding information.

[0089] A "URL" is a string of characters used to specify resources on the Internet, and stands for Uniform Resource Locator.

[0090] A "dataset" refers to a collection of multiple pieces of data that an AI model uses for training.

[0091] A "Do-Not-Learn flag" is a flag set in metadata to prevent a creative work from being used to train AI models.

[0092] "TF-IDF vectorization" is a method for extracting important words from a document and converting them into feature vectors.

[0093] "Cosine similarity" is a method for measuring similarity based on the angle between two vectors, with the closer it is to 1, the more similar it is.

[0094] "Report" means a report containing the results of the similarity measurement and the identification of the matching creative works.

[0095] "System" refers to a set of hardware and software configurations that have functions such as receiving creative works, generating identification information, managing metadata, generating QR codes, and detecting fraud.

[0096] This invention relates to a system that protects copyrights of creative works and prevents their unauthorized use by generative AI models. This system allows users to upload their creative works to a server, where they generate and manage identifying information as metadata. The server then generates a QR code or URL based on the metadata and embeds it in the creative work to protect it. When the AI ​​model learns a new dataset, it can refer to this metadata and exclude creative works with specific flags from the learning process. Furthermore, the system has the ability to measure and report the similarity of newly generated content to existing creative works.

[0097] Upload your creations and generate metadata

[0098] First, a user uploads their creation to the server. The server generates an identifier for the received creation using the SHA-256 hash algorithm. This identifier is saved as metadata for the creation. Furthermore, if the user indicates that they do not want the creation to be used for learning, the server adds a flag "Do-Not-Learn: True" to the metadata.

[0099] Generate and embed a QR code or URL

[0100] The server generates a QR code or URL based on the generated metadata. The generated QR code or URL is used to provide access to the metadata of the creative work. For example, by creating a QR code using a QR code generation library and embedding it in a creative work, users can easily access it using a smartphone or other device.

[0101] Processing the training dataset

[0102] When a device (specifically an AI model) loads a new dataset, it checks the metadata of each piece of data. If it contains any creative work with a "Do-Not-Learn" flag set to true, it will exclude that data from the learning process. This prevents certain creative work from being used in learning without permission.

[0103] Measuring and reporting similarity

[0104] When a user uploads newly generated content to the server, the server receives the content and matches it with an existing database of creative works. During this process, the server converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works.

[0105] Specific examples

[0106] As a concrete example, consider the case where a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then uses this metadata to generate a QR code and embeds it in the novel.

[0107] Next, when the AI ​​model trains on a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0108] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0109] Prompt Sentence Examples

[0110] Examples of prompts for generative AI models include:

[0111] "Measure the similarity between newly uploaded sentences and sentences in the existing database, and generate a report if the similarity is 70% or higher."

[0112] "Use the SHA-256 hashing algorithm to generate an identifier for the uploaded file, include it in the metadata, and generate a QR code."

[0113] This will prevent unauthorized use of creative works and enable rapid detection of copyright infringement.

[0114] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0115] Step 1: Upload your creation

[0116] The user uploads his / her creation to the server. As input, the creation file selected by the user is required. Specifically, the user uses the server's upload form, selects the file, and clicks the upload button. As output, the server saves the received creation file and prepares it for the next processing step.

[0117] Step 2: Generate hash values ​​and create metadata

[0118] The server generates identification information by running the received creative file through the SHA-256 hash algorithm. The saved creative file data is required as input. Specifically, the server calls the "hash value generation module" to calculate the hash value of the creative work. As output, the generated hash value is included in the metadata. At this time, if the user specifies that they do not want the file to be used for learning, the "Do-Not-Learn: True" flag is added to the metadata.

[0119] Step 3: Generate a QR code or URL

[0120] The server generates a QR code or a URL based on the generated metadata. The metadata (identification information and "Do-Not-Learn" flag) is required as input. Specifically, the server generates a QR code using a "QR code generation module" and saves the QR code as image data using a file operation library. The generated QR code or URL is obtained as output.

[0121] Step 4: Embed a QR code or URL

[0122] The server embeds the generated QR code or URL into the creative work. The input required is the file data of the creative work and the generated QR code or URL. Specifically, the server uses a file manipulation library to embed the QR code image into the cover of the creative work, etc. The output is a creative work file with the QR code or URL embedded.

[0123] Step 5: Processing the training data set

[0124] The terminal (specifically, the AI ​​model) checks the metadata of each data when loading a new dataset. As input, it requires the new dataset and its metadata. Specifically, the terminal runs a "metadata analysis module" and checks the "Do-Not-Learn" flag in the metadata. As output, data whose "Do-Not-Learn" flag is true is filtered out and the data is not used in the learning process.

[0125] Step 6: Measure and report similarity

[0126] When a user uploads new content to the server, the server compares it with the existing creative database. The inputs required are the new content and the existing creative database. Specifically, the server converts the new content into vector format using TF-IDF vectorization and measures the similarity using a "cosine similarity calculation module." The output is the identification information and similarity of highly similar content.

[0127] Step 7: Generate and notify the similarity report

[0128] The server generates a report based on the measured similarity and notifies the user or administrator. As input, the server requires the similarity measurement results and an existing database of creative works. Specifically, the server uses a "report generation module" to create a report that includes the identification information and similarity of highly similar creative works. As output, the report is sent to the user or administrator, who is notified of its contents.

[0129] Each step in the system prevents unauthorized use of creative works and allows for rapid detection of copyright infringement.

[0130] (Application example 1)

[0131] 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."

[0132] In recent years, the development of generative AI models has made it easier to use and copy creative works without permission. This has made copyright protection of creative works more difficult and increased the risk of copyright infringement. To solve this problem, a system is needed that can effectively protect the copyright of creative works and prevent unauthorized use and copying.

[0133] 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.

[0134] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, means for reporting the identification information and similarity of highly similar creative works, and means for notifying users based on the highly similar creative works, thereby making it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0135] A "creative work" is a work that has a form or content of expression that is protected by copyright law.

[0136] "Identification information" is information that uniquely identifies a particular creative work.

[0137] "Metadata" is supplementary information about a creative work, including identifying information and specific flags.

[0138] A "QR code" is a two-dimensional barcode and a means of visually representing data.

[0139] A "URL" is an address that indicates a resource on the Internet.

[0140] A "dataset" is a collection of data that is the subject of learning or analysis.

[0141] A "flag" is a piece of metadata that indicates a particular condition or state.

[0142] "Similarity" is a numerical representation of the similarity between two pieces of content.

[0143] A "notification" is the act of informing a user about a particular event or condition.

[0144] A "server" is a computer system that provides specific services and manages data on a network.

[0145] "User" means any individual or organization that wishes to use this system to manage or protect their creative works.

[0146] The present invention provides a system for protecting creative works and preventing unauthorized use, which includes the following processes:

[0147] First, a user uploads their creation to the server. The server generates a unique identifier for the received creation using the SHA-256 hash algorithm and includes it in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a flag called "Do-Not-Learn: True" to the metadata. This metadata includes the identifier and flag.

[0148] The server then generates a QR code or a URL based on the generated metadata and embeds it in the creative work, which provides access to the metadata about the creative work.

[0149] Additionally, when a device (e.g., a computer system containing an AI model) loads a new dataset, it checks the metadata of each piece of data and, if it contains any creative work with a "Do-Not-Learn" flag set to "True," the data is excluded from the learning process.

[0150] When a user uploads newly generated content to the server, the server receives the content and compares it with the existing database of creative works. It converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, it generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works. It also notifies users based on the most similar creative works.

[0151] The hardware mainly consists of servers and terminals connected to a network. The software uses various Python libraries (e.g., hashlib, qrcode, scikit-learn, json). These libraries perform processes such as generating identification information, generating QR codes, converting text to vectors, and calculating cosine similarity.

[0152] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, generates metadata including a flag "Do-Not-Learn: True", and then generates a QR code based on this metadata and embeds it in the novel.

[0153] Also, the following prompt sentence will be used as an input example.

[0154] Hash the following text content and convert it as a TF-IDF vector: "This is new text."

[0155] Then calculate the cosine similarity with the existing content database and display the content in descending order of similarity to the existing content in the database.

[0156] In this way, unauthorized use of creative works can be prevented and copyright infringement can be detected quickly.

[0157] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0158] Step 1:

[0159] Users upload their creations. Users upload their creations (e.g., text, images, videos) to the server, along with basic information about the creation (title, author, description, etc.).

[0160] Input: User-uploaded creative works and basic information about them.

[0161] Output: Creations and basic information stored on the server.

[0162] Specific operation: When the user selects a file on the application's upload screen and presses the "Upload" button, the creation is sent to the server.

[0163] Step 2:

[0164] The server generates the identifier and metadata. The server generates a unique identifier for the received creative work using the SHA-256 hash algorithm. It also generates metadata containing the "Do-Not-Learn: True" flag based on the user's preferences.

[0165] Input: Uploaded creative work and user preference (whether or not to use it for learning).

[0166] Output: Generated identification information and metadata.

[0167] Specifically, the server inputs the content of the creative work into the SHA-256 algorithm, obtains a hash value (identification information) as the output, and then generates metadata that combines the identification information and flags.

[0168] Step 3:

[0169] The server generates a QR code or a URL and embeds it in the creative work.The server generates a QR code or a URL based on the metadata and embeds it in the creative work.

[0170] Input: Generated metadata.

[0171] Output: Your creation with an embedded QR code or URL.

[0172] Specific operation: The server uses the qrcode library to generate a QR code that visualizes the metadata, places it in the appropriate location on the work, and saves the generated QR code as a file.

[0173] Step 4:

[0174] When the device loads a new dataset, it checks the metadata. Before adding a new dataset to the learning process, the device checks the metadata of each data and excludes data with the "Do-Not-Learn" flag.

[0175] Input: A new dataset.

[0176] Output: Data excluded from the training process.

[0177] What it does: When the device loads a dataset, it analyzes the metadata of each piece of data and adds flagged data to an exclusion list.

[0178] Step 5:

[0179] A user uploads new content and the server measures the similarity. When a user uploads new content to the server, the server measures the similarity with existing creations in the database. TF-IDF vectorization and cosine similarity are used to calculate the similarity between the new content and existing content.

[0180] Input: New content.

[0181] Output: Similarity measurement results.

[0182] Specific operation: The server performs TF-IDF vectorization and calculates the cosine similarity between the resulting vectors. For content with high similarity, the server notifies the user of the results in a report.

[0183] Step 6:

[0184] The server notifies the user of the similarity measurement results. If the similarity exceeds a certain threshold, the server notifies the user of the content's identification information and the similarity.

[0185] Input: Similarity measurement results.

[0186] Output: Notification to the user.

[0187] Specific operation: The server analyzes the similarity measurement results, and if the threshold is exceeded, an email or in-app notification is sent to the user.

[0188] In this way, the system of the present invention takes specific steps to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0189] 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.

[0190] This invention combines an emotion engine with a system that strengthens copyright protection for creative works and prevents unauthorized use by AI. This system recognizes users' emotions and provides feedback based on uploading creative works and judging their similarity. Below, we will generate a program for this system and explain its processing in natural language.

[0191] Overall system overview

[0192] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities. By combining this process with an emotion engine, it is possible to provide feedback and warnings that reflect the user's emotions.

[0193] Program processing

[0194] Upload your creations and generate metadata

[0195] Users upload their creations to the server. The server generates identification information for the received creations using the SHA-256 hash algorithm and includes this information in the metadata. If the user indicates that they do not want their creation to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata. Furthermore, the server uses an emotion engine to analyze the user's emotions at the time of uploading and records the results in the metadata.

[0196] Generate a QR code or URL

[0197] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0198] Processing the training dataset

[0199] When a device administrator (including an AI model) loads a new dataset, the device retrieves the metadata for each piece of data. It checks the metadata for each piece of data, and if the "Do-Not-Learn" flag is set to "True," it excludes that data from the learning process. This prevents specific creative works from being improperly learned by the AI ​​model.

[0200] Measuring and reporting similarity

[0201] Users upload newly created content to the server. The server receives the content and compares it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and the user or administrator is notified. An emotion engine is built into this notification process, providing feedback based on the user's emotions.

[0202] Specific examples

[0203] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then analyzes the user's emotions using an emotion engine and records the result (e.g., "joy") in the metadata. The server then uses this metadata to generate a QR code and embeds it in the novel.

[0204] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0205] Finally, when a user uploads a newly generated sentence, the server compares it with the existing novel database and notifies the user if there is a high match. At this time, the emotion engine evaluates the user's current emotion and provides appropriate feedback (e.g., "The generated sentence is 85% consistent. Please be careful.").

[0206] As described above, the system of the present invention can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide appropriate feedback according to the user's emotions.

[0207] The processing flow will be explained below.

[0208] Step 1:

[0209] Users upload their creations to the server. At this time, they can also indicate their preference that they do not want their creations to be used for learning. The emotion engine then analyzes the user's emotions in real time.

[0210] Step 2:

[0211] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0212] Step 3:

[0213] The server generates identification information and adds a flag "Do-Not-Learn: True" to the metadata based on the user's preference not to have it used for learning.

[0214] Step 4:

[0215] The server checks the analysis results of the emotion engine and records the user's emotions (e.g., "happiness," "anxiety," "anger," etc.) in metadata, allowing us to track emotional information related to creative works.

[0216] Step 5:

[0217] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0218] Step 6:

[0219] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0220] Step 7:

[0221] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0222] Step 8:

[0223] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0224] Step 9:

[0225] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0226] Step 10:

[0227] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0228] Step 11:

[0229] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0230] Step 12:

[0231] The server generates a report of the similarity scores and the identification information of the most similar creations, and notifies the user or administrator. This notification process incorporates an emotion engine, which provides feedback based on the user's emotions.

[0232] Step 13:

[0233] The server uses an emotion engine to provide appropriate feedback based on the user's emotions. For example, it may notify the user that "The generated text is 85% consistent with existing novel A. The user's emotions are uneasy, so we recommend further confirmation."

[0234] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works, protect copyrights, and provide appropriate feedback according to the user's emotions.

[0235] Example 2

[0236] 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."

[0237] Systems that focus on protecting the copyright of creative works and preventing their unauthorized use generally lack the ability to control widespread use. They also often lack feedback functions that take user sentiment into account. Therefore, there is a need for a system that provides feedback based on user sentiment while strengthening copyright protection and preventing unauthorized use by AI.

[0238] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving a creative work, means for assigning unique identification information to the creative work and generating metadata, means for analyzing user emotions and recording the results in the metadata, means for generating a QR code or a URL based on the generated metadata, means for embedding the generated QR code or URL in the creative work, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for providing feedback based on the user's emotions when reporting the identification information and similarity of highly similar creative works. This makes it possible to provide feedback according to the user's emotions while preventing unauthorized use of creative works and reducing the risk of copyright infringement.

[0239] A "creative work" is an original expression or work that may take various forms, such as text, images, audio, or video.

[0240] "Server" refers to a computer system capable of receiving, processing, storing, and transmitting data over a network.

[0241] "User" refers to a person who uses the System to upload and manage creative works.

[0242] "Identifying Information" means a hash value or other information generated to uniquely identify a particular creative work.

[0243] "Metadata" is attribute information related to a creative work, and includes identification information, user emotions, and information regarding usage restrictions.

[0244] A "QR code" is a representation of metadata in the form of a two-dimensional barcode that provides access to information related to a creative work.

[0245] "URL" means a Uniform Resource Locator that provides direct access to metadata and information related to a creative work.

[0246] A "flag" refers to true / false information that is set in metadata to indicate a specific condition or state.

[0247] "Emotion engine" refers to a software module that analyzes user input data (e.g., comments and options) and evaluates their emotions.

[0248] A "dataset" refers to a collection of data that has been collected and organized for the purpose of learning or analysis.

[0249] "TF-IDF" refers to a technology that vectorizes text data based on the document frequency and inverse document frequency of words.

[0250] "Cosine similarity" is a mathematical method for measuring the similarity between vectors, and expresses the similarity using the cosine value of an angle.

[0251] "Feedback" refers to the reactions and information provided to the user by the system, and in this case refers to the information based on the user's emotions using the emotion engine.

[0252] This invention is a system that strengthens copyright protection for creative works and prevents unauthorized use by generative AI models. This system combines multiple means, including a server, a terminal, and a user, to assign unique identification information to creative works and provide feedback based on the user's emotions using an emotion engine.

