System

Abstract

A system is provided.SOLUTION: A system, comprising: means for receiving content from a user to determine a content creator; means for preprocessing the received content; means for extracting features from the preprocessed content; means for calculating probabilities of AI generation and human generation based on the extracted features; means for determining a rating rank for the content based on the calculated probabilities; and means for providing the rating rank to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] With the advancement of AI technology, it has become commonplace for AI to generate content in a variety of genres. This has made it difficult for users to determine whether the content was created by an AI or a human, making it difficult to correctly evaluate the value of the content. Furthermore, the lack of a reliable method for determining this makes it difficult for consumers to use content with confidence. The present invention aims to solve these problems by accurately identifying the creator of content and providing the results to users. [Means for solving the problem]

[0005] The present invention solves the above-mentioned problems by the following means. Specifically, it provides a system including a means for receiving content from a user, a means for preprocessing the received content, a means for extracting features from the preprocessed content, a means for calculating the probability of AI generation or human generation based on the extracted features, a means for determining an evaluation rank for the content based on the calculated probability, and a means for providing the evaluation rank to the user. This allows users to easily determine whether the content was generated by an AI or a human, and obtain an accurate and reliable evaluation. Furthermore, by checking the evaluation rank, users can understand the value of the content and use it with confidence.

[0006] "User" means a person or organization that uses the system to upload content and obtain evaluation results.

[0007] "Content" refers to digital data such as text, music, and video that the system receives.

[0008] A "server" is a central computer system that processes content, analyzes it, and provides the results.

[0009] "Terminal" refers to a device used by a user to upload content and receive evaluation results, and includes a PC, tablet, smartphone, etc.

[0010] "Preprocessing" is a process that standardizes the format of received content, deletes unnecessary data, and prepares the content in a state suitable for analysis.

[0011] "Feature extraction" is the process of analyzing and extracting important patterns and attributes from content; in the case of text, this refers to stylistic analysis, and in the case of music, it refers to melody analysis.

[0012] "Probability calculation" is a process that uses a machine learning model based on extracted features to numerically express the likelihood that content was generated by AI or a human.

[0013] The "evaluation rank" is a scale for evaluating content based on the results of probability calculations and classifying it into categories, and is expressed in the form of "A," "B," "C," "D," etc.

[0014] A "machine learning model" is a collection of algorithms and models that learn patterns based on past data and make predictions and judgments about new data. [Brief explanation of the drawings]

[0015] [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

[0016] 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.

[0017] First, the terms used in the following description will be explained.

[0018] 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).

[0019] 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.

[0020] 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.

[0021] 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.

[0022] 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."

[0023] [First embodiment]

[0024] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0025] 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.

[0026] 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).

[0027] 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.

[0028] 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.

[0029] 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.

[0030] 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.

[0031] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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."

[0036] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of the program for this system will be described below.

[0037] Overall overview

[0038] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. The process involves the following main steps:

[0039] Uploading content

[0040] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0041] Content Preprocessing

[0042] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0043] Feature Extraction

[0044] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0045] Judgment and probability calculation

[0046] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0047] Rating ranking

[0048] Based on the calculated probability, the server determines the content's rating. This rating is based on pre-set criteria and is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0049] Providing and displaying results

[0050] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is B."

[0051] Specific examples

[0052] Example 1: Text content

[0053] 1. The user uploads their novel from their device to the server.

[0054] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0055] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0056] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0057] 5. The server determines that there is a high probability of AI generation and assigns a rank of C.

[0058] 6. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0059] Example 2: Music content

[0060] 1. The user uploads the music file they composed from their device to the server.

[0061] 2. The server standardizes the sample rate and format of music files.

[0062] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0063] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0064] 5. The server determines that there is a high probability of generating a human and assigns a rank of A.

[0065] 6. The server sends the results to the device, and the user confirms that "There is a 60% chance that this music was created by a human, and it has an evaluation rank of A."

[0066] In this way, the system of the present invention accurately identifies the creator of content and provides the result to the user, thereby providing an environment in which content can be used with peace of mind.

[0067] The processing flow will be explained below.

[0068] Step 1:

[0069] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[0070] Step 2:

[0071] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[0072] Step 3:

[0073] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, etc. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[0074] Step 4:

[0075] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[0076] Step 5:

[0077] The server assigns an evaluation rank based on the calculated probability. Based on pre-set criteria, it determines an evaluation rank such as "A," "B," "C," or "D" for each piece of content. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0078] Step 6:

[0079] The server sends the evaluation results and rank to the device. It then creates a message containing the generated results and evaluation rank and sends it back to the device. A specific example would be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank is B."