[0253] Upload your creations and generate metadata

[0254] Users upload their creations to the server through an interface. Uploaded data can be in various formats, including text, images, and audio. The interface can be a browser or a dedicated application. For each uploaded creation, the server generates identification information (hash value) using the SHA-256 hash algorithm and stores this information as metadata.

[0255] Specific behavior:

[0256] The server calculates a SHA-256 hash of the uploaded creative work and stores the result as metadata.

[0257] If the user does not want the data to be used for learning, the flag "Do-Not-Learn: True" is added to the metadata.

[0258] User sentiment analysis

[0259] The server uses an emotion engine to analyze the user's emotions when uploading a creative work, and the emotion engine evaluates the emotions based on the user's input (e.g., text comments and options) and records the results in the metadata.

[0260] Specific behavior:

[0261] The server passes the comments and options entered by the user to the emotion engine.

[0262] The emotion engine analyzes the input data and stores emotions (e.g., "joy" or "anxiety") as metadata.

[0263] QR code and URL generation

[0264] The server generates a QR code based on the generated metadata and embeds it in the creative work or generates a URL associated with the creative work, allowing access to the metadata about the creative work via the QR code or URL.

[0265] Specific behavior:

[0266] The server passes the metadata to a QR code generator to generate a QR code image.

[0267] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[0268] Processing the training dataset

[0269] When an administrator of a device (the hardware that runs an AI model) loads a new dataset, the device checks the metadata of each piece of data and, if it contains any creative content flagged as "Do-Not-Learn: True," it excludes that data from the learning process.

[0270] Specific behavior:

[0271] The device accesses the metadata of each creation in the loaded dataset and checks the flag.

[0272] Exclude creations with the flag set to "True" from the AI ​​model's training set.

[0273] Similarity measurement and feedback

[0274] Users upload newly created content to the server. The server then uses TF-IDF vectorization to measure cosine similarity to match the content with existing creative content databases. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. This report includes feedback based on the user's emotions generated by the emotion engine.

[0275] Specific behavior:

[0276] The server vectorizes the newly uploaded content using the TF-IDF algorithm.

[0277] The cosine similarity is calculated and the similarity with the creations in the existing database is checked.

[0278] If the similarity exceeds a threshold, a notification report is generated and sent to the user.

[0279] The notification content is customized according to the user's emotions analyzed by the emotion engine.

[0280] Specific examples

[0281] For example, consider a scenario where a user uploads a novel they have written to a server. The server generates an identifier for the novel using a SHA-256 hash and stores this information as metadata. It also adds a "Do-Not-Learn: True" flag to the metadata and records the user's emotional response (e.g., "Delight") as analyzed by the emotion engine. The server then uses this metadata to generate a QR code and embeds it in the novel. Later, when the AI ​​model trains a new dataset, it checks the metadata for each piece of data loaded and excludes any data with the "True" flag from training.

[0282] Prompt Sentence Examples

[0283] Some examples of prompts that can be input to a generative AI model include:

[0284] "Compare the content of your new novel. Calculate the similarity with the existing database and report the results."

[0285] or

[0286] "Please upload this image and exclude it from AI training data. Add the reason 'Do not want to use for training' to the metadata."

[0287] As a result, this system can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide feedback that reflects the user's emotions.

[0288] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0289] Processing flow

[0290] Step 1: Upload your creation

[0291] Input: The user selects a creative file and uploads it through the interface.

[0292] Specific behavior:

[0293] The user selects the creative file and clicks the "Upload" button.

[0294] The server receives the uploaded data.

[0295] Output: Creative files are saved to the server.

[0296] Step 2: Generate Metadata

[0297] Input: Uploaded creative data.

[0298] Specific behavior:

[0299] The server generates a hash value of the received creative work using the SHA-256 hashing algorithm.

[0300] The identification information (hash value) is saved as metadata, and if the user does not want it to be used for learning, "Do-Not-Learn: True" is added to the metadata.

[0301] Output: Metadata is generated and includes hash values ​​and flags.

[0302] Step 3: Conducting sentiment analysis

[0303] Input: User input data (comments and choices).

[0304] Specific behavior:

[0305] The server passes the user's input data to the emotion engine for emotion analysis.

[0306] The emotion engine analyzes the input data and identifies emotions (e.g., "joy" or "anxiety").

[0307] The identified emotion is recorded in the metadata.

[0308] Output: The sentiment result is added to the metadata.

[0309] Step 4: Generate a QR code or URL

[0310] Input: The generated metadata.

[0311] Specific behavior:

[0312] The server passes the metadata to a QR code generator to generate a QR code.

[0313] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[0314] Output: A QR code or URL is created and embedded in the creation.

[0315] Step 5: Processing the training data set

[0316] Input: The newly loaded dataset.

[0317] Specific behavior:

[0318] The terminal obtains metadata for each creation in the data set.

[0319] Check the "Do-Not-Learn" flag in the metadata and exclude those that are "True" from the learning process.

[0320] Output: A list of creative works that were excluded from the training dataset.

[0321] Step 6: Measure and report similarity

[0322] Input: The newly generated content.

[0323] Specific behavior:

[0324] A user uploads new content to the server.

[0325] The server converts the new content into vector format using TF-IDF vectorization and matches it with an existing creative database.

[0326] It calculates the cosine similarity and generates a report if the similarity exceeds a certain threshold.

[0327] Use the emotion engine to add feedback to reports based on user emotions.

[0328] Output: A report containing information on highly similar creative works is generated and notified to the user or administrator.

[0329] The above are the specific processing steps of the program of this system, which effectively achieves the protection of creative works and the prevention of unauthorized use, while providing feedback based on the user's emotions.

[0330] (Application example 2)

[0331] 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."

[0332] Conventional copyright protection systems lacked sufficient functionality to prevent the unauthorized use of creative works, and they faced challenges in preventing AI from learning copyrighted works without permission and generating similar content. Furthermore, they lacked functionality to provide feedback based on user emotions, which resulted in a lack of improvement in the user experience.

[0333] 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.

[0334] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for reporting the identification information and similarity of highly similar creative works and providing feedback according to the user's emotions using an emotion engine. This makes it possible to effectively prevent unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[0335] "Creative works" are original works that should be protected by copyright, such as text, music, video, and art.

[0336] "Identifying Information" is a string of data, such as a SHA-256 hash, used to uniquely identify a creative work.

[0337] "Metadata" is additional information related to a creative work, including data such as "Do-Not-Learn" flags and user sentiment.

[0338] A "QR code" is a two-dimensional barcode that can visually embed information related to a creative work.

[0339] "URL" stands for Uniform Resource Locator, an internet address for accessing the metadata of a creative work.

[0340] A "dataset" is a collection of data used to train an AI model.

[0341] A "flag" is a data field for indicating a specific condition, and in this case has a value such as "Do-Not-Learn: True."

[0342] "Similarity" is an index that indicates the degree of correspondence between new content and existing creations.

[0343] "Emotion Engine" is an AI technology that analyzes and evaluates users' emotions in real time.

[0344] "Feedback" is information or notifications provided to the user by the system, based on the user's actions and emotions.

[0345] Specific embodiments for carrying out the present invention are described below.

[0346] Overall system configuration

[0347] The system receives creative works, assigns identifying information and generates metadata, then generates QR codes and URLs based on the metadata and embeds them in the creative works. It also checks the metadata of newly loaded datasets and excludes creative works with specific flags from learning. It also measures the similarity of the generated new content with creative works in the existing database, and provides feedback to the user if the similarity is high. This feedback is provided using an emotion engine.

[0348] Hardware and Software Used

[0349] Cloud-based servers: Use AWS® or Google® Cloud.

[0350] Emotion engine: Uses Microsoft(R) Azure(R) Emotion API.

[0351] Database: Data is managed using MySQL (registered trademark) or PostgreSQL.

[0352] QR code generation library: Uses the Python qrcode library.

[0353] Similarity calculation: Uses the Python sklearn library.

[0354] Program processing overview

[0355] Upload your creations and generate metadata

[0356] Users upload their creations to the server using a smartphone app, which generates a unique identifier using the SHA-256 algorithm. The server then analyzes the user's emotions with an AI emotion engine and adds a "Do-Not-Learn" flag and emotional information as metadata to the generated identifier.

[0357] Generate a QR code or URL

[0358] The server generates a QR code or a URL from the generated metadata and embeds it in the creation.

[0359] Processing the training dataset

[0360] When a dataset is loaded, the server checks the metadata of each piece of data. If the "Do-Not-Learn: True" flag is set, the creation is excluded from learning.

[0361] Measuring and reporting similarity

[0362] When newly created content is uploaded, the server matches it with existing creations in the database using TF-IDF vectorization and cosine similarity calculation. If the similarity is high, the server sends a notification to the user with emotional feedback.

[0363] Examples of specific examples and prompts

[0364] Examples:

[0365] 1. When a user uploads a short story they have written, the server generates a SHA-256 hash of it and adds a "Do-Not-Learn: True" flag and sentiment information (e.g., "Satisfied") to the metadata.

[0366] 2. Generate a QR code from the metadata and embed it in the novel file.

[0367] 3. When another user uploads a sentence, the server compares the sentence with the existing database and finds that the similarity is 85%, so it sends the user emotional feedback saying, "Your sentence is 85% similar. Please reconsider it from a different perspective."

[0368] Example prompt sentence:

[0369] Input text: "A user has uploaded a poem of their own. Compare it to a database of existing poems."

[0370] system:

[0371] 1. Calculate the SHA-256 hash value of the input text.

[0372] 2. Add "Do-Not-Learn" flag and user sentiment to metadata.

[0373] 3. TF-IDF vectorization and calculation of cosine similarity.

[0374] 4. Generate feedback based on user emotions.

[0375] This invention is a system that can effectively prevent the unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[0376] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0377] Step 1:

[0378] Users upload their creations to the server using a smartphone app. The file of the creation they upload becomes the input. The server receives this file.

[0379] Step 2:

[0380] The server generates an identifier for the received creative work using the SHA-256 hash algorithm, with the identifier being the input and the generated hash value being the output.

[0381] Step 3:

[0382] The server saves the generated hash value as metadata. At the same time, if the user indicates that they do not want their data to be used for learning, a flag "Do-Not-Learn: True" is added to the metadata. Furthermore, an AI emotion engine is used to analyze the user's emotions at the time of upload and the results are recorded in the metadata. The emotion engine is the input, and the emotion result is the output.

[0383] Step 4:

[0384] The server generates a QR code or a URL based on the generated metadata. This metadata is the input, and the generated QR code or URL is the output.

[0385] Step 5:

[0386] The server embeds the generated QR code or URL into the creative work. At this time, the creative work file and the QR code or URL are input, and the creative work file with the embedded QR code or URL is output.

[0387] Step 6:

[0388] When a device administrator loads a new dataset, the server checks the metadata of each data set. The new dataset is the input, and the verification result is the output. If the flag "Do-Not-Learn: True" is present, the creation is excluded from learning.

[0389] Step 7:

[0390] Users upload newly generated content to the server, which then matches this content with a database of existing creations.

[0391] Step 8:

[0392] The server measures the similarity between the new content and existing creative works by performing TF-IDF vectorization and cosine similarity calculation. The new content is the input, and the similarity calculation result is the output.

[0393] Step 9:

[0394] If the similarity exceeds a certain threshold, the server generates a report and notifies the user or administrator. The notification process incorporates an emotion engine, which provides feedback based on the user's emotions. The similarity calculation result is the input, and the feedback notification is the output.

[0395] 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.

[0396] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0397] 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.

[0398] [Second embodiment]

[0399] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0400] 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.

[0401] 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).

[0402] 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.

[0403] 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.

[0404] 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).

[0405] 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.

[0406] 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.

[0407] 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.

[0408] 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.

[0409] In the smart glasses 214, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0410] 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."

[0411] This invention relates to a system that protects the copyright of creative works and prevents their unauthorized use by AI. Below, we will generate a program for this system and explain its processing in natural language.

[0412] Overall system overview

[0413] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities.

[0414] Program processing

[0415] Upload your creations and generate metadata

[0416] Users upload their creations to the server, which then generates identification information for the received creations using the SHA-256 hash algorithm and includes that information in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata.

[0417] Generate a QR code or URL

[0418] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0419] Processing the training dataset

[0420] When a device (including an AI model) loads a new dataset, it checks the metadata of each piece of data. If the "Do-Not-Learn" flag is set to "True," the data is excluded from the learning process. This prevents specific creative works from being used for learning without permission.

[0421] Measuring and reporting similarity

[0422] A user uploads newly generated content to the server. The server receives the content and matches it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and notified to the user or administrator. The report includes the identity and similarity of the matching creative works.

[0423] Specific examples

[0424] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, saves it as metadata with a flag "Do-Not-Learn: True," and then uses this metadata to generate a QR code, which is then embedded in the novel.

[0425] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0426] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0427] The system of the present invention makes it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0428] The processing flow will be explained below.

[0429] Step 1:

[0430] Users upload their own creations to the server, and can also indicate their preference that they do not want their creations to be used for study purposes.

[0431] Step 2:

[0432] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0433] Step 3:

[0434] The server generates identification information and adds a flag "Do-Not-Learn: True" to the creative work as metadata based on the user's request that the work not be used for learning purposes.

[0435] Step 4:

[0436] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0437] Step 5:

[0438] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0439] Step 6:

[0440] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0441] Step 7:

[0442] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0443] Step 8:

[0444] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0445] Step 9:

[0446] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0447] Step 10:

[0448] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0449] Step 11:

[0450] The server generates a report containing the identification information of highly similar creative works and the results of the similarity, and notifies the user or administrator. This notification enables a prompt response when there is a risk of copyright infringement.

[0451] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works and protect copyrights.

[0452] Example 1

[0453] 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."

[0454] There is a need to prevent copyright infringement by preventing the unauthorized use of creative works in training datasets. Furthermore, a system is needed to quickly report information when newly generated content by a generative AI model is similar to existing creative works. However, current systems are sometimes insufficient to protect creative works or detect unauthorized use.

[0455] 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.

[0456] In this invention, the server includes means for receiving creative works, means for generating identification information for the received creative works using the SHA-256 hash algorithm and creating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded dataset and excluding creative works for which the "Do-Not-Learn" flag is true, means for measuring the similarity of the generated new content with creative works in an existing database using TF-IDF vectorization, and means for reporting the identification information and similarity of highly similar creative works, thereby preventing unauthorized use of creative works and enabling rapid detection of copyright infringement.

[0457] "Creative works" refers to all works such as documents, images, audio, and video created by users themselves.

[0458] The "SHA-256 hash algorithm" is a cryptographic hash function that converts input data into a fixed-length 256-bit hash value.

[0459] "Identifying Information" means a hash value or other unique data generated to uniquely identify a Creative Work.

[0460] "Metadata" is auxiliary data containing information about the creative work, such as identifying information and "Do-Not-Learn" flags.

[0461] A "QR code" is a type of two-dimensional barcode, a data representation format for visually encoding information.

[0462] A "URL" is a string of characters used to specify resources on the Internet, and stands for Uniform Resource Locator.

[0463] A "dataset" refers to a collection of multiple pieces of data that an AI model uses for training.

[0464] A "Do-Not-Learn flag" is a flag set in metadata to prevent a creative work from being used to train AI models.

[0465] "TF-IDF vectorization" is a method for extracting important words from a document and converting them into feature vectors.

[0466] "Cosine similarity" is a method for measuring similarity based on the angle between two vectors, with the closer it is to 1, the more similar it is.

[0467] "Report" means a report containing the results of the similarity measurement and the identification of the matching creative works.

[0468] "System" refers to a set of hardware and software configurations that have functions such as receiving creative works, generating identification information, managing metadata, generating QR codes, and detecting fraud.

[0469] This invention relates to a system that protects copyrights of creative works and prevents their unauthorized use by generative AI models. This system allows users to upload their creative works to a server, where they generate and manage identifying information as metadata. The server then generates a QR code or URL based on the metadata and embeds it in the creative work to protect it. When the AI ​​model learns a new dataset, it can refer to this metadata and exclude creative works with specific flags from the learning process. Furthermore, the system has the ability to measure and report the similarity of newly generated content to existing creative works.

[0470] Upload your creations and generate metadata

[0471] First, a user uploads their creation to the server. The server generates an identifier for the received creation using the SHA-256 hash algorithm. This identifier is saved as metadata for the creation. Furthermore, if the user indicates that they do not want the creation to be used for learning, the server adds a flag "Do-Not-Learn: True" to the metadata.

[0472] Generate and embed a QR code or URL

[0473] The server generates a QR code or URL based on the generated metadata. The generated QR code or URL is used to provide access to the metadata of the creative work. For example, by creating a QR code using a QR code generation library and embedding it in a creative work, users can easily access it using a smartphone or other device.

[0474] Processing the training dataset

[0475] When a device (specifically an AI model) loads a new dataset, it checks the metadata of each piece of data. If it contains any creative work with a "Do-Not-Learn" flag set to true, it will exclude that data from the learning process. This prevents certain creative work from being used in learning without permission.

[0476] Measuring and reporting similarity

[0477] When a user uploads newly generated content to the server, the server receives the content and matches it with an existing database of creative works. During this process, the server converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works.

[0478] Specific examples

[0479] As a concrete example, consider the case where a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then uses this metadata to generate a QR code and embeds it in the novel.

[0480] Next, when the AI ​​model trains on a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0481] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0482] Prompt Sentence Examples

[0483] Examples of prompts for generative AI models include:

[0484] "Measure the similarity between newly uploaded sentences and sentences in the existing database, and generate a report if the similarity is 70% or higher."

[0485] "Use the SHA-256 hashing algorithm to generate an identifier for the uploaded file, include it in the metadata, and generate a QR code."

[0486] This will prevent unauthorized use of creative works and enable rapid detection of copyright infringement.

[0487] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0488] Step 1: Upload your creation

[0489] The user uploads his / her creation to the server. As input, the creation file selected by the user is required. Specifically, the user uses the server's upload form, selects the file, and clicks the upload button. As output, the server saves the received creation file and prepares it for the next processing step.

[0490] Step 2: Generate hash values ​​and create metadata

[0491] The server generates identification information by running the received creative file through the SHA-256 hash algorithm. The saved creative file data is required as input. Specifically, the server calls the "hash value generation module" to calculate the hash value of the creative work. As output, the generated hash value is included in the metadata. At this time, if the user specifies that they do not want the file to be used for learning, the "Do-Not-Learn: True" flag is added to the metadata.

[0492] Step 3: Generate a QR code or URL

[0493] The server generates a QR code or a URL based on the generated metadata. The metadata (identification information and "Do-Not-Learn" flag) is required as input. Specifically, the server generates a QR code using a "QR code generation module" and saves the QR code as image data using a file operation library. The generated QR code or URL is obtained as output.

[0494] Step 4: Embed a QR code or URL

[0495] The server embeds the generated QR code or URL into the creative work. The input required is the file data of the creative work and the generated QR code or URL. Specifically, the server uses a file manipulation library to embed the QR code image into the cover of the creative work, etc. The output is a creative work file with the QR code or URL embedded.

[0496] Step 5: Processing the training data set

[0497] The terminal (specifically, the AI ​​model) checks the metadata of each data when loading a new dataset. As input, it requires the new dataset and its metadata. Specifically, the terminal runs a "metadata analysis module" and checks the "Do-Not-Learn" flag in the metadata. As output, data whose "Do-Not-Learn" flag is true is filtered out and the data is not used in the learning process.

[0498] Step 6: Measure and report similarity

[0499] When a user uploads new content to the server, the server compares it with the existing creative database. The inputs required are the new content and the existing creative database. Specifically, the server converts the new content into vector format using TF-IDF vectorization and measures the similarity using a "cosine similarity calculation module." The output is the identification information and similarity of highly similar content.

[0500] Step 7: Generate and notify the similarity report

[0501] The server generates a report based on the measured similarity and notifies the user or administrator. As input, the server requires the similarity measurement results and an existing database of creative works. Specifically, the server uses a "report generation module" to create a report that includes the identification information and similarity of highly similar creative works. As output, the report is sent to the user or administrator, who is notified of its contents.

[0502] Each step in the system prevents unauthorized use of creative works and allows for rapid detection of copyright infringement.

[0503] (Application example 1)

[0504] 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."

[0505] In recent years, the development of generative AI models has made it easier to use and copy creative works without permission. This has made copyright protection of creative works more difficult and increased the risk of copyright infringement. To solve this problem, a system is needed that can effectively protect the copyright of creative works and prevent unauthorized use and copying.

[0506] 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.

[0507] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, means for reporting the identification information and similarity of highly similar creative works, and means for notifying users based on the highly similar creative works, thereby making it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0508] A "creative work" is a work that has a form or content of expression that is protected by copyright law.

[0509] "Identification information" is information that uniquely identifies a particular creative work.

[0510] "Metadata" is supplementary information about a creative work, including identifying information and specific flags.

[0511] A "QR code" is a two-dimensional barcode and a means of visually representing data.

[0512] A "URL" is an address that indicates a resource on the Internet.

[0513] A "dataset" is a collection of data that is the subject of learning or analysis.

[0514] A "flag" is a piece of metadata that indicates a particular condition or state.

[0515] "Similarity" is a numerical representation of the similarity between two pieces of content.

[0516] A "notification" is the act of informing a user about a particular event or condition.

[0517] A "server" is a computer system that provides specific services and manages data on a network.

[0518] "User" means any individual or organization that wishes to use this system to manage or protect their creative works.

[0519] The present invention provides a system for protecting creative works and preventing unauthorized use, which includes the following processes:

[0520] First, a user uploads their creation to the server. The server generates a unique identifier for the received creation using the SHA-256 hash algorithm and includes it in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a flag called "Do-Not-Learn: True" to the metadata. This metadata includes the identifier and flag.

[0521] The server then generates a QR code or a URL based on the generated metadata and embeds it in the creative work, which provides access to the metadata about the creative work.

[0522] Additionally, when a device (e.g., a computer system containing an AI model) loads a new dataset, it checks the metadata of each piece of data and, if it contains any creative work with a "Do-Not-Learn" flag set to "True," the data is excluded from the learning process.

[0523] When a user uploads newly generated content to the server, the server receives the content and compares it with the existing database of creative works. It converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, it generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works. It also notifies users based on the most similar creative works.

[0524] The hardware mainly consists of servers and terminals connected to a network. The software uses various Python libraries (e.g., hashlib, qrcode, scikit-learn, json). These libraries perform processes such as generating identification information, generating QR codes, converting text to vectors, and calculating cosine similarity.

[0525] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, generates metadata including a flag "Do-Not-Learn: True", and then generates a QR code based on this metadata and embeds it in the novel.

[0526] Also, the following prompt sentence will be used as an input example.

[0527] Hash the following text content and convert it as a TF-IDF vector: "This is new text."

[0528] Then calculate the cosine similarity with the existing content database and display the content in descending order of similarity to the existing content in the database.

[0529] In this way, unauthorized use of creative works can be prevented and copyright infringement can be detected quickly.

[0530] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0531] Step 1:

[0532] Users upload their creations. Users upload their creations (e.g., text, images, videos) to the server, along with basic information about the creation (title, author, description, etc.).

[0533] Input: User-uploaded creative works and basic information about them.

[0534] Output: Creations and basic information stored on the server.

[0535] Specific operation: When the user selects a file on the application's upload screen and presses the "Upload" button, the creation is sent to the server.

[0536] Step 2:

[0537] The server generates the identifier and metadata. The server generates a unique identifier for the received creative work using the SHA-256 hash algorithm. It also generates metadata containing the "Do-Not-Learn: True" flag based on the user's preferences.

[0538] Input: Uploaded creative work and user preference (whether or not to use it for learning).

[0539] Output: Generated identification information and metadata.

[0540] Specifically, the server inputs the content of the creative work into the SHA-256 algorithm, obtains a hash value (identification information) as the output, and then generates metadata that combines the identification information and flags.

[0541] Step 3:

[0542] The server generates a QR code or a URL and embeds it in the creative work.The server generates a QR code or a URL based on the metadata and embeds it in the creative work.

[0543] Input: Generated metadata.

[0544] Output: Your creation with an embedded QR code or URL.

[0545] Specific operation: The server uses the qrcode library to generate a QR code that visualizes the metadata, places it in the appropriate location on the work, and saves the generated QR code as a file.

[0546] Step 4:

[0547] When the device loads a new dataset, it checks the metadata. Before adding a new dataset to the learning process, the device checks the metadata of each data and excludes data with the "Do-Not-Learn" flag.

[0548] Input: A new dataset.

[0549] Output: Data excluded from the training process.

[0550] What it does: When the device loads a dataset, it analyzes the metadata of each piece of data and adds flagged data to an exclusion list.

[0551] Step 5:

[0552] A user uploads new content and the server measures the similarity. When a user uploads new content to the server, the server measures the similarity with existing creations in the database. TF-IDF vectorization and cosine similarity are used to calculate the similarity between the new content and existing content.

[0553] Input: New content.

[0554] Output: Similarity measurement results.

[0555] Specific operation: The server performs TF-IDF vectorization and calculates the cosine similarity between the resulting vectors. For content with high similarity, the server notifies the user of the results in a report.

[0556] Step 6:

[0557] The server notifies the user of the similarity measurement results. If the similarity exceeds a certain threshold, the server notifies the user of the content's identification information and the similarity.

[0558] Input: Similarity measurement results.

[0559] Output: Notification to the user.

[0560] Specific operation: The server analyzes the similarity measurement results, and if the threshold is exceeded, an email or in-app notification is sent to the user.

[0561] In this way, the system of the present invention takes specific steps to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0562] 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.

[0563] This invention combines an emotion engine with a system that strengthens copyright protection for creative works and prevents unauthorized use by AI. This system recognizes users' emotions and provides feedback based on uploading creative works and judging their similarity. Below, we will generate a program for this system and explain its processing in natural language.

[0564] Overall system overview

[0565] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities. By combining this process with an emotion engine, it is possible to provide feedback and warnings that reflect the user's emotions.

[0566] Program processing

[0567] Upload your creations and generate metadata

[0568] Users upload their creations to the server. The server generates identification information for the received creations using the SHA-256 hash algorithm and includes this information in the metadata. If the user indicates that they do not want their creation to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata. Furthermore, the server uses an emotion engine to analyze the user's emotions at the time of uploading and records the results in the metadata.

[0569] Generate a QR code or URL

[0570] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0571] Processing the training dataset

[0572] When a device administrator (including an AI model) loads a new dataset, the device retrieves the metadata for each piece of data. It checks the metadata for each piece of data, and if the "Do-Not-Learn" flag is set to "True," it excludes that data from the learning process. This prevents specific creative works from being improperly learned by the AI ​​model.

[0573] Measuring and reporting similarity

[0574] Users upload newly created content to the server. The server receives the content and compares it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and the user or administrator is notified. An emotion engine is built into this notification process, providing feedback based on the user's emotions.

[0575] Specific examples

[0576] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then analyzes the user's emotions using an emotion engine and records the result (e.g., "joy") in the metadata. The server then uses this metadata to generate a QR code and embeds it in the novel.

[0577] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0578] Finally, when a user uploads a newly generated sentence, the server compares it with the existing novel database and notifies the user if there is a high match. At this time, the emotion engine evaluates the user's current emotion and provides appropriate feedback (e.g., "The generated sentence is 85% consistent. Please be careful.").

[0579] As described above, the system of the present invention can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide appropriate feedback according to the user's emotions.

[0580] The processing flow will be explained below.

[0581] Step 1:

[0582] Users upload their creations to the server. At this time, they can also indicate their preference that they do not want their creations to be used for learning. The emotion engine then analyzes the user's emotions in real time.

[0583] Step 2:

[0584] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0585] Step 3:

[0586] The server generates identification information and adds a flag "Do-Not-Learn: True" to the metadata based on the user's preference not to have it used for learning.

[0587] Step 4:

[0588] The server checks the analysis results of the emotion engine and records the user's emotions (e.g., "happiness," "anxiety," "anger," etc.) in metadata, allowing us to track emotional information related to creative works.

[0589] Step 5:

[0590] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0591] Step 6:

[0592] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0593] Step 7:

[0594] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0595] Step 8:

[0596] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0597] Step 9:

[0598] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0599] Step 10:

[0600] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0601] Step 11:

[0602] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0603] Step 12:

[0604] The server generates a report of the similarity scores and the identification information of the most similar creations, and notifies the user or administrator. This notification process incorporates an emotion engine, which provides feedback based on the user's emotions.

[0605] Step 13:

[0606] The server uses an emotion engine to provide appropriate feedback based on the user's emotions. For example, it may notify the user that "The generated text is 85% consistent with existing novel A. The user's emotions are uneasy, so we recommend further confirmation."

[0607] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works, protect copyrights, and provide appropriate feedback according to the user's emotions.

[0608] Example 2

[0609] 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."

[0610] Systems that focus on protecting the copyright of creative works and preventing their unauthorized use generally lack the ability to control widespread use. They also often lack feedback functions that take user sentiment into account. Therefore, there is a need for a system that provides feedback based on user sentiment while strengthening copyright protection and preventing unauthorized use by AI.

[0611] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving a creative work, means for assigning unique identification information to the creative work and generating metadata, means for analyzing user emotions and recording the results in the metadata, means for generating a QR code or a URL based on the generated metadata, means for embedding the generated QR code or URL in the creative work, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for providing feedback based on the user's emotions when reporting the identification information and similarity of highly similar creative works. This makes it possible to provide feedback according to the user's emotions while preventing unauthorized use of creative works and reducing the risk of copyright infringement.

[0612] A "creative work" is an original expression or work that may take various forms, such as text, images, audio, or video.

[0613] "Server" refers to a computer system capable of receiving, processing, storing, and transmitting data over a network.

[0614] "User" refers to a person who uses the System to upload and manage creative works.

[0615] "Identifying Information" means a hash value or other information generated to uniquely identify a particular creative work.

[0616] "Metadata" is attribute information related to a creative work, and includes identification information, user emotions, and information regarding usage restrictions.

[0617] A "QR code" is a representation of metadata in the form of a two-dimensional barcode that provides access to information related to a creative work.

[0618] "URL" means a Uniform Resource Locator that provides direct access to metadata and information related to a creative work.

[0619] A "flag" refers to true / false information that is set in metadata to indicate a specific condition or state.

[0620] "Emotion engine" refers to a software module that analyzes user input data (e.g., comments and options) and evaluates their emotions.

[0621] A "dataset" refers to a collection of data that has been collected and organized for the purpose of learning or analysis.

[0622] "TF-IDF" refers to a technology that vectorizes text data based on the document frequency and inverse document frequency of words.

[0623] "Cosine similarity" is a mathematical method for measuring the similarity between vectors, and expresses the similarity using the cosine value of an angle.

[0624] "Feedback" refers to the reactions and information provided to the user by the system, and in this case refers to the information based on the user's emotions using the emotion engine.

[0625] This invention is a system that strengthens copyright protection for creative works and prevents unauthorized use by generative AI models. This system combines multiple means, including a server, a terminal, and a user, to assign unique identification information to creative works and provide feedback based on the user's emotions using an emotion engine.

[0626] Upload your creations and generate metadata

[0627] Users upload their creations to the server through an interface. Uploaded data can be in various formats, including text, images, and audio. The interface can be a browser or a dedicated application. For each uploaded creation, the server generates identification information (hash value) using the SHA-256 hash algorithm and stores this information as metadata.

[0628] Specific behavior:

[0629] The server calculates a SHA-256 hash of the uploaded creative work and stores the result as metadata.

[0630] If the user does not want the data to be used for learning, the flag "Do-Not-Learn: True" is added to the metadata.

[0631] User sentiment analysis

[0632] The server uses an emotion engine to analyze the user's emotions when uploading a creative work, and the emotion engine evaluates the emotions based on the user's input (e.g., text comments and options) and records the results in the metadata.

[0633] Specific behavior:

[0634] The server passes the comments and options entered by the user to the emotion engine.

[0635] The emotion engine analyzes the input data and stores emotions (e.g., "joy" or "anxiety") as metadata.

[0636] QR code and URL generation

[0637] The server generates a QR code based on the generated metadata and embeds it in the creative work or generates a URL associated with the creative work, allowing access to the metadata about the creative work via the QR code or URL.