[0080] Step 7:

[0081] The user checks the results on their device. The device displays the received message, allowing the user to see whether their content was generated by AI or a human, as well as its rating. For example, it might say, "There's a 70% chance this text was created by AI, and it has a rating of B."

[0082] Example 1

[0083] 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."

[0084] Conventional content evaluation systems lack an effective method for determining whether content is generated by AI or humans. This makes it difficult for users to judge the reliability of content. While identifying the creator is particularly important for creative works such as text and music, current technology is unable to adequately address this issue.

[0085] 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.

[0086] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, and means for providing the evaluation rank to the user. This makes it possible to determine with high accuracy whether content was generated by AI or human and provide the result to the user.

[0087] "User" refers to an entity that uploads content for evaluation using the system.

[0088] "Content" refers to data such as text or music files that users upload to a server and wish to be evaluated.

[0089] "Terminal" refers to devices used by users, such as computers, tablets, and smartphones.

[0090] "Server" refers to the core part of the system that receives content, pre-processes it, extracts features, judges it, determines its rating, and provides the results.

[0091] "Preprocessing" refers to processing such as standardizing the format and deleting unnecessary information that the server performs on the content it receives.

[0092] "Features" refer to the data characteristics used by machine learning models to determine, for example, sentence length, word diversity, or melody lines extracted from preprocessed content.

[0093] "Extraction" refers to the operation of extracting features from preprocessed content.

[0094] "Probability" refers to the percentage of likelihood that the content was generated by AI or a human, calculated by the server based on the extracted features.

[0095] The "rating rank" refers to a grade indicating the level of the content's rating, determined by the server based on the calculated probability.

[0096] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of this system will be described below.

[0097] Uploading content

[0098] Users use their own devices (PCs, tablets, smartphones, etc.) to select and upload the content they want to rate. For example, a user might select a file such as "my_novel.txt" on their PC. The device then imports the selected content and sends it to the server.

[0099] Content Preprocessing

[0100] The server pre-processes the received content. This pre-processing includes the following specific steps:

[0101] Text format conversion: The server standardizes the format to UTF-8. For example, convert "my_novel.txt" to UTF-8.

[0102] Remove unnecessary information: Remove special characters and HTML tags. For example, 」「 Remove tags such as ".

[0103] Standardize data: Standardize line breaks and spaces. For example, standardize multiple spaces into one.

[0104] Feature Extraction

[0105] The server then uses the machine learning model to extract features from the pre-processed content. This process includes the following specific steps:

[0106] Text analysis: The server extracts sentence length, word variety, and word frequency. For example, from "my_novel.txt" we measure the average sentence length and the number of unique words.

[0107] Music analysis: The server analyzes the melody line and beat pattern of the music file. For example, it obtains the melody pitch from "my_music.mp3".

[0108] Judgment and probability calculation

[0109] The server uses machine learning models to calculate the decision and probability based on the extracted features. This process includes the following specific steps:

[0110] Calculating AI generation rate: The server uses the trained model to calculate the probability of AI generation and the probability of human generation. For example, analyzing "my_novel.txt" determines that 65% of the text is AI-generated.

[0111] Probability determination: The server determines the probability as a result. For example, "65% probability of AI generation, 35% probability of human generation."

[0112] Rating ranking

[0113] The server determines the reputation rank based on the calculated probability. This process includes the following specific steps:

[0114] Criteria-based ranking: The server assigns a ranking based on pre-defined criteria. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if it is 70% or higher, it will be ranked A.

[0115] Generating evaluation results: The server generates evaluation results such as "There is a 65% chance that this content was AI-generated, and the evaluation rank is C."

[0116] Providing and displaying results

[0117] The server then sends the final judgment result and evaluation rank to the user's device. This process includes the following specific steps:

[0118] Sending the result: The server sends the evaluation result to the user's terminal. For example, it sends the evaluation result of "my_novel.txt".

[0119] Display of results: The user checks the evaluation results on their device. For example, the user's screen may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0120] Examples of prompt statements

[0121] "There is a 60% chance that this sentence was created by an AI. Please rate it."

[0122] "Please tell me who created this music file and show me the probability."