[0638] Specific behavior:

[0639] The server passes the metadata to a QR code generator to generate a QR code image.

[0640] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[0641] Processing the training dataset

[0642] When an administrator of a device (the hardware that runs an AI model) loads a new dataset, the device checks the metadata of each piece of data and, if it contains any creative content flagged as "Do-Not-Learn: True," it excludes that data from the learning process.

[0643] Specific behavior:

[0644] The device accesses the metadata of each creation in the loaded dataset and checks the flag.

[0645] Exclude creations with the flag set to "True" from the AI ​​model's training set.

[0646] Similarity measurement and feedback

[0647] Users upload newly created content to the server. The server then uses TF-IDF vectorization to measure cosine similarity to match the content with existing creative content databases. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. This report includes feedback based on the user's emotions generated by the emotion engine.

[0648] Specific behavior:

[0649] The server vectorizes the newly uploaded content using the TF-IDF algorithm.

[0650] The cosine similarity is calculated and the similarity with the creations in the existing database is checked.

[0651] If the similarity exceeds a threshold, a notification report is generated and sent to the user.

[0652] The notification content is customized according to the user's emotions analyzed by the emotion engine.

[0653] Specific examples

[0654] For example, consider a scenario where a user uploads a novel they have written to a server. The server generates an identifier for the novel using a SHA-256 hash and stores this information as metadata. It also adds a "Do-Not-Learn: True" flag to the metadata and records the user's emotional response (e.g., "Delight") as analyzed by the emotion engine. The server then uses this metadata to generate a QR code and embeds it in the novel. Later, when the AI ​​model trains a new dataset, it checks the metadata for each piece of data loaded and excludes any data with the "True" flag from training.

[0655] Prompt Sentence Examples

[0656] Some examples of prompts that can be input to a generative AI model include:

[0657] "Compare the content of your new novel. Calculate the similarity with the existing database and report the results."

[0658] or

[0659] "Please upload this image and exclude it from AI training data. Add the reason 'Do not want to use for training' to the metadata."

[0660] As a result, this system can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide feedback that reflects the user's emotions.

[0661] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0662] Processing flow

[0663] Step 1: Upload your creation

[0664] Input: The user selects a creative file and uploads it through the interface.

[0665] Specific behavior:

[0666] The user selects the creative file and clicks the "Upload" button.

[0667] The server receives the uploaded data.

[0668] Output: Creative files are saved to the server.

[0669] Step 2: Generate Metadata

[0670] Input: Uploaded creative data.

[0671] Specific behavior:

[0672] The server generates a hash value of the received creative work using the SHA-256 hashing algorithm.

[0673] The identification information (hash value) is saved as metadata, and if the user does not want it to be used for learning, "Do-Not-Learn: True" is added to the metadata.

[0674] Output: Metadata is generated and includes hash values ​​and flags.

[0675] Step 3: Conducting sentiment analysis

[0676] Input: User input data (comments and choices).

[0677] Specific behavior:

[0678] The server passes the user's input data to the emotion engine for emotion analysis.

[0679] The emotion engine analyzes the input data and identifies emotions (e.g., "joy" or "anxiety").

[0680] The identified emotion is recorded in the metadata.

[0681] Output: The sentiment result is added to the metadata.

[0682] Step 4: Generate a QR code or URL

[0683] Input: The generated metadata.

[0684] Specific behavior:

[0685] The server passes the metadata to a QR code generator to generate a QR code.

[0686] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[0687] Output: A QR code or URL is created and embedded in the creation.

[0688] Step 5: Processing the training data set

[0689] Input: The newly loaded dataset.

[0690] Specific behavior:

[0691] The terminal obtains metadata for each creation in the data set.

[0692] Check the "Do-Not-Learn" flag in the metadata and exclude those that are "True" from the learning process.

[0693] Output: A list of creative works that were excluded from the training dataset.

[0694] Step 6: Measure and report similarity

[0695] Input: The newly generated content.

[0696] Specific behavior:

[0697] A user uploads new content to the server.

[0698] The server converts the new content into vector format using TF-IDF vectorization and matches it with an existing creative database.

[0699] It calculates the cosine similarity and generates a report if the similarity exceeds a certain threshold.

[0700] Use the emotion engine to add feedback to reports based on user emotions.

[0701] Output: A report containing information on highly similar creative works is generated and notified to the user or administrator.

[0702] The above are the specific processing steps of the program of this system, which effectively achieves the protection of creative works and the prevention of unauthorized use, while providing feedback based on the user's emotions.

[0703] (Application example 2)

[0704] 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."

[0705] Conventional copyright protection systems lacked sufficient functionality to prevent the unauthorized use of creative works, and they faced challenges in preventing AI from learning copyrighted works without permission and generating similar content. Furthermore, they lacked functionality to provide feedback based on user emotions, which resulted in a lack of improvement in the user experience.

[0706] 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.

[0707] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for reporting the identification information and similarity of highly similar creative works and providing feedback according to the user's emotions using an emotion engine. This makes it possible to effectively prevent unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[0708] "Creative works" are original works that should be protected by copyright, such as text, music, video, and art.

[0709] "Identifying Information" is a string of data, such as a SHA-256 hash, used to uniquely identify a creative work.

[0710] "Metadata" is additional information related to a creative work, including data such as "Do-Not-Learn" flags and user sentiment.

[0711] A "QR code" is a two-dimensional barcode that can visually embed information related to a creative work.

[0712] "URL" stands for Uniform Resource Locator, an internet address for accessing the metadata of a creative work.

[0713] A "dataset" is a collection of data used to train an AI model.

[0714] A "flag" is a data field for indicating a specific condition, and in this case has a value such as "Do-Not-Learn: True."

[0715] "Similarity" is an index that indicates the degree of correspondence between new content and existing creations.

[0716] "Emotion Engine" is an AI technology that analyzes and evaluates users' emotions in real time.

[0717] "Feedback" is information or notifications provided to the user by the system, based on the user's actions and emotions.

[0718] Specific embodiments for carrying out the present invention are described below.

[0719] Overall system configuration

[0720] The system receives creative works, assigns identifying information and generates metadata, then generates QR codes and URLs based on the metadata and embeds them in the creative works. It also checks the metadata of newly loaded datasets and excludes creative works with specific flags from learning. It also measures the similarity of the generated new content with creative works in the existing database, and provides feedback to the user if the similarity is high. This feedback is provided using an emotion engine.

[0721] Hardware and Software Used

[0722] Cloud-based servers: Use AWS or Google Cloud.

[0723] Emotion engine: Uses Microsoft Azure Emotion API.

[0724] Database: Manage data with MySQL or PostgreSQL.

[0725] QR code generation library: Uses the Python qrcode library.

[0726] Similarity calculation: Uses the Python sklearn library.

[0727] Program processing overview

[0728] Upload your creations and generate metadata

[0729] Users upload their creations to the server using a smartphone app, which generates a unique identifier using the SHA-256 algorithm. The server then analyzes the user's emotions with an AI emotion engine and adds a "Do-Not-Learn" flag and emotional information as metadata to the generated identifier.

[0730] Generate a QR code or URL

[0731] The server generates a QR code or a URL from the generated metadata and embeds it in the creation.

[0732] Processing the training dataset

[0733] When a dataset is loaded, the server checks the metadata of each piece of data. If the "Do-Not-Learn: True" flag is set, the creation is excluded from learning.

[0734] Measuring and reporting similarity

[0735] When newly created content is uploaded, the server matches it with existing creations in the database using TF-IDF vectorization and cosine similarity calculation. If the similarity is high, the server sends a notification to the user with emotional feedback.

[0736] Examples of specific examples and prompts

[0737] Examples:

[0738] 1. When a user uploads a short story they have written, the server generates a SHA-256 hash of it and adds a "Do-Not-Learn: True" flag and sentiment information (e.g., "Satisfied") to the metadata.

[0739] 2. Generate a QR code from the metadata and embed it in the novel file.

[0740] 3. When another user uploads a sentence, the server compares the sentence with the existing database and finds that the similarity is 85%, so it sends the user emotional feedback saying, "Your sentence is 85% similar. Please reconsider it from a different perspective."

[0741] Example prompt sentence:

[0742] Input text: "A user has uploaded a poem of their own. Compare it to a database of existing poems."

[0743] system:

[0744] 1. Calculate the SHA-256 hash value of the input text.

[0745] 2. Add "Do-Not-Learn" flag and user sentiment to metadata.

[0746] 3. TF-IDF vectorization and calculation of cosine similarity.

[0747] 4. Generate feedback based on user emotions.

[0748] This invention is a system that can effectively prevent the unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[0749] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0750] Step 1:

[0751] Users upload their creations to the server using a smartphone app. The file of the creation they upload becomes the input. The server receives this file.

[0752] Step 2:

[0753] The server generates an identifier for the received creative work using the SHA-256 hash algorithm, with the identifier being the input and the generated hash value being the output.

[0754] Step 3:

[0755] The server saves the generated hash value as metadata. At the same time, if the user indicates that they do not want their data to be used for learning, a flag "Do-Not-Learn: True" is added to the metadata. Furthermore, an AI emotion engine is used to analyze the user's emotions at the time of upload and the results are recorded in the metadata. The emotion engine is the input, and the emotion result is the output.

[0756] Step 4:

[0757] The server generates a QR code or a URL based on the generated metadata. This metadata is the input, and the generated QR code or URL is the output.

[0758] Step 5:

[0759] The server embeds the generated QR code or URL into the creative work. At this time, the creative work file and the QR code or URL are input, and the creative work file with the embedded QR code or URL is output.

[0760] Step 6:

[0761] When a device administrator loads a new dataset, the server checks the metadata of each data set. The new dataset is the input, and the verification result is the output. If the flag "Do-Not-Learn: True" is present, the creation is excluded from learning.

[0762] Step 7:

[0763] Users upload newly generated content to the server, which then matches this content with a database of existing creations.

[0764] Step 8:

[0765] The server measures the similarity between the new content and existing creative works by performing TF-IDF vectorization and cosine similarity calculation. The new content is the input, and the similarity calculation result is the output.

[0766] Step 9:

[0767] If the similarity exceeds a certain threshold, the server generates a report and notifies the user or administrator. The notification process incorporates an emotion engine, which provides feedback based on the user's emotions. The similarity calculation result is the input, and the feedback notification is the output.

[0768] 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.

[0769] 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.

[0770] 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.

[0771] [Third embodiment]

[0772] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0773] 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.

[0774] 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).

[0775] 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.

[0776] 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.

[0777] 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).

[0778] 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.

[0779] 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.

[0780] 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.

[0781] 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.

[0782] 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.

[0783] 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."

[0784] This invention relates to a system that protects the copyright of creative works and prevents their unauthorized use by AI. Below, we will generate a program for this system and explain its processing in natural language.

[0785] Overall system overview

[0786] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities.

[0787] Program processing

[0788] Upload your creations and generate metadata

[0789] Users upload their creations to the server, which then generates identification information for the received creations using the SHA-256 hash algorithm and includes that information in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata.

[0790] Generate a QR code or URL

[0791] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0792] Processing the training dataset

[0793] When a device (including an AI model) loads a new dataset, it checks the metadata of each piece of data. If the "Do-Not-Learn" flag is set to "True," the data is excluded from the learning process. This prevents specific creative works from being used for learning without permission.

[0794] Measuring and reporting similarity

[0795] A user uploads newly generated content to the server. The server receives the content and matches it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and notified to the user or administrator. The report includes the identity and similarity of the matching creative works.

[0796] Specific examples

[0797] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, saves it as metadata with a flag "Do-Not-Learn: True," and then uses this metadata to generate a QR code, which is then embedded in the novel.

[0798] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0799] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0800] The system of the present invention makes it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0801] The processing flow will be explained below.

[0802] Step 1:

[0803] Users upload their own creations to the server, and can also indicate their preference that they do not want their creations to be used for study purposes.

[0804] Step 2:

[0805] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0806] Step 3:

[0807] The server generates identification information and adds a flag "Do-Not-Learn: True" to the creative work as metadata based on the user's request that the work not be used for learning purposes.

[0808] Step 4:

[0809] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0810] Step 5:

[0811] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0812] Step 6:

[0813] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0814] Step 7:

[0815] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0816] Step 8:

[0817] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0818] Step 9:

[0819] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0820] Step 10:

[0821] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0822] Step 11:

[0823] The server generates a report containing the identification information of highly similar creative works and the results of the similarity, and notifies the user or administrator. This notification enables a prompt response when there is a risk of copyright infringement.

[0824] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works and protect copyrights.

[0825] Example 1

[0826] 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."

[0827] There is a need to prevent copyright infringement by preventing the unauthorized use of creative works in training datasets. Furthermore, a system is needed to quickly report information when newly generated content by a generative AI model is similar to existing creative works. However, current systems are sometimes insufficient to protect creative works or detect unauthorized use.

[0828] 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.

[0829] In this invention, the server includes means for receiving creative works, means for generating identification information for the received creative works using the SHA-256 hash algorithm and creating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded dataset and excluding creative works for which the "Do-Not-Learn" flag is true, means for measuring the similarity of the generated new content with creative works in an existing database using TF-IDF vectorization, and means for reporting the identification information and similarity of highly similar creative works, thereby preventing unauthorized use of creative works and enabling rapid detection of copyright infringement.

[0830] "Creative works" refers to all works such as documents, images, audio, and video created by users themselves.

[0831] The "SHA-256 hash algorithm" is a cryptographic hash function that converts input data into a fixed-length 256-bit hash value.

[0832] "Identifying Information" means a hash value or other unique data generated to uniquely identify a Creative Work.

[0833] "Metadata" is auxiliary data containing information about the creative work, such as identifying information and "Do-Not-Learn" flags.

[0834] A "QR code" is a type of two-dimensional barcode, a data representation format for visually encoding information.

[0835] A "URL" is a string of characters used to specify resources on the Internet, and stands for Uniform Resource Locator.

[0836] A "dataset" refers to a collection of multiple pieces of data that an AI model uses for training.

[0837] A "Do-Not-Learn flag" is a flag set in metadata to prevent a creative work from being used to train AI models.

[0838] "TF-IDF vectorization" is a method for extracting important words from a document and converting them into feature vectors.

[0839] "Cosine similarity" is a method for measuring similarity based on the angle between two vectors, with the closer it is to 1, the more similar it is.

[0840] "Report" means a report containing the results of the similarity measurement and the identification of the matching creative works.

[0841] "System" refers to a set of hardware and software configurations that have functions such as receiving creative works, generating identification information, managing metadata, generating QR codes, and detecting fraud.

[0842] This invention relates to a system that protects copyrights of creative works and prevents their unauthorized use by generative AI models. This system allows users to upload their creative works to a server, where they generate and manage identifying information as metadata. The server then generates a QR code or URL based on the metadata and embeds it in the creative work to protect it. When the AI ​​model learns a new dataset, it can refer to this metadata and exclude creative works with specific flags from the learning process. Furthermore, the system has the ability to measure and report the similarity of newly generated content to existing creative works.

[0843] Upload your creations and generate metadata

[0844] First, a user uploads their creation to the server. The server generates an identifier for the received creation using the SHA-256 hash algorithm. This identifier is saved as metadata for the creation. Furthermore, if the user indicates that they do not want the creation to be used for learning, the server adds a flag "Do-Not-Learn: True" to the metadata.

[0845] Generate and embed a QR code or URL

[0846] The server generates a QR code or URL based on the generated metadata. The generated QR code or URL is used to provide access to the metadata of the creative work. For example, by creating a QR code using a QR code generation library and embedding it in a creative work, users can easily access it using a smartphone or other device.

[0847] Processing the training dataset

[0848] When a device (specifically an AI model) loads a new dataset, it checks the metadata of each piece of data. If it contains any creative work with a "Do-Not-Learn" flag set to true, it will exclude that data from the learning process. This prevents certain creative work from being used in learning without permission.

[0849] Measuring and reporting similarity

[0850] When a user uploads newly generated content to the server, the server receives the content and matches it with an existing database of creative works. During this process, the server converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works.

[0851] Specific examples

[0852] As a concrete example, consider the case where a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then uses this metadata to generate a QR code and embeds it in the novel.

[0853] Next, when the AI ​​model trains on a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0854] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[0855] Prompt Sentence Examples

[0856] Examples of prompts for generative AI models include:

[0857] "Measure the similarity between newly uploaded sentences and sentences in the existing database, and generate a report if the similarity is 70% or higher."

[0858] "Use the SHA-256 hashing algorithm to generate an identifier for the uploaded file, include it in the metadata, and generate a QR code."

[0859] This will prevent unauthorized use of creative works and enable rapid detection of copyright infringement.

[0860] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0861] Step 1: Upload your creation

[0862] The user uploads his / her creation to the server. As input, the creation file selected by the user is required. Specifically, the user uses the server's upload form, selects the file, and clicks the upload button. As output, the server saves the received creation file and prepares it for the next processing step.

[0863] Step 2: Generate hash values ​​and create metadata

[0864] The server generates identification information by running the received creative file through the SHA-256 hash algorithm. The saved creative file data is required as input. Specifically, the server calls the "hash value generation module" to calculate the hash value of the creative work. As output, the generated hash value is included in the metadata. At this time, if the user specifies that they do not want the file to be used for learning, the "Do-Not-Learn: True" flag is added to the metadata.

[0865] Step 3: Generate a QR code or URL

[0866] The server generates a QR code or a URL based on the generated metadata. The metadata (identification information and "Do-Not-Learn" flag) is required as input. Specifically, the server generates a QR code using a "QR code generation module" and saves the QR code as image data using a file operation library. The generated QR code or URL is obtained as output.

[0867] Step 4: Embed a QR code or URL

[0868] The server embeds the generated QR code or URL into the creative work. The input required is the file data of the creative work and the generated QR code or URL. Specifically, the server uses a file manipulation library to embed the QR code image into the cover of the creative work, etc. The output is a creative work file with the QR code or URL embedded.

[0869] Step 5: Processing the training data set

[0870] The terminal (specifically, the AI ​​model) checks the metadata of each data when loading a new dataset. As input, it requires the new dataset and its metadata. Specifically, the terminal runs a "metadata analysis module" and checks the "Do-Not-Learn" flag in the metadata. As output, data whose "Do-Not-Learn" flag is true is filtered out and the data is not used in the learning process.

[0871] Step 6: Measure and report similarity

[0872] When a user uploads new content to the server, the server compares it with the existing creative database. The inputs required are the new content and the existing creative database. Specifically, the server converts the new content into vector format using TF-IDF vectorization and measures the similarity using a "cosine similarity calculation module." The output is the identification information and similarity of highly similar content.

[0873] Step 7: Generate and notify the similarity report

[0874] The server generates a report based on the measured similarity and notifies the user or administrator. As input, the server requires the similarity measurement results and an existing database of creative works. Specifically, the server uses a "report generation module" to create a report that includes the identification information and similarity of highly similar creative works. As output, the report is sent to the user or administrator, who is notified of its contents.

[0875] Each step in the system prevents unauthorized use of creative works and allows for rapid detection of copyright infringement.

[0876] (Application example 1)

[0877] 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."

[0878] In recent years, the development of generative AI models has made it easier to use and copy creative works without permission. This has made copyright protection of creative works more difficult and increased the risk of copyright infringement. To solve this problem, a system is needed that can effectively protect the copyright of creative works and prevent unauthorized use and copying.

[0879] 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.

[0880] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, means for reporting the identification information and similarity of highly similar creative works, and means for notifying users based on the highly similar creative works, thereby making it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0881] A "creative work" is a work that has a form or content of expression that is protected by copyright law.

[0882] "Identification information" is information that uniquely identifies a particular creative work.

[0883] "Metadata" is supplementary information about a creative work, including identifying information and specific flags.

[0884] A "QR code" is a two-dimensional barcode and a means of visually representing data.

[0885] A "URL" is an address that indicates a resource on the Internet.

[0886] A "dataset" is a collection of data that is the subject of learning or analysis.

[0887] A "flag" is a piece of metadata that indicates a particular condition or state.

[0888] "Similarity" is a numerical representation of the similarity between two pieces of content.

[0889] A "notification" is the act of informing a user about a particular event or condition.

[0890] A "server" is a computer system that provides specific services and manages data on a network.

[0891] "User" means any individual or organization that wishes to use this system to manage or protect their creative works.

[0892] The present invention provides a system for protecting creative works and preventing unauthorized use, which includes the following processes:

[0893] First, a user uploads their creation to the server. The server generates a unique identifier for the received creation using the SHA-256 hash algorithm and includes it in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a flag called "Do-Not-Learn: True" to the metadata. This metadata includes the identifier and flag.

[0894] The server then generates a QR code or a URL based on the generated metadata and embeds it in the creative work, which provides access to the metadata about the creative work.

[0895] Additionally, when a device (e.g., a computer system containing an AI model) loads a new dataset, it checks the metadata of each piece of data and, if it contains any creative work with a "Do-Not-Learn" flag set to "True," the data is excluded from the learning process.

[0896] When a user uploads newly generated content to the server, the server receives the content and compares it with the existing database of creative works. It converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, it generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works. It also notifies users based on the most similar creative works.

[0897] The hardware mainly consists of servers and terminals connected to a network. The software uses various Python libraries (e.g., hashlib, qrcode, scikit-learn, json). These libraries perform processes such as generating identification information, generating QR codes, converting text to vectors, and calculating cosine similarity.

[0898] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, generates metadata including a flag "Do-Not-Learn: True", and then generates a QR code based on this metadata and embeds it in the novel.

[0899] Also, the following prompt sentence will be used as an input example.

[0900] Hash the following text content and convert it as a TF-IDF vector: "This is new text."

[0901] Then calculate the cosine similarity with the existing content database and display the content in descending order of similarity to the existing content in the database.

[0902] In this way, unauthorized use of creative works can be prevented and copyright infringement can be detected quickly.

[0903] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0904] Step 1:

[0905] Users upload their creations. Users upload their creations (e.g., text, images, videos) to the server, along with basic information about the creation (title, author, description, etc.).

[0906] Input: User-uploaded creative works and basic information about them.

[0907] Output: Creations and basic information stored on the server.

[0908] Specific operation: When the user selects a file on the application's upload screen and presses the "Upload" button, the creation is sent to the server.

[0909] Step 2:

[0910] The server generates the identifier and metadata. The server generates a unique identifier for the received creative work using the SHA-256 hash algorithm. It also generates metadata containing the "Do-Not-Learn: True" flag based on the user's preferences.

[0911] Input: Uploaded creative work and user preference (whether or not to use it for learning).

[0912] Output: Generated identification information and metadata.

[0913] Specifically, the server inputs the content of the creative work into the SHA-256 algorithm, obtains a hash value (identification information) as the output, and then generates metadata that combines the identification information and flags.

[0914] Step 3:

[0915] The server generates a QR code or a URL and embeds it in the creative work.The server generates a QR code or a URL based on the metadata and embeds it in the creative work.

[0916] Input: Generated metadata.

[0917] Output: Your creation with an embedded QR code or URL.

[0918] Specific operation: The server uses the qrcode library to generate a QR code that visualizes the metadata, places it in the appropriate location on the work, and saves the generated QR code as a file.

[0919] Step 4:

[0920] When the device loads a new dataset, it checks the metadata. Before adding a new dataset to the learning process, the device checks the metadata of each data and excludes data with the "Do-Not-Learn" flag.

[0921] Input: A new dataset.

[0922] Output: Data excluded from the training process.

[0923] What it does: When the device loads a dataset, it analyzes the metadata of each piece of data and adds flagged data to an exclusion list.

[0924] Step 5:

[0925] A user uploads new content and the server measures the similarity. When a user uploads new content to the server, the server measures the similarity with existing creations in the database. TF-IDF vectorization and cosine similarity are used to calculate the similarity between the new content and existing content.

[0926] Input: New content.

[0927] Output: Similarity measurement results.

[0928] Specific operation: The server performs TF-IDF vectorization and calculates the cosine similarity between the resulting vectors. For content with high similarity, the server notifies the user of the results in a report.

[0929] Step 6:

[0930] The server notifies the user of the similarity measurement results. If the similarity exceeds a certain threshold, the server notifies the user of the content's identification information and the similarity.

[0931] Input: Similarity measurement results.

[0932] Output: Notification to the user.

[0933] Specific operation: The server analyzes the similarity measurement results, and if the threshold is exceeded, an email or in-app notification is sent to the user.

[0934] In this way, the system of the present invention takes specific steps to prevent unauthorized use of creative works and quickly detect copyright infringement.

[0935] 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.

[0936] This invention combines an emotion engine with a system that strengthens copyright protection for creative works and prevents unauthorized use by AI. This system recognizes users' emotions and provides feedback based on uploading creative works and judging their similarity. Below, we will generate a program for this system and explain its processing in natural language.

[0937] Overall system overview

[0938] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities. By combining this process with an emotion engine, it is possible to provide feedback and warnings that reflect the user's emotions.

[0939] Program processing

[0940] Upload your creations and generate metadata

[0941] Users upload their creations to the server. The server generates identification information for the received creations using the SHA-256 hash algorithm and includes this information in the metadata. If the user indicates that they do not want their creation to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata. Furthermore, the server uses an emotion engine to analyze the user's emotions at the time of uploading and records the results in the metadata.

[0942] Generate a QR code or URL

[0943] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[0944] Processing the training dataset

[0945] When a device administrator (including an AI model) loads a new dataset, the device retrieves the metadata for each piece of data. It checks the metadata for each piece of data, and if the "Do-Not-Learn" flag is set to "True," it excludes that data from the learning process. This prevents specific creative works from being improperly learned by the AI ​​model.

[0946] Measuring and reporting similarity

[0947] Users upload newly created content to the server. The server receives the content and compares it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and the user or administrator is notified. An emotion engine is built into this notification process, providing feedback based on the user's emotions.

[0948] Specific examples

[0949] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then analyzes the user's emotions using an emotion engine and records the result (e.g., "joy") in the metadata. The server then uses this metadata to generate a QR code and embeds it in the novel.

[0950] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[0951] Finally, when a user uploads a newly generated sentence, the server compares it with the existing novel database and notifies the user if there is a high match. At this time, the emotion engine evaluates the user's current emotion and provides appropriate feedback (e.g., "The generated sentence is 85% consistent. Please be careful.").

[0952] As described above, the system of the present invention can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide appropriate feedback according to the user's emotions.

[0953] The processing flow will be explained below.

[0954] Step 1:

[0955] Users upload their creations to the server. At this time, they can also indicate their preference that they do not want their creations to be used for learning. The emotion engine then analyzes the user's emotions in real time.

[0956] Step 2:

[0957] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[0958] Step 3:

[0959] The server generates identification information and adds a flag "Do-Not-Learn: True" to the metadata based on the user's preference not to have it used for learning.

[0960] Step 4:

[0961] The server checks the analysis results of the emotion engine and records the user's emotions (e.g., "happiness," "anxiety," "anger," etc.) in metadata, allowing us to track emotional information related to creative works.

[0962] Step 5:

[0963] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[0964] Step 6:

[0965] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[0966] Step 7:

[0967] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[0968] Step 8:

[0969] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[0970] Step 9:

[0971] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[0972] Step 10:

[0973] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[0974] Step 11:

[0975] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[0976] Step 12:

[0977] The server generates a report of the similarity scores and the identification information of the most similar creations, and notifies the user or administrator. This notification process incorporates an emotion engine, which provides feedback based on the user's emotions.

[0978] Step 13:

[0979] The server uses an emotion engine to provide appropriate feedback based on the user's emotions. For example, it may notify the user that "The generated text is 85% consistent with existing novel A. The user's emotions are uneasy, so we recommend further confirmation."

[0980] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works, protect copyrights, and provide appropriate feedback according to the user's emotions.

[0981] Example 2

[0982] 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."

[0983] Systems that focus on protecting the copyright of creative works and preventing their unauthorized use generally lack the ability to control widespread use. They also often lack feedback functions that take user sentiment into account. Therefore, there is a need for a system that provides feedback based on user sentiment while strengthening copyright protection and preventing unauthorized use by AI.

[0984] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving a creative work, means for assigning unique identification information to the creative work and generating metadata, means for analyzing user emotions and recording the results in the metadata, means for generating a QR code or a URL based on the generated metadata, means for embedding the generated QR code or URL in the creative work, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for providing feedback based on the user's emotions when reporting the identification information and similarity of highly similar creative works. This makes it possible to provide feedback according to the user's emotions while preventing unauthorized use of creative works and reducing the risk of copyright infringement.

[0985] A "creative work" is an original expression or work that may take various forms, such as text, images, audio, or video.

[0986] "Server" refers to a computer system capable of receiving, processing, storing, and transmitting data over a network.

[0987] "User" refers to a person who uses the System to upload and manage creative works.

[0988] "Identifying Information" means a hash value or other information generated to uniquely identify a particular creative work.

[0989] "Metadata" is attribute information related to a creative work, and includes identification information, user emotions, and information regarding usage restrictions.

[0990] A "QR code" is a representation of metadata in the form of a two-dimensional barcode that provides access to information related to a creative work.

[0991] "URL" means a Uniform Resource Locator that provides direct access to metadata and information related to a creative work.

[0992] A "flag" refers to true / false information that is set in metadata to indicate a specific condition or state.

[0993] "Emotion engine" refers to a software module that analyzes user input data (e.g., comments and options) and evaluates their emotions.

[0994] A "dataset" refers to a collection of data that has been collected and organized for the purpose of learning or analysis.

[0995] "TF-IDF" refers to a technology that vectorizes text data based on the document frequency and inverse document frequency of words.

[0996] "Cosine similarity" is a mathematical method for measuring the similarity between vectors, and expresses the similarity using the cosine value of an angle.

[0997] "Feedback" refers to the reactions and information provided to the user by the system, and in this case refers to the information based on the user's emotions using the emotion engine.

[0998] This invention is a system that strengthens copyright protection for creative works and prevents unauthorized use by generative AI models. This system combines multiple means, including a server, a terminal, and a user, to assign unique identification information to creative works and provide feedback based on the user's emotions using an emotion engine.

[0999] Upload your creations and generate metadata

[1000] Users upload their creations to the server through an interface. Uploaded data can be in various formats, including text, images, and audio. The interface can be a browser or a dedicated application. For each uploaded creation, the server generates identification information (hash value) using the SHA-256 hash algorithm and stores this information as metadata.

[1001] Specific behavior:

[1002] The server calculates a SHA-256 hash of the uploaded creative work and stores the result as metadata.

[1003] If the user does not want the data to be used for learning, the flag "Do-Not-Learn: True" is added to the metadata.

[1004] User sentiment analysis

[1005] The server uses an emotion engine to analyze the user's emotions when uploading a creative work, and the emotion engine evaluates the emotions based on the user's input (e.g., text comments and options) and records the results in the metadata.

[1006] Specific behavior:

[1007] The server passes the comments and options entered by the user to the emotion engine.

[1008] The emotion engine analyzes the input data and stores emotions (e.g., "joy" or "anxiety") as metadata.

[1009] QR code and URL generation

[1010] The server generates a QR code based on the generated metadata and embeds it in the creative work or generates a URL associated with the creative work, allowing access to the metadata about the creative work via the QR code or URL.

[1011] Specific behavior:

[1012] The server passes the metadata to a QR code generator to generate a QR code image.

[1013] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[1014] Processing the training dataset

[1015] When an administrator of a device (the hardware that runs an AI model) loads a new dataset, the device checks the metadata of each piece of data and, if it contains any creative content flagged as "Do-Not-Learn: True," it excludes that data from the learning process.

[1016] Specific behavior:

[1017] The device accesses the metadata of each creation in the loaded dataset and checks the flag.

[1018] Exclude creations with the flag set to "True" from the AI ​​model's training set.

[1019] Similarity measurement and feedback

[1020] Users upload newly created content to the server. The server then uses TF-IDF vectorization to measure cosine similarity to match the content with existing creative content databases. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. This report includes feedback based on the user's emotions generated by the emotion engine.

[1021] Specific behavior:

[1022] The server vectorizes the newly uploaded content using the TF-IDF algorithm.

[1023] The cosine similarity is calculated and the similarity with the creations in the existing database is checked.

[1024] If the similarity exceeds a threshold, a notification report is generated and sent to the user.

[1025] The notification content is customized according to the user's emotions analyzed by the emotion engine.