[0123] As described above, this system can analyze content uploaded by users with high accuracy and determine whether the content was generated by AI or humans. Based on these results, it determines an evaluation rank and provides it to users to help them evaluate the reliability of the content.

[0124] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0125] Step 1:

[0126] A user selects the content they want to rate on their device and uploads it. For example, a user selects a text file called "my_novel.txt" on their computer and performs the upload. The device sends the selected file to the server. The input is the content selected by the user, and the output is the content sent to the server.

[0127] Step 2:

[0128] The server preprocesses the received content. This preprocessing includes converting text formats, removing unnecessary information, and standardizing data. For example, converting "my_novel.txt" to UTF-8, removing special characters and HTML tags, and standardizing line breaks and spaces. The input is the uploaded content, and the output is the preprocessed content.

[0129] Step 3:

[0130] The server extracts features from the preprocessed content. In this step, a machine learning model is used to analyze text such as sentence length, word diversity, and word frequency. For example, the average sentence length and number of unique words are measured from "my_novel.txt." The input is the preprocessed content, and the output is the extracted feature data.

[0131] Step 4:

[0132] The server calculates the probability of AI-generated versus human-generated sentences based on the extracted feature data. In this step, a pre-trained machine learning model is used to generate results such as "This sentence has a 65% probability of being AI-generated and a 35% probability of being human-generated." The input is the extracted feature data, and the output is the probability of AI-generated versus human-generated sentences.

[0133] Step 5:

[0134] The server determines the evaluation rank based on the calculated probability. In this step, a rank is assigned according to the probability based on pre-set criteria. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if it is 70% or higher, it is ranked A. The input is the probability of AI generation and human generation, and the output is the determined evaluation rank.

[0135] Step 6:

[0136] The server sends the final judgment result and evaluation rank to the user's device. In this step, the evaluation result is sent to the user's device so that the user can check the result. For example, it may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C." The input is the evaluation rank and judgment result, and the output is the result displayed on the user's device.

[0137] (Application example 1)

[0138] 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."

[0139] In recent years, advances in generative AI models have made it difficult to distinguish between human-generated and AI-generated content. This has made it difficult for users to find reliable information sources. In particular, there is a demand for mechanisms to evaluate the reliability of content in content distribution services. The objective of this invention is to solve this problem and provide an environment in which users can use content with peace of mind.

[0140] 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.

[0141] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, means for providing the evaluation rank to the user, and means for displaying the result as an application installed on the smart device. This makes it possible to accurately determine whether content uploaded by a user was generated by AI or a human, and visualize the result on the smart device.

[0142] A "user" is an operator of a terminal that generates content or uploads content for evaluation.

[0143] "Content" refers to information expressions such as text, music, and images.

[0144] A "receiving means" is a device or system that has the function of transmitting content from a user to a server.

[0145] A "preprocessing means" is a system that has processing capabilities to standardize received content and remove unnecessary information.

[0146] A "means for extracting features" is a system that has the function of extracting specific patterns or characteristics from content.

[0147] The "means for calculating probability" is a system that calculates the probability of AI-generated or human-generated based on the extracted features.

[0148] The "means for determining an evaluation rank" is a system having a function of assigning an evaluation rank to content based on probability.

[0149] The "means for displaying results" is an application that has the function of visualizing the evaluation rank and the creator's judgment results to the user on a smart device.

[0150] A "smart device" is a mobile device that can connect to the Internet, such as a smartphone, tablet, or smart glasses.

[0151] The present invention relates to a system that determines whether user-generated content is created by AI or a human, and provides the result as an evaluation rank. The following describes an embodiment of the present invention.

[0152] System Overview

[0153] The system consists of the following main components:

[0154] 1. User's device: The smart device (smartphone, tablet, PC, etc.) used to upload content.

[0155] 2. Server: Performs content preprocessing, feature extraction, judgment, and evaluation ranking.

[0156] 3. Smart device application: An interface that displays the evaluation results to the user.

[0157] Specific hardware and software

[0158] This system uses the following hardware and software:

[0159] Server: A server with high-performance computing resources. For example, you can use a cloud server such as Amazon Web Services (AWS) or your own server.

[0160] User device: Any smart device that can connect to the Internet, including iPhones, Android smartphones, iPads, and Windows PCs.

[0161] Smart device applications: Implemented as iOS apps, Android apps, web applications, etc.

[0162] Program processing

[0163] In this system, the following processing is performed.