[1026] Specific examples

[1027] For example, consider a scenario where a user uploads a novel they have written to a server. The server generates an identifier for the novel using a SHA-256 hash and stores this information as metadata. It also adds a "Do-Not-Learn: True" flag to the metadata and records the user's emotional response (e.g., "Delight") as analyzed by the emotion engine. The server then uses this metadata to generate a QR code and embeds it in the novel. Later, when the AI ​​model trains a new dataset, it checks the metadata for each piece of data loaded and excludes any data with the "True" flag from training.

[1028] Prompt Sentence Examples

[1029] Some examples of prompts that can be input to a generative AI model include:

[1030] "Compare the content of your new novel. Calculate the similarity with the existing database and report the results."

[1031] or

[1032] "Please upload this image and exclude it from AI training data. Add the reason 'Do not want to use for training' to the metadata."

[1033] As a result, this system can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide feedback that reflects the user's emotions.

[1034] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1035] Processing flow

[1036] Step 1: Upload your creation

[1037] Input: The user selects a creative file and uploads it through the interface.

[1038] Specific behavior:

[1039] The user selects the creative file and clicks the "Upload" button.

[1040] The server receives the uploaded data.

[1041] Output: Creative files are saved to the server.

[1042] Step 2: Generate Metadata

[1043] Input: Uploaded creative data.

[1044] Specific behavior:

[1045] The server generates a hash value of the received creative work using the SHA-256 hashing algorithm.

[1046] The identification information (hash value) is saved as metadata, and if the user does not want it to be used for learning, "Do-Not-Learn: True" is added to the metadata.

[1047] Output: Metadata is generated and includes hash values ​​and flags.

[1048] Step 3: Conducting sentiment analysis

[1049] Input: User input data (comments and choices).

[1050] Specific behavior:

[1051] The server passes the user's input data to the emotion engine for emotion analysis.

[1052] The emotion engine analyzes the input data and identifies emotions (e.g., "joy" or "anxiety").

[1053] The identified emotion is recorded in the metadata.

[1054] Output: The sentiment result is added to the metadata.

[1055] Step 4: Generate a QR code or URL

[1056] Input: The generated metadata.

[1057] Specific behavior:

[1058] The server passes the metadata to a QR code generator to generate a QR code.

[1059] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[1060] Output: A QR code or URL is created and embedded in the creation.

[1061] Step 5: Processing the training data set

[1062] Input: The newly loaded dataset.

[1063] Specific behavior:

[1064] The terminal obtains metadata for each creation in the data set.

[1065] Check the "Do-Not-Learn" flag in the metadata and exclude those that are "True" from the learning process.

[1066] Output: A list of creative works that were excluded from the training dataset.

[1067] Step 6: Measure and report similarity

[1068] Input: The newly generated content.

[1069] Specific behavior:

[1070] A user uploads new content to the server.

[1071] The server converts the new content into vector format using TF-IDF vectorization and matches it with an existing creative database.

[1072] It calculates the cosine similarity and generates a report if the similarity exceeds a certain threshold.

[1073] Use the emotion engine to add feedback to reports based on user emotions.

[1074] Output: A report containing information on highly similar creative works is generated and notified to the user or administrator.

[1075] The above are the specific processing steps of the program of this system, which effectively achieves the protection of creative works and the prevention of unauthorized use, while providing feedback based on the user's emotions.

[1076] (Application example 2)

[1077] 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."

[1078] Conventional copyright protection systems lacked sufficient functionality to prevent the unauthorized use of creative works, and they faced challenges in preventing AI from learning copyrighted works without permission and generating similar content. Furthermore, they lacked functionality to provide feedback based on user emotions, which resulted in a lack of improvement in the user experience.

[1079] 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.

[1080] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for reporting the identification information and similarity of highly similar creative works and providing feedback according to the user's emotions using an emotion engine. This makes it possible to effectively prevent unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[1081] "Creative works" are original works that should be protected by copyright, such as text, music, video, and art.

[1082] "Identifying Information" is a string of data, such as a SHA-256 hash, used to uniquely identify a creative work.

[1083] "Metadata" is additional information related to a creative work, including data such as "Do-Not-Learn" flags and user sentiment.

[1084] A "QR code" is a two-dimensional barcode that can visually embed information related to a creative work.

[1085] "URL" stands for Uniform Resource Locator, an internet address for accessing the metadata of a creative work.

[1086] A "dataset" is a collection of data used to train an AI model.

[1087] A "flag" is a data field for indicating a specific condition, and in this case has a value such as "Do-Not-Learn: True."

[1088] "Similarity" is an index that indicates the degree of correspondence between new content and existing creations.

[1089] "Emotion Engine" is an AI technology that analyzes and evaluates users' emotions in real time.

[1090] "Feedback" is information or notifications provided to the user by the system, based on the user's actions and emotions.

[1091] Specific embodiments for carrying out the present invention are described below.

[1092] Overall system configuration

[1093] The system receives creative works, assigns identifying information and generates metadata, then generates QR codes and URLs based on the metadata and embeds them in the creative works. It also checks the metadata of newly loaded datasets and excludes creative works with specific flags from learning. It also measures the similarity of the generated new content with creative works in the existing database, and provides feedback to the user if the similarity is high. This feedback is provided using an emotion engine.

[1094] Hardware and Software Used

[1095] Cloud-based servers: Use AWS or Google Cloud.

[1096] Emotion engine: Uses Microsoft Azure Emotion API.

[1097] Database: Manage data with MySQL or PostgreSQL.

[1098] QR code generation library: Uses the Python qrcode library.

[1099] Similarity calculation: Uses the Python sklearn library.

[1100] Program processing overview

[1101] Upload your creations and generate metadata

[1102] Users upload their creations to the server using a smartphone app, which generates a unique identifier using the SHA-256 algorithm. The server then analyzes the user's emotions with an AI emotion engine and adds a "Do-Not-Learn" flag and emotional information as metadata to the generated identifier.

[1103] Generate a QR code or URL

[1104] The server generates a QR code or a URL from the generated metadata and embeds it in the creation.

[1105] Processing the training dataset

[1106] When a dataset is loaded, the server checks the metadata of each piece of data. If the "Do-Not-Learn: True" flag is set, the creation is excluded from learning.

[1107] Measuring and reporting similarity

[1108] When newly created content is uploaded, the server matches it with existing creations in the database using TF-IDF vectorization and cosine similarity calculation. If the similarity is high, the server sends a notification to the user with emotional feedback.

[1109] Examples of specific examples and prompts

[1110] Examples:

[1111] 1. When a user uploads a short story they have written, the server generates a SHA-256 hash of it and adds a "Do-Not-Learn: True" flag and sentiment information (e.g., "Satisfied") to the metadata.

[1112] 2. Generate a QR code from the metadata and embed it in the novel file.

[1113] 3. When another user uploads a sentence, the server compares the sentence with the existing database and finds that the similarity is 85%, so it sends the user emotional feedback saying, "Your sentence is 85% similar. Please reconsider it from a different perspective."

[1114] Example prompt sentence:

[1115] Input text: "A user has uploaded a poem of their own. Compare it to a database of existing poems."

[1116] system:

[1117] 1. Calculate the SHA-256 hash value of the input text.

[1118] 2. Add "Do-Not-Learn" flag and user sentiment to metadata.

[1119] 3. TF-IDF vectorization and calculation of cosine similarity.

[1120] 4. Generate feedback based on user emotions.

[1121] This invention is a system that can effectively prevent the unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[1122] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1123] Step 1:

[1124] Users upload their creations to the server using a smartphone app. The file of the creation they upload becomes the input. The server receives this file.

[1125] Step 2:

[1126] The server generates an identifier for the received creative work using the SHA-256 hash algorithm, with the identifier being the input and the generated hash value being the output.

[1127] Step 3:

[1128] The server saves the generated hash value as metadata. At the same time, if the user indicates that they do not want their data to be used for learning, a flag "Do-Not-Learn: True" is added to the metadata. Furthermore, an AI emotion engine is used to analyze the user's emotions at the time of upload and the results are recorded in the metadata. The emotion engine is the input, and the emotion result is the output.

[1129] Step 4:

[1130] The server generates a QR code or a URL based on the generated metadata. This metadata is the input, and the generated QR code or URL is the output.

[1131] Step 5:

[1132] The server embeds the generated QR code or URL into the creative work. At this time, the creative work file and the QR code or URL are input, and the creative work file with the embedded QR code or URL is output.

[1133] Step 6:

[1134] When a device administrator loads a new dataset, the server checks the metadata of each data set. The new dataset is the input, and the verification result is the output. If the flag "Do-Not-Learn: True" is present, the creation is excluded from learning.

[1135] Step 7:

[1136] Users upload newly generated content to the server, which then matches this content with a database of existing creations.

[1137] Step 8:

[1138] The server measures the similarity between the new content and existing creative works by performing TF-IDF vectorization and cosine similarity calculation. The new content is the input, and the similarity calculation result is the output.

[1139] Step 9:

[1140] If the similarity exceeds a certain threshold, the server generates a report and notifies the user or administrator. The notification process incorporates an emotion engine, which provides feedback based on the user's emotions. The similarity calculation result is the input, and the feedback notification is the output.

[1141] 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.

[1142] 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.

[1143] 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.

[1144] [Fourth embodiment]

[1145] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1146] 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.

[1147] 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).

[1148] 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.

[1149] 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.

[1150] 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).

[1151] 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.

[1152] 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.

[1153] 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.

[1154] 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.

[1155] 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.

[1156] 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.

[1157] 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."

[1158] This invention relates to a system that protects the copyright of creative works and prevents their unauthorized use by AI. Below, we will generate a program for this system and explain its processing in natural language.

[1159] Overall system overview

[1160] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities.

[1161] Program processing

[1162] Upload your creations and generate metadata

[1163] Users upload their creations to the server, which then generates identification information for the received creations using the SHA-256 hash algorithm and includes that information in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata.

[1164] Generate a QR code or URL

[1165] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[1166] Processing the training dataset

[1167] When a device (including an AI model) loads a new dataset, it checks the metadata of each piece of data. If the "Do-Not-Learn" flag is set to "True," the data is excluded from the learning process. This prevents specific creative works from being used for learning without permission.

[1168] Measuring and reporting similarity

[1169] A user uploads newly generated content to the server. The server receives the content and matches it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and notified to the user or administrator. The report includes the identity and similarity of the matching creative works.

[1170] Specific examples

[1171] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, saves it as metadata with a flag "Do-Not-Learn: True," and then uses this metadata to generate a QR code, which is then embedded in the novel.

[1172] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[1173] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[1174] The system of the present invention makes it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[1175] The processing flow will be explained below.

[1176] Step 1:

[1177] Users upload their own creations to the server, and can also indicate their preference that they do not want their creations to be used for study purposes.

[1178] Step 2:

[1179] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[1180] Step 3:

[1181] The server generates identification information and adds a flag "Do-Not-Learn: True" to the creative work as metadata based on the user's request that the work not be used for learning purposes.

[1182] Step 4:

[1183] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[1184] Step 5:

[1185] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[1186] Step 6:

[1187] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[1188] Step 7:

[1189] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[1190] Step 8:

[1191] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[1192] Step 9:

[1193] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[1194] Step 10:

[1195] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[1196] Step 11:

[1197] The server generates a report containing the identification information of highly similar creative works and the results of the similarity, and notifies the user or administrator. This notification enables a prompt response when there is a risk of copyright infringement.

[1198] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works and protect copyrights.

[1199] Example 1

[1200] 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."

[1201] There is a need to prevent copyright infringement by preventing the unauthorized use of creative works in training datasets. Furthermore, a system is needed to quickly report information when newly generated content by a generative AI model is similar to existing creative works. However, current systems are sometimes insufficient to protect creative works or detect unauthorized use.

[1202] 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.

[1203] In this invention, the server includes means for receiving creative works, means for generating identification information for the received creative works using the SHA-256 hash algorithm and creating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded dataset and excluding creative works for which the "Do-Not-Learn" flag is true, means for measuring the similarity of the generated new content with creative works in an existing database using TF-IDF vectorization, and means for reporting the identification information and similarity of highly similar creative works, thereby preventing unauthorized use of creative works and enabling rapid detection of copyright infringement.

[1204] "Creative works" refers to all works such as documents, images, audio, and video created by users themselves.

[1205] The "SHA-256 hash algorithm" is a cryptographic hash function that converts input data into a fixed-length 256-bit hash value.

[1206] "Identifying Information" means a hash value or other unique data generated to uniquely identify a Creative Work.

[1207] "Metadata" is auxiliary data containing information about the creative work, such as identifying information and "Do-Not-Learn" flags.

[1208] A "QR code" is a type of two-dimensional barcode, a data representation format for visually encoding information.

[1209] A "URL" is a string of characters used to specify resources on the Internet, and stands for Uniform Resource Locator.

[1210] A "dataset" refers to a collection of multiple pieces of data that an AI model uses for training.

[1211] A "Do-Not-Learn flag" is a flag set in metadata to prevent a creative work from being used to train AI models.

[1212] "TF-IDF vectorization" is a method for extracting important words from a document and converting them into feature vectors.

[1213] "Cosine similarity" is a method for measuring similarity based on the angle between two vectors, with the closer it is to 1, the more similar it is.

[1214] "Report" means a report containing the results of the similarity measurement and the identification of the matching creative works.

[1215] "System" refers to a set of hardware and software configurations that have functions such as receiving creative works, generating identification information, managing metadata, generating QR codes, and detecting fraud.

[1216] This invention relates to a system that protects copyrights of creative works and prevents their unauthorized use by generative AI models. This system allows users to upload their creative works to a server, where they generate and manage identifying information as metadata. The server then generates a QR code or URL based on the metadata and embeds it in the creative work to protect it. When the AI ​​model learns a new dataset, it can refer to this metadata and exclude creative works with specific flags from the learning process. Furthermore, the system has the ability to measure and report the similarity of newly generated content to existing creative works.

[1217] Upload your creations and generate metadata

[1218] First, a user uploads their creation to the server. The server generates an identifier for the received creation using the SHA-256 hash algorithm. This identifier is saved as metadata for the creation. Furthermore, if the user indicates that they do not want the creation to be used for learning, the server adds a flag "Do-Not-Learn: True" to the metadata.

[1219] Generate and embed a QR code or URL

[1220] The server generates a QR code or URL based on the generated metadata. The generated QR code or URL is used to provide access to the metadata of the creative work. For example, by creating a QR code using a QR code generation library and embedding it in a creative work, users can easily access it using a smartphone or other device.

[1221] Processing the training dataset

[1222] When a device (specifically an AI model) loads a new dataset, it checks the metadata of each piece of data. If it contains any creative work with a "Do-Not-Learn" flag set to true, it will exclude that data from the learning process. This prevents certain creative work from being used in learning without permission.

[1223] Measuring and reporting similarity

[1224] When a user uploads newly generated content to the server, the server receives the content and matches it with an existing database of creative works. During this process, the server converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works.

[1225] Specific examples

[1226] As a concrete example, consider the case where a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then uses this metadata to generate a QR code and embeds it in the novel.

[1227] Next, when the AI ​​model trains on a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[1228] Finally, when a user uploads a newly generated sentence, the server compares it with an existing novel database and generates a report to notify the user if there is a high degree of match. For example, if the generated sentence matches 85% with existing novel A, the server notifies the user of this information and prompts them to take appropriate action.

[1229] Prompt Sentence Examples

[1230] Examples of prompts for generative AI models include:

[1231] "Measure the similarity between newly uploaded sentences and sentences in the existing database, and generate a report if the similarity is 70% or higher."

[1232] "Use the SHA-256 hashing algorithm to generate an identifier for the uploaded file, include it in the metadata, and generate a QR code."

[1233] This will prevent unauthorized use of creative works and enable rapid detection of copyright infringement.

[1234] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1235] Step 1: Upload your creation

[1236] The user uploads his / her creation to the server. As input, the creation file selected by the user is required. Specifically, the user uses the server's upload form, selects the file, and clicks the upload button. As output, the server saves the received creation file and prepares it for the next processing step.

[1237] Step 2: Generate hash values ​​and create metadata

[1238] The server generates identification information by running the received creative file through the SHA-256 hash algorithm. The saved creative file data is required as input. Specifically, the server calls the "hash value generation module" to calculate the hash value of the creative work. As output, the generated hash value is included in the metadata. At this time, if the user specifies that they do not want the file to be used for learning, the "Do-Not-Learn: True" flag is added to the metadata.