[0164] 1. Uploading Content

[0165] Content is uploaded from the user's terminal to the server through a smart device application. Examples include short stories, music files, and image files.

[0166] 2. Content Preprocessing

[0167] The server standardizes the received content and removes unnecessary information, including standardizing text encoding and removing special characters.

[0168] 3. Feature Extraction

[0169] Features are extracted from the preprocessed content: for text, this includes sentence length, word diversity, and word frequency list generation; for music files, this involves analysis of melody line, beat pattern, and harmonic spectrum.

[0170] 4. Judgment and evaluation ranking

[0171] Based on the extracted features, the server calculates the probability of whether the sentence is AI-generated or human-generated. To do this, a pre-trained generative AI model is used. For example, for a particular sentence, the result may be "60% probability of AI generation" or "40% probability of human generation." Based on the probability, the evaluation rank is determined as "A," "B," "C," or "D."

[0172] 5. Providing and displaying results

[0173] The final evaluation rank and judgment results are provided to the user via a smart device application. Users can view specific feedback on their devices, such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0174] Examples and Prompts

[0175] For example, a user uploads the following short story:

[0176] There stands an old tree in the forest. For hundreds of years, many birds have gathered around this tree, watching over the passage of time.

[0177] This prompt sentence is sent to the server, and after going through the steps above, a result such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C" is obtained. This result is displayed in the application on the smart device.

[0178] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0179] Step 1:

[0180] A user uploads content.

[0181] Users use their devices (smartphones, tablets, PCs, etc.) to select the content they want to rate and send it to the server via an upload form.

[0182] Input: User-uploaded content (text, music, images, etc.)

[0183] Output: Uploaded content is saved on the server

[0184] Step 2:

[0185] The server pre-processes the content.

[0186] The server standardizes the received content and removes unnecessary information, for example, standardizing text encoding and removing special characters.

[0187] Input: User-uploaded content

[0188] Output: Standardized and cleaned content

[0189] Step 3:

[0190] The server extracts the features.

[0191] The server extracts certain patterns and characteristics from the pre-processed content: for text, this includes sentence length, word diversity, and generating a list of frequent words, while for music files, it analyzes melody lines, beat patterns, and harmonic spectrum.

[0192] Input: Preprocessed content

[0193] Output: Patterns and characteristics extracted as content features

[0194] Step 4:

[0195] The server calculates the probability.

[0196] Based on the extracted features, the server calculates the probability of AI or human generation using a pre-trained generative AI model. For example, the server can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0197] Input: Extracted features

[0198] Output: AI-generated and human-generated probabilities

[0199] Step 5:

[0200] The server determines the evaluation rank.

[0201] Based on the calculated probability, the content is assigned an evaluation rank. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if the probability of human generation is 70% or higher, it is ranked A.

[0202] Input: AI-generated and human-generated probabilities

[0203] Output: Evaluation rank (A, B, C, D, etc.)

[0204] Step 6:

[0205] The server provides the results.

[0206] The final judgment result and evaluation rank are sent to the user's device, where the user can check the results.

[0207] Input: Evaluation rank and generator probability

[0208] Output: Display of results on the user's device (judgment result and evaluation rank)

[0209] Step 7:

[0210] The user checks the results.

[0211] The user can check the results on their device's application or web page, and receive feedback such as, "There is a 65% chance that this sentence was created by AI, and it has been given a rating of C."

[0212] Input: Final result sent from the server

[0213] Output: Evaluation rank and judgment result for user confirmation

[0214] 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.

[0215] This invention combines an emotion engine with a system that determines whether the creator of content is an AI or a human and provides the result to the user. Specific implementation methods for the program of this system are described below.

[0216] Overall overview

[0217] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. It also incorporates an emotion engine that recognizes the user's emotions. In addition to the traditional evaluation process, the emotion engine analyzes the user's emotional state and adjusts the evaluation results accordingly.

[0218] Uploading content

[0219] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0220] Content Preprocessing

[0221] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0222] Feature Extraction

[0223] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0224] Judgment and probability calculation

[0225] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0226] Emotional engine regulation

[0227] Before determining the rating rank, the server uses an emotion engine to recognize the user's emotions. When users upload content, they provide voice input or facial expression data via a camera. The server analyzes this data and classifies the user's emotional state (positive, negative, neutral, etc.). The server adjusts the rating rank based on the output of the emotion engine.

[0228] Rating ranking

[0229] Based on the adjusted results, the server determines the final evaluation rank, which is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, it will be ranked B, and if it is negative, it will be ranked C.

[0230] Providing and displaying results

[0231] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is C based on the user's sentiment."

[0232] Specific examples

[0233] Example 1: Text content

[0234] 1. The user uploads their novel from their device to the server.

[0235] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0236] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0237] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0238] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[0239] 6. The server adjusts the evaluation rank from B to A based on the output of the emotion engine.

[0240] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[0241] Example 2: Music content

[0242] 1. The user uploads the music file they composed from their device to the server.

[0243] 2. The server standardizes the sample rate and format of music files.

[0244] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0245] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0246] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[0247] 6. The server adjusts the evaluation rank from A to B based on the output of the emotion engine.

[0248] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[0249] In this way, the system of the present invention not only accurately identifies the creator of the content and provides the result to the user, but also analyzes the user's emotions and adjusts the evaluation results, thereby providing a more personalized evaluation.

[0250] The processing flow will be explained below.

[0251] Step 1:

[0252] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[0253] Step 2:

[0254] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[0255] Step 3:

[0256] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, and frequency word lists. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[0257] Step 4:

[0258] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[0259] Step 5:

[0260] The user provides emotional data to the emotion engine. When uploading content, the user uses a device (microphone, camera) to capture voice and facial expressions. The device then sends the collected voice and image data to the server.

[0261] Step 6:

[0262] The server uses an emotion engine to analyze the user's emotional state. For voice data, an emotion analysis algorithm is used to identify emotional categories such as positive, negative, or neutral from the tone and patterns of the voice. For image data, an expression recognition algorithm is used to recognize the emotional state from facial expressions and subtle movements.

[0263] Step 7:

[0264] The server adjusts the evaluation rank based on the output of the emotion engine. For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, the evaluation rank is adjusted from B to A. If the user's emotion is negative, the evaluation rank is adjusted from B to C.

[0265] Step 8:

[0266] The server sends the final evaluation result and rank to the device. A message containing the generated result and evaluation rank is created and sent back to the device. For example, it could be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank based on the user's sentiment is A."

[0267] Step 9:

[0268] The user checks the results on the device. The device displays the received message, allowing the user to check whether their content was generated by AI or a human, as well as its rating. For example, the device may display a message such as, "There is a 70% chance that this text was created by AI, and the rating based on user sentiment is A."

[0269] Example 2

[0270] 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."

[0271] Conventional content rating systems only distinguish whether the creator is an AI or a human, and are unable to adjust the rating taking into account the user's emotional state. As a result, ratings that do not reflect the individual needs of users based on their emotions are not provided, resulting in problems with reduced accuracy and reliability of the ratings. Furthermore, the user experience is uniform, making it difficult to provide individual responses.

[0272] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0273] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation and human generation based on the extracted features, means for acquiring user emotional data as input, means for analyzing the acquired emotional data and determining the user's emotional state, means for determining an evaluation rank based on the determined emotional state and the calculated probability of AI generation and human generation, and means for providing the evaluation rank to the user. This enables evaluation that takes user emotions into consideration and provides evaluation results that can meet individual requests.

[0274] A "server" is a computer that processes data over a network and provides services and resources to other computer systems.

[0275] A "terminal" is a device operated by a user, and is a hardware device such as a computer, tablet, or smartphone that inputs and displays content.

[0276] A "user" is a person who uses the system, uploads content, and receives the results.

[0277] "Content" refers to information resources such as text and music, and is the data or files that are the subject of evaluation.

[0278] "Preprocessing" is a general term for the process of standardizing received content and removing unnecessary information.

[0279] "Feature extraction" is the process of extracting characteristics and patterns from preprocessed content that are necessary for evaluation.

[0280] "Probability of AI vs. human generation" is a number that indicates the probability that content was generated by artificial intelligence or by a human.

[0281] "Emotion data" is information that indicates the user's emotional state, and is composed of voice data, facial expression data, and the like.

[0282] "Emotion analysis" is a process of analyzing acquired emotion data and determining the user's emotional state (positive, negative, neutral, etc.).

[0283] The "evaluation rank" is a rank that quantitatively indicates the value and quality of content based on the extracted features and the results of sentiment analysis.

[0284] "Providing an evaluation" is the act of reporting the determined evaluation rank to the user.

[0285] This invention is a system that determines whether a content creator is an AI or a human, and provides the result to the user. This system has the function of analyzing the user's emotions and adjusting the evaluation results. Specifically, the system is configured using the following hardware and software.