[1239] Step 3: Generate a QR code or URL

[1240] The server generates a QR code or a URL based on the generated metadata. The metadata (identification information and "Do-Not-Learn" flag) is required as input. Specifically, the server generates a QR code using a "QR code generation module" and saves the QR code as image data using a file operation library. The generated QR code or URL is obtained as output.

[1241] Step 4: Embed a QR code or URL

[1242] The server embeds the generated QR code or URL into the creative work. The input required is the file data of the creative work and the generated QR code or URL. Specifically, the server uses a file manipulation library to embed the QR code image into the cover of the creative work, etc. The output is a creative work file with the QR code or URL embedded.

[1243] Step 5: Processing the training data set

[1244] The terminal (specifically, the AI ​​model) checks the metadata of each data when loading a new dataset. As input, it requires the new dataset and its metadata. Specifically, the terminal runs a "metadata analysis module" and checks the "Do-Not-Learn" flag in the metadata. As output, data whose "Do-Not-Learn" flag is true is filtered out and the data is not used in the learning process.

[1245] Step 6: Measure and report similarity

[1246] When a user uploads new content to the server, the server compares it with the existing creative database. The inputs required are the new content and the existing creative database. Specifically, the server converts the new content into vector format using TF-IDF vectorization and measures the similarity using a "cosine similarity calculation module." The output is the identification information and similarity of highly similar content.

[1247] Step 7: Generate and notify the similarity report

[1248] The server generates a report based on the measured similarity and notifies the user or administrator. As input, the server requires the similarity measurement results and an existing database of creative works. Specifically, the server uses a "report generation module" to create a report that includes the identification information and similarity of highly similar creative works. As output, the report is sent to the user or administrator, who is notified of its contents.

[1249] Each step in the system prevents unauthorized use of creative works and allows for rapid detection of copyright infringement.

[1250] (Application example 1)

[1251] 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."

[1252] In recent years, the development of generative AI models has made it easier to use and copy creative works without permission. This has made copyright protection of creative works more difficult and increased the risk of copyright infringement. To solve this problem, a system is needed that can effectively protect the copyright of creative works and prevent unauthorized use and copying.

[1253] 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.

[1254] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of the newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, means for reporting the identification information and similarity of highly similar creative works, and means for notifying users based on the highly similar creative works, thereby making it possible to prevent unauthorized use of creative works and quickly detect copyright infringement.

[1255] A "creative work" is a work that has a form or content of expression that is protected by copyright law.

[1256] "Identification information" is information that uniquely identifies a particular creative work.

[1257] "Metadata" is supplementary information about a creative work, including identifying information and specific flags.

[1258] A "QR code" is a two-dimensional barcode and a means of visually representing data.

[1259] A "URL" is an address that indicates a resource on the Internet.

[1260] A "dataset" is a collection of data that is the subject of learning or analysis.

[1261] A "flag" is a piece of metadata that indicates a particular condition or state.

[1262] "Similarity" is a numerical representation of the similarity between two pieces of content.

[1263] A "notification" is the act of informing a user about a particular event or condition.

[1264] A "server" is a computer system that provides specific services and manages data on a network.

[1265] "User" means any individual or organization that wishes to use this system to manage or protect their creative works.

[1266] The present invention provides a system for protecting creative works and preventing unauthorized use, which includes the following processes:

[1267] First, a user uploads their creation to the server. The server generates a unique identifier for the received creation using the SHA-256 hash algorithm and includes it in the metadata. If the user indicates that they do not want the work to be used for learning, the server adds a flag called "Do-Not-Learn: True" to the metadata. This metadata includes the identifier and flag.

[1268] The server then generates a QR code or a URL based on the generated metadata and embeds it in the creative work, which provides access to the metadata about the creative work.

[1269] Additionally, when a device (e.g., a computer system containing an AI model) loads a new dataset, it checks the metadata of each piece of data and, if it contains any creative work with a "Do-Not-Learn" flag set to "True," the data is excluded from the learning process.

[1270] When a user uploads newly generated content to the server, the server receives the content and compares it with the existing database of creative works. It converts the text into vector form using TF-IDF vectorization and calculates the cosine similarity. If the similarity exceeds a certain threshold, it generates a report of the results and notifies the user or administrator. The report includes the identity and similarity of the matched creative works. It also notifies users based on the most similar creative works.

[1271] The hardware mainly consists of servers and terminals connected to a network. The software uses various Python libraries (e.g., hashlib, qrcode, scikit-learn, json). These libraries perform processes such as generating identification information, generating QR codes, converting text to vectors, and calculating cosine similarity.

[1272] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256, generates metadata including a flag "Do-Not-Learn: True", and then generates a QR code based on this metadata and embeds it in the novel.

[1273] Also, the following prompt sentence will be used as an input example.

[1274] Hash the following text content and convert it as a TF-IDF vector: "This is new text."

[1275] Then calculate the cosine similarity with the existing content database and display the content in descending order of similarity to the existing content in the database.

[1276] In this way, unauthorized use of creative works can be prevented and copyright infringement can be detected quickly.

[1277] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1278] Step 1:

[1279] Users upload their creations. Users upload their creations (e.g., text, images, videos) to the server, along with basic information about the creation (title, author, description, etc.).

[1280] Input: User-uploaded creative works and basic information about them.

[1281] Output: Creations and basic information stored on the server.

[1282] Specific operation: When the user selects a file on the application's upload screen and presses the "Upload" button, the creation is sent to the server.

[1283] Step 2:

[1284] The server generates the identifier and metadata. The server generates a unique identifier for the received creative work using the SHA-256 hash algorithm. It also generates metadata containing the "Do-Not-Learn: True" flag based on the user's preferences.

[1285] Input: Uploaded creative work and user preference (whether or not to use it for learning).

[1286] Output: Generated identification information and metadata.

[1287] Specifically, the server inputs the content of the creative work into the SHA-256 algorithm, obtains a hash value (identification information) as the output, and then generates metadata that combines the identification information and flags.

[1288] Step 3:

[1289] The server generates a QR code or a URL and embeds it in the creative work.The server generates a QR code or a URL based on the metadata and embeds it in the creative work.

[1290] Input: Generated metadata.

[1291] Output: Your creation with an embedded QR code or URL.

[1292] Specific operation: The server uses the qrcode library to generate a QR code that visualizes the metadata, places it in the appropriate location on the work, and saves the generated QR code as a file.

[1293] Step 4:

[1294] When the device loads a new dataset, it checks the metadata. Before adding a new dataset to the learning process, the device checks the metadata of each data and excludes data with the "Do-Not-Learn" flag.

[1295] Input: A new dataset.

[1296] Output: Data excluded from the training process.

[1297] What it does: When the device loads a dataset, it analyzes the metadata of each piece of data and adds flagged data to an exclusion list.

[1298] Step 5:

[1299] A user uploads new content and the server measures the similarity. When a user uploads new content to the server, the server measures the similarity with existing creations in the database. TF-IDF vectorization and cosine similarity are used to calculate the similarity between the new content and existing content.

[1300] Input: New content.

[1301] Output: Similarity measurement results.

[1302] Specific operation: The server performs TF-IDF vectorization and calculates the cosine similarity between the resulting vectors. For content with high similarity, the server notifies the user of the results in a report.

[1303] Step 6:

[1304] The server notifies the user of the similarity measurement results. If the similarity exceeds a certain threshold, the server notifies the user of the content's identification information and the similarity.

[1305] Input: Similarity measurement results.

[1306] Output: Notification to the user.

[1307] Specific operation: The server analyzes the similarity measurement results, and if the threshold is exceeded, an email or in-app notification is sent to the user.

[1308] In this way, the system of the present invention takes specific steps to prevent unauthorized use of creative works and quickly detect copyright infringement.

[1309] 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.

[1310] This invention combines an emotion engine with a system that strengthens copyright protection for creative works and prevents unauthorized use by AI. This system recognizes users' emotions and provides feedback based on uploading creative works and judging their similarity. Below, we will generate a program for this system and explain its processing in natural language.

[1311] Overall system overview

[1312] This system uploads creative works to a server, generates unique identification information for them, and manages it as metadata. It then uses this metadata to generate QR codes and URLs, which are embedded in the creative works. The generated metadata includes a flag indicating that the work should not be used for learning. When the AI ​​model learns a new dataset, this flag is used to exclude specific creative works from the learning process. It also compares the similarity of new content generated by the generative AI with existing creative works, and reports any high similarities. By combining this process with an emotion engine, it is possible to provide feedback and warnings that reflect the user's emotions.

[1313] Program processing

[1314] Upload your creations and generate metadata

[1315] Users upload their creations to the server. The server generates identification information for the received creations using the SHA-256 hash algorithm and includes this information in the metadata. If the user indicates that they do not want their creation to be used for learning, the server adds a "Do-Not-Learn: True" flag to the metadata. Furthermore, the server uses an emotion engine to analyze the user's emotions at the time of uploading and records the results in the metadata.

[1316] Generate a QR code or URL

[1317] The server generates a QR code or a URL based on the generated metadata and embeds it in the creative work. The QR code or the URL provides access to the metadata about the creative work.

[1318] Processing the training dataset

[1319] When a device administrator (including an AI model) loads a new dataset, the device retrieves the metadata for each piece of data. It checks the metadata for each piece of data, and if the "Do-Not-Learn" flag is set to "True," it excludes that data from the learning process. This prevents specific creative works from being improperly learned by the AI ​​model.

[1320] Measuring and reporting similarity

[1321] Users upload newly created content to the server. The server receives the content and compares it with an existing database of creative works. TF-IDF vectorization is used to convert the text into vector form and calculate the cosine similarity. If the similarity exceeds a certain threshold, a report is generated and the user or administrator is notified. An emotion engine is built into this notification process, providing feedback based on the user's emotions.

[1322] Specific examples

[1323] As a concrete example, suppose a user uploads a novel they have written to a server. The server generates a hash value for the novel using SHA-256 and saves it as metadata along with a flag "Do-Not-Learn: True." It then analyzes the user's emotions using an emotion engine and records the result (e.g., "joy") in the metadata. The server then uses this metadata to generate a QR code and embeds it in the novel.

[1324] Next, when the AI ​​model is training a new dataset, it checks the metadata of each creative work in that dataset, and if it finds a novel flagged as "Do-Not-Learn: True," it will exclude that novel from training.

[1325] Finally, when a user uploads a newly generated sentence, the server compares it with the existing novel database and notifies the user if there is a high match. At this time, the emotion engine evaluates the user's current emotion and provides appropriate feedback (e.g., "The generated sentence is 85% consistent. Please be careful.").

[1326] As described above, the system of the present invention can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide appropriate feedback according to the user's emotions.

[1327] The processing flow will be explained below.

[1328] Step 1:

[1329] Users upload their creations to the server. At this time, they can also indicate their preference that they do not want their creations to be used for learning. The emotion engine then analyzes the user's emotions in real time.

[1330] Step 2:

[1331] The server generates an identifier for the uploaded creative work using the SHA-256 hash algorithm, which uniquely identifies the creative work.

[1332] Step 3:

[1333] The server generates identification information and adds a flag "Do-Not-Learn: True" to the metadata based on the user's preference not to have it used for learning.

[1334] Step 4:

[1335] The server checks the analysis results of the emotion engine and records the user's emotions (e.g., "happiness," "anxiety," "anger," etc.) in metadata, allowing us to track emotional information related to creative works.

[1336] Step 5:

[1337] The server generates a QR code or URL based on the generated metadata, allowing easy access to the metadata of the creative work.

[1338] Step 6:

[1339] The server embeds the generated QR code or URL in the creation and provides it to the user, allowing the creation's metadata to be always available for reference.

[1340] Step 7:

[1341] When the administrator of the device (including the AI ​​model) loads a new dataset, the device retrieves the metadata for each piece of data.

[1342] Step 8:

[1343] The device checks the metadata of each piece of data it acquires, and if the "Do-Not-Learn: True" flag is set, it excludes that data from the learning process, preventing specific creative works from being improperly learned by the AI ​​model.

[1344] Step 9:

[1345] A user uploads newly generated content to a server, which may have been generated by an AI model or similar.

[1346] Step 10:

[1347] The server receives the uploaded generated content and compares it with a database of existing creative works. For this comparison, the server converts the text into vector form using TF-IDF vectorization.

[1348] Step 11:

[1349] The server calculates the cosine similarity between the vectors of the generated content and existing creations to measure the degree of match, identifies creations with high similarity, and records their identification information and the degree of similarity.

[1350] Step 12:

[1351] The server generates a report of the similarity scores and the identification information of the most similar creations, and notifies the user or administrator. This notification process incorporates an emotion engine, which provides feedback based on the user's emotions.

[1352] Step 13:

[1353] The server uses an emotion engine to provide appropriate feedback based on the user's emotions. For example, it may notify the user that "The generated text is 85% consistent with existing novel A. The user's emotions are uneasy, so we recommend further confirmation."

[1354] The above is the processing flow of the system of the present invention, which makes it possible to prevent unauthorized use of creative works, protect copyrights, and provide appropriate feedback according to the user's emotions.

[1355] Example 2

[1356] 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."

[1357] Systems that focus on protecting the copyright of creative works and preventing their unauthorized use generally lack the ability to control widespread use. They also often lack feedback functions that take user sentiment into account. Therefore, there is a need for a system that provides feedback based on user sentiment while strengthening copyright protection and preventing unauthorized use by AI.

[1358] The identification process by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for receiving a creative work, means for assigning unique identification information to the creative work and generating metadata, means for analyzing user emotions and recording the results in the metadata, means for generating a QR code or a URL based on the generated metadata, means for embedding the generated QR code or URL in the creative work, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for providing feedback based on the user's emotions when reporting the identification information and similarity of highly similar creative works. This makes it possible to provide feedback according to the user's emotions while preventing unauthorized use of creative works and reducing the risk of copyright infringement.

[1359] A "creative work" is an original expression or work that may take various forms, such as text, images, audio, or video.

[1360] "Server" refers to a computer system capable of receiving, processing, storing, and transmitting data over a network.

[1361] "User" refers to a person who uses the System to upload and manage creative works.

[1362] "Identifying Information" means a hash value or other information generated to uniquely identify a particular creative work.

[1363] "Metadata" is attribute information related to a creative work, and includes identification information, user emotions, and information regarding usage restrictions.

[1364] A "QR code" is a representation of metadata in the form of a two-dimensional barcode that provides access to information related to a creative work.

[1365] "URL" means a Uniform Resource Locator that provides direct access to metadata and information related to a creative work.

[1366] A "flag" refers to true / false information that is set in metadata to indicate a specific condition or state.

[1367] "Emotion engine" refers to a software module that analyzes user input data (e.g., comments and options) and evaluates their emotions.

[1368] A "dataset" refers to a collection of data that has been collected and organized for the purpose of learning or analysis.

[1369] "TF-IDF" refers to a technology that vectorizes text data based on the document frequency and inverse document frequency of words.

[1370] "Cosine similarity" is a mathematical method for measuring the similarity between vectors, and expresses the similarity using the cosine value of an angle.

[1371] "Feedback" refers to the reactions and information provided to the user by the system, and in this case refers to the information based on the user's emotions using the emotion engine.

[1372] This invention is a system that strengthens copyright protection for creative works and prevents unauthorized use by generative AI models. This system combines multiple means, including a server, a terminal, and a user, to assign unique identification information to creative works and provide feedback based on the user's emotions using an emotion engine.

[1373] Upload your creations and generate metadata

[1374] Users upload their creations to the server through an interface. Uploaded data can be in various formats, including text, images, and audio. The interface can be a browser or a dedicated application. For each uploaded creation, the server generates identification information (hash value) using the SHA-256 hash algorithm and stores this information as metadata.

[1375] Specific behavior:

[1376] The server calculates a SHA-256 hash of the uploaded creative work and stores the result as metadata.

[1377] If the user does not want the data to be used for learning, the flag "Do-Not-Learn: True" is added to the metadata.

[1378] User sentiment analysis

[1379] The server uses an emotion engine to analyze the user's emotions when uploading a creative work, and the emotion engine evaluates the emotions based on the user's input (e.g., text comments and options) and records the results in the metadata.

[1380] Specific behavior:

[1381] The server passes the comments and options entered by the user to the emotion engine.