[0286] Hardware and software used

[0287] The devices used by users include PCs, tablets, smartphones, etc. These devices have the ability to upload content. The devices communicate with the server via a web browser or dedicated application.

[0288] The server is the central computer that processes and analyzes the content it receives. The server has the following software and libraries installed:

[0289] Text processing libraries (e.g. NLTK, spaCy)

[0290] Machine learning libraries (e.g., Scikit-learn, TensorFlow)

[0291] Music informatics library (e.g., LibROSA)

[0292] Sentiment analysis API (e.g., general sentiment analysis services)

[0293] Uploading content

[0294] Users select and upload the content they want to rate from their device. For example, if a user wants to upload a novel manuscript file from their smartphone, they click the file upload button and then press the "Upload" button, which transfers the file to the server.

[0295] Content Preprocessing

[0296] The server preprocesses the received content. This preprocessing includes format conversion, removal of unnecessary information, and standardization. Specifically, it uses the Python libraries NLTK and BeautifulSoup to remove special characters and standardize encoding. For example, in the case of a text file, the server removes special characters (e.g., $, &,) from the text and converts it to UTF-8 encoding.

[0297] Feature Extraction

[0298] After preprocessing is complete, the server uses a machine learning model to extract features from the content. For text content, feature extraction involves extracting characteristics and patterns such as sentence length, word diversity, and word frequency. Specifically, libraries such as Scikit-learn and TensorFlow are used. For music content, music informatics libraries such as LibROSA are used to extract melody lines and rhythmic patterns.

[0299] Judgment and probability calculation

[0300] Based on the extracted features, the server uses a pre-trained machine learning model to calculate the probability of AI generation versus human generation. For example, it can produce a result such as "There is a 60% probability that this sentence was generated by an AI." This can be achieved using models such as neural networks and support vector machines (SVMs).

[0301] Emotion analysis

[0302] Users provide voice data and facial expression data via a camera. The server receives this data and uses an emotion analysis API to classify the user's emotional state. For example, when a user provides voice input, the microphone is used to collect emotion data, which is then sent to the server.

[0303] Determining the evaluation rank

[0304] Based on the results of the emotion analysis, the server determines the evaluation rank. The evaluation rank is dynamically determined based on the calculated probability of AI generation and the user's emotional state. For example, the evaluation rank may be adjusted as follows: "The probability of AI generation is 65% and the user's emotion is positive, so the evaluation rank is A."

[0305] Providing and displaying results

[0306] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, a web browser might display "AI generation probability is 65%, evaluation rank is A."

[0307] Specific examples

[0308] Example 1: Text content

[0309] 1. The user uploads their novel from their device to the server.

[0310] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0311] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0312] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0313] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[0314] 6. The server adjusts the rating from B to A based on the results of the sentiment analysis.

[0315] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[0316] Example 2: Music content

[0317] 1. The user uploads the music file they composed from their device to the server.

[0318] 2. The server standardizes the sample rate and format of music files.

[0319] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0320] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0321] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[0322] 6. The server adjusts the rating from A to B based on the results of the sentiment analysis.

[0323] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[0324] Prompt Sentence Examples

[0325] "Is it likely that this novel was generated by AI?"

[0326] "Please rate this music. Please take into account the user's feelings."

[0327] The above description demonstrates how the system judges content creators and provides ratings that take user sentiment into account.

[0328] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0329] Step 1:

[0330] A user uploads content. Using their device, the user selects a text file or music file they want to rate and presses the upload button. The device then sends the selected content to the server. The input is the content file sent from the user's device, and the output is the content file uploaded to the server. Specifically, this includes the user selecting a novel text file from their smartphone and pressing the "upload" button.

[0331] Step 2:

[0332] This preprocesses the content received by the server, standardizing the format of the received content and removing unnecessary information. The input is the content file uploaded to the server, and the output is the preprocessed content data. Specifically, this includes using Python's NLTK and BeautifulSoup to remove special characters from the text and standardize the encoding to UTF-8.

[0333] Step 3:

[0334] The server extracts features from the preprocessed content. Based on the preprocessed data, features such as sentence length, word diversity, word frequency, melody line, and rhythmic patterns are analyzed. The input is the preprocessed content data, and the output is the extracted feature data. Specifically, this includes operations such as tokenizing text and generating feature vectors using Scikit-learn, and analyzing the frequency spectrum of music files using LibROSA.