[1382] The emotion engine analyzes the input data and stores emotions (e.g., "joy" or "anxiety") as metadata.

[1383] QR code and URL generation

[1384] The server generates a QR code based on the generated metadata and embeds it in the creative work or generates a URL associated with the creative work, allowing access to the metadata about the creative work via the QR code or URL.

[1385] Specific behavior:

[1386] The server passes the metadata to a QR code generator to generate a QR code image.

[1387] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[1388] Processing the training dataset

[1389] When an administrator of a device (the hardware that runs an AI model) loads a new dataset, the device checks the metadata of each piece of data and, if it contains any creative content flagged as "Do-Not-Learn: True," it excludes that data from the learning process.

[1390] Specific behavior:

[1391] The device accesses the metadata of each creation in the loaded dataset and checks the flag.

[1392] Exclude creations with the flag set to "True" from the AI ​​model's training set.

[1393] Similarity measurement and feedback

[1394] Users upload newly created content to the server. The server then uses TF-IDF vectorization to measure cosine similarity to match the content with existing creative content databases. If the similarity exceeds a certain threshold, the server generates a report of the results and notifies the user or administrator. This report includes feedback based on the user's emotions generated by the emotion engine.

[1395] Specific behavior:

[1396] The server vectorizes the newly uploaded content using the TF-IDF algorithm.

[1397] The cosine similarity is calculated and the similarity with the creations in the existing database is checked.

[1398] If the similarity exceeds a threshold, a notification report is generated and sent to the user.

[1399] The notification content is customized according to the user's emotions analyzed by the emotion engine.

[1400] Specific examples

[1401] For example, consider a scenario where a user uploads a novel they have written to a server. The server generates an identifier for the novel using a SHA-256 hash and stores this information as metadata. It also adds a "Do-Not-Learn: True" flag to the metadata and records the user's emotional response (e.g., "Delight") as analyzed by the emotion engine. The server then uses this metadata to generate a QR code and embeds it in the novel. Later, when the AI ​​model trains a new dataset, it checks the metadata for each piece of data loaded and excludes any data with the "True" flag from training.

[1402] Prompt Sentence Examples

[1403] Some examples of prompts that can be input to a generative AI model include:

[1404] "Compare the content of your new novel. Calculate the similarity with the existing database and report the results."

[1405] or

[1406] "Please upload this image and exclude it from AI training data. Add the reason 'Do not want to use for training' to the metadata."

[1407] As a result, this system can prevent unauthorized use of creative works, reduce the risk of copyright infringement, and provide feedback that reflects the user's emotions.

[1408] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1409] Processing flow

[1410] Step 1: Upload your creation

[1411] Input: The user selects a creative file and uploads it through the interface.

[1412] Specific behavior:

[1413] The user selects the creative file and clicks the "Upload" button.

[1414] The server receives the uploaded data.

[1415] Output: Creative files are saved to the server.

[1416] Step 2: Generate Metadata

[1417] Input: Uploaded creative data.

[1418] Specific behavior:

[1419] The server generates a hash value of the received creative work using the SHA-256 hashing algorithm.

[1420] The identification information (hash value) is saved as metadata, and if the user does not want it to be used for learning, "Do-Not-Learn: True" is added to the metadata.

[1421] Output: Metadata is generated and includes hash values ​​and flags.

[1422] Step 3: Conducting sentiment analysis

[1423] Input: User input data (comments and choices).

[1424] Specific behavior:

[1425] The server passes the user's input data to the emotion engine for emotion analysis.

[1426] The emotion engine analyzes the input data and identifies emotions (e.g., "joy" or "anxiety").

[1427] The identified emotion is recorded in the metadata.

[1428] Output: The sentiment result is added to the metadata.

[1429] Step 4: Generate a QR code or URL

[1430] Input: The generated metadata.

[1431] Specific behavior:

[1432] The server passes the metadata to a QR code generator to generate a QR code.

[1433] The generated QR code can be embedded in the image data of the creation or saved as a URL.

[1434] Output: A QR code or URL is created and embedded in the creation.

[1435] Step 5: Processing the training data set

[1436] Input: The newly loaded dataset.

[1437] Specific behavior:

[1438] The terminal obtains metadata for each creation in the data set.

[1439] Check the "Do-Not-Learn" flag in the metadata and exclude those that are "True" from the learning process.

[1440] Output: A list of creative works that were excluded from the training dataset.

[1441] Step 6: Measure and report similarity

[1442] Input: The newly generated content.

[1443] Specific behavior:

[1444] A user uploads new content to the server.

[1445] The server converts the new content into vector format using TF-IDF vectorization and matches it with an existing creative database.

[1446] It calculates the cosine similarity and generates a report if the similarity exceeds a certain threshold.

[1447] Use the emotion engine to add feedback to reports based on user emotions.

[1448] Output: A report containing information on highly similar creative works is generated and notified to the user or administrator.

[1449] The above are the specific processing steps of the program of this system, which effectively achieves the protection of creative works and the prevention of unauthorized use, while providing feedback based on the user's emotions.

[1450] (Application example 2)

[1451] 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."

[1452] Conventional copyright protection systems lacked sufficient functionality to prevent the unauthorized use of creative works, and they faced challenges in preventing AI from learning copyrighted works without permission and generating similar content. Furthermore, they lacked functionality to provide feedback based on user emotions, which resulted in a lack of improvement in the user experience.

[1453] 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.

[1454] In this invention, the server includes means for receiving creative works, means for assigning unique identification information to the creative works and generating metadata, means for generating a QR code or URL based on the generated metadata, means for embedding the generated QR code or URL in the creative works, means for checking the metadata of a newly loaded data set and excluding creative works with specific flags, means for measuring the similarity of the generated new content with creative works in an existing database, and means for reporting the identification information and similarity of highly similar creative works and providing feedback according to the user's emotions using an emotion engine. This makes it possible to effectively prevent unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[1455] "Creative works" are original works that should be protected by copyright, such as text, music, video, and art.

[1456] "Identifying Information" is a string of data, such as a SHA-256 hash, used to uniquely identify a creative work.

[1457] "Metadata" is additional information related to a creative work, including data such as "Do-Not-Learn" flags and user sentiment.

[1458] A "QR code" is a two-dimensional barcode that can visually embed information related to a creative work.

[1459] "URL" stands for Uniform Resource Locator, an internet address for accessing the metadata of a creative work.

[1460] A "dataset" is a collection of data used to train an AI model.

[1461] A "flag" is a data field for indicating a specific condition, and in this case has a value such as "Do-Not-Learn: True."

[1462] "Similarity" is an index that indicates the degree of correspondence between new content and existing creations.

[1463] "Emotion Engine" is an AI technology that analyzes and evaluates users' emotions in real time.

[1464] "Feedback" is information or notifications provided to the user by the system, based on the user's actions and emotions.

[1465] Specific embodiments for carrying out the present invention are described below.

[1466] Overall system configuration

[1467] The system receives creative works, assigns identifying information and generates metadata, then generates QR codes and URLs based on the metadata and embeds them in the creative works. It also checks the metadata of newly loaded datasets and excludes creative works with specific flags from learning. It also measures the similarity of the generated new content with creative works in the existing database, and provides feedback to the user if the similarity is high. This feedback is provided using an emotion engine.

[1468] Hardware and Software Used

[1469] Cloud-based servers: Use AWS or Google Cloud.

[1470] Emotion engine: Uses Microsoft Azure Emotion API.

[1471] Database: Manage data with MySQL or PostgreSQL.

[1472] QR code generation library: Uses the Python qrcode library.

[1473] Similarity calculation: Uses the Python sklearn library.

[1474] Program processing overview

[1475] Upload your creations and generate metadata

[1476] Users upload their creations to the server using a smartphone app, which generates a unique identifier using the SHA-256 algorithm. The server then analyzes the user's emotions with an AI emotion engine and adds a "Do-Not-Learn" flag and emotional information as metadata to the generated identifier.

[1477] Generate a QR code or URL

[1478] The server generates a QR code or a URL from the generated metadata and embeds it in the creation.

[1479] Processing the training dataset

[1480] When a dataset is loaded, the server checks the metadata of each piece of data. If the "Do-Not-Learn: True" flag is set, the creation is excluded from learning.

[1481] Measuring and reporting similarity

[1482] When newly created content is uploaded, the server matches it with existing creations in the database using TF-IDF vectorization and cosine similarity calculation. If the similarity is high, the server sends a notification to the user with emotional feedback.

[1483] Examples of specific examples and prompts

[1484] Examples:

[1485] 1. When a user uploads a short story they have written, the server generates a SHA-256 hash of it and adds a "Do-Not-Learn: True" flag and sentiment information (e.g., "Satisfied") to the metadata.

[1486] 2. Generate a QR code from the metadata and embed it in the novel file.

[1487] 3. When another user uploads a sentence, the server compares the sentence with the existing database and finds that the similarity is 85%, so it sends the user emotional feedback saying, "Your sentence is 85% similar. Please reconsider it from a different perspective."

[1488] Example prompt sentence:

[1489] Input text: "A user has uploaded a poem of their own. Compare it to a database of existing poems."

[1490] system:

[1491] 1. Calculate the SHA-256 hash value of the input text.

[1492] 2. Add "Do-Not-Learn" flag and user sentiment to metadata.

[1493] 3. TF-IDF vectorization and calculation of cosine similarity.

[1494] 4. Generate feedback based on user emotions.

[1495] This invention is a system that can effectively prevent the unauthorized use of creative works and provide appropriate feedback based on the user's emotions.

[1496] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1497] Step 1:

[1498] Users upload their creations to the server using a smartphone app. The file of the creation they upload becomes the input. The server receives this file.

[1499] Step 2:

[1500] The server generates an identifier for the received creative work using the SHA-256 hash algorithm, with the identifier being the input and the generated hash value being the output.

[1501] Step 3:

[1502] The server saves the generated hash value as metadata. At the same time, if the user indicates that they do not want their data to be used for learning, a flag "Do-Not-Learn: True" is added to the metadata. Furthermore, an AI emotion engine is used to analyze the user's emotions at the time of upload and the results are recorded in the metadata. The emotion engine is the input, and the emotion result is the output.

[1503] Step 4:

[1504] The server generates a QR code or a URL based on the generated metadata. This metadata is the input, and the generated QR code or URL is the output.

[1505] Step 5:

[1506] The server embeds the generated QR code or URL into the creative work. At this time, the creative work file and the QR code or URL are input, and the creative work file with the embedded QR code or URL is output.

[1507] Step 6:

[1508] When a device administrator loads a new dataset, the server checks the metadata of each data set. The new dataset is the input, and the verification result is the output. If the flag "Do-Not-Learn: True" is present, the creation is excluded from learning.

[1509] Step 7:

[1510] Users upload newly generated content to the server, which then matches this content with a database of existing creations.

[1511] Step 8:

[1512] The server measures the similarity between the new content and existing creative works by performing TF-IDF vectorization and cosine similarity calculation. The new content is the input, and the similarity calculation result is the output.

[1513] Step 9:

[1514] If the similarity exceeds a certain threshold, the server generates a report and notifies the user or administrator. The notification process incorporates an emotion engine, which provides feedback based on the user's emotions. The similarity calculation result is the input, and the feedback notification is the output.

[1515] 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.

[1516] 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.

[1517] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1518] 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.

[1519] FIG. 9 illustrates an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and behaviors arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1520] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[1521] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[1522] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1523] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[1524] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[1525] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[1526] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[1527] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[1528] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[1529] It is not necessary to store all of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[1530] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[1531] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Also, the hardware resource that executes the specific processing may be a single processor.

[1532] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[1533] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[1534] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[1535] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[1536] The following is further disclosed regarding the above embodiment.

[1537] (Claim 1)

[1538] [Means for receiving creative works;

[1539] [Means for providing unique identification information for creative works and generating metadata;

[1540] [means for generating a QR code or a URL based on the generated metadata;

[1541] [Means for embedding the generated QR code or URL into a creative work;

[1542] [Means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and

[1543] [Means for measuring the similarity of the new content generated with existing creations in the database; and

[1544] [A system including means for reporting the identification information and similarity of highly similar creative works.

[1545] (Claim 2)

[1546] [The system according to claim 1, which associates the generated new content creation with metadata and stores it.

[1547] (Claim 3)

[1548] [The system of claim 1, wherein if new content is generated based on a creative work that has a particular flag set, the creation is rejected.

[1549] "Example 1"

[1550] (Claim 1)

[1551] [Means for receiving creative works;

[1552] [Means for generating identification information for the received creative work using the SHA-256 hash algorithm and creating metadata;

[1553] [means for generating a QR code or a URL based on the generated metadata;

[1554] [Means for embedding the generated QR code or URL into a creative work;

[1555] [Means of checking the metadata of newly loaded datasets and filtering out creations where the "Do-Not-Learn" flag is true; and

[1556] [Means for measuring the similarity of new content generated using TF-IDF vectorization with existing creations in the database;

[1557] [A system including means for reporting the identification information and similarity of highly similar creative works.

[1558] (Claim 2)

[1559] The system according to claim 1, wherein a report is generated based on the similarity of the generated new content and notifies the user or administrator.

[1560] (Claim 3)

[1561] [The system of claim 1, wherein when new content is generated based on a creative work that has a specific flag set, the system detects the creation and notifies the user of the creation.

[1562] "Application Example 1"

[1563] (Claim 1)

[1564] [Means for receiving creative works;

[1565] [Means for providing unique identification information for creative works and generating metadata;

[1566] [means for generating a QR code or a URL based on the generated metadata;

[1567] [Means for embedding the generated QR code or URL into a creative work;

[1568] [Means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and

[1569] [Means for measuring the similarity of the new content generated with existing creations in the database; and

[1570] [Means for reporting the identification information and similarity of highly similar creative works;

[1571] [A system including a means for notifying users based on highly similar creative works.

[1572] (Claim 2)

[1573] [The system according to claim 1, which links the generated new content creations with metadata, stores them, and manages similarities.

[1574] (Claim 3)

[1575] [The system of claim 1, wherein if new content is generated based on a creative work that has a particular flag set, the creation is rejected.

[1576] "Example 2: Combining Emotion Engines"

[1577] (Claim 1)

[1578] [Means for receiving creative works;

[1579] [Means for providing unique identification information for creative works and generating metadata;

[1580] [Means for analyzing user sentiment and recording the results in metadata;

[1581] [means for generating a QR code or a URL based on the generated metadata;

[1582] [Means for embedding the generated QR code or URL into a creative work;

[1583] [Means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and

[1584] [Means for measuring the similarity of the new content generated with existing creations in the database; and

[1585] [A system including means for providing feedback based on the user's emotions when reporting identification information of highly similar creations and the similarity.

[1586] (Claim 2)

[1587] [The system according to claim 1, which associates the generated new content creation with metadata and stores it.

[1588] (Claim 3)

[1589] [The system of claim 1, wherein if new content is generated based on a creative work that has a particular flag set, the creation is rejected.

[1590] "Application example 2 when combining emotion engines"

[1591] (Claim 1)

[1592] [Means for receiving creative works;

[1593] [Means for providing unique identification information for creative works and generating metadata;

[1594] [means for generating a QR code or a URL based on the generated metadata;

[1595] [Means for embedding the generated QR code or URL into a creative work;

[1596] [Means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and

[1597] [Means for measuring the similarity of the new content generated with existing creations in the database; and

[1598] [A system including a means for reporting the identification information and similarity of highly similar creative works, and providing feedback according to the user's emotions using an emotion engine.

[1599] (Claim 2)

[1600] [The system according to claim 1, which associates the generated new content creation with metadata and stores it.

[1601] (Claim 3)

[1602] [The system of claim 1, wherein if new content is generated based on a creative work that has a particular flag set, the creation is rejected. [Explanation of symbols]

[1603] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for receiving the creative work; a means for assigning unique identification information to the creative work and generating metadata; means for generating a QR code or a URL based on the generated metadata; A means for embedding the generated QR code or URL into a creative work; A means of checking the metadata of newly loaded datasets and filtering out creative works with specific flags; and A means for measuring the similarity of the new content generated with existing creations in the database; A system including means for reporting the identification information and similarity of highly similar creative works.

2. The system according to claim 1, wherein the generated new content creation is associated with metadata and stored.

3. 10. The system of claim 1, wherein if new content is generated based on a creative work that has a particular flag set, the creation is rejected.

Images