[0335] Step 4:

[0336] The server calculates the probability of AI-generated or human-generated content based on the extracted features. It uses a machine learning model to calculate the probability that the content is AI-generated or human-generated. The input is the extracted feature data, and the output is the probability of AI-generated content and the probability of human-generated content. Specifically, this involves using a trained neural network model or support vector machine to calculate the probability.

[0337] Step 5:

[0338] The user provides emotion data. The user provides facial expression data using voice input or a camera. The input is the user's voice data and facial expression data, and the output is emotion data. Specifically, this includes operations such as recording voice with a microphone and sending it to a server.

[0339] Step 6:

[0340] The server analyzes the acquired emotional data and determines the user's emotional state. Using the emotion analysis API, the emotional state is classified as positive, negative, or neutral. The input is the provided emotional data, and the output is the determined emotional state. Specifically, this includes analyzing voice and facial expressions through the emotion analysis API and determining a "positive" state.

[0341] Step 7:

[0342] The server determines the evaluation rank based on the determined emotional state and the probability of AI generation and human generation. The evaluation is given as a rank from A to D. The inputs are the probability of AI generation and human generation, and the determined emotional state, and the output is the determined evaluation rank. Specifically, it includes an action to determine rank A when the probability of AI generation is high with positive emotions.

[0343] Step 8:

[0344] The server provides the evaluation rank to the user. The final judgment result and evaluation rank are sent to the user's device and displayed. The input is the determined evaluation rank, and the output is the evaluation result displayed on the user's device. Specifically, this includes generating an HTTP response and displaying in the browser, "There is a 65% probability that this content was generated by AI, and the evaluation rank is A."

[0345] (Application example 2)

[0346] 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."

[0347] Current content distribution services lack a mechanism for determining whether the content being viewed is generated by AI or humans, and for providing appropriate ratings based on the results. Furthermore, viewers' emotional state is not reflected in content ratings, and personalized ratings are not provided. This makes it difficult to accurately convey the reliability and value of content to viewers.

[0348] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI-generated or human-generated content based on the extracted features, means for recognizing the user's emotional state, means for adjusting the evaluation rank of the content based on the calculated probability and the recognized emotional state, and means for providing the evaluation rank to the user. This allows the user to not only determine whether the content they are viewing is AI-generated or human-generated, but also to receive a personalized evaluation based on their own emotional state.

[0349] The "means for receiving content from a user" is a mechanism for a user to send content created by the user using his / her own terminal to a server.

[0350] The "means for preprocessing received content" is a mechanism for standardizing the content received by the server into a specific format and performing processing to delete unnecessary information.

[0351] The "means for extracting features from preprocessed content" is a mechanism for analyzing the internal information of the preprocessed content and extracting its features as data.

[0352] "Means for calculating the probability of AI-generated or human-generated content based on extracted features" refers to algorithms and devices for calculating the probability that content is AI-generated or human-generated based on extracted features.

[0353] The "means for recognizing the user's emotional state" refers to a device and algorithm that analyzes the audio data and image data provided by the user when uploading content and determines the user's emotional state.

[0354] The "means for adjusting the evaluation rank of content based on the calculated probability and the recognized emotional state" refers to a mechanism that comprehensively evaluates the AI-generated probability of content and the user's emotional state, and determines and adjusts a visual or numerical evaluation rank.

[0355] The "means for providing the evaluation rank to the user" refers to a communication and display device for transmitting the finally determined evaluation rank to the user's terminal and displaying it.

[0356] This invention provides a system that determines whether the content a user is viewing is generated by AI or a human, and provides a rating based on the result. The system also has the ability to recognize the user's emotional state and adjust the rating based on that emotional state.

[0357] Hardware and software used

[0358] Hardware: Smartphone, webcam

[0359] Software: Python, OpenCV, Keras, scikit-learn

[0360] System configuration

[0361] 1. User's device: A device used by a user to upload content they create, such as a smartphone or tablet.

[0362] 2. Server: Performs key processing such as content preprocessing, feature extraction, probability calculation, emotion recognition, and rating adjustment.

[0363] Program processing

[0364] 1. A method for receiving content from users: Users upload text or music content to the server using devices such as smartphones or tablets. An example of user input is a prompt such as "Please rate this poem."

[0365] 2. A means of pre-processing the received content: The server standardizes the received content to a specific format and removes unnecessary information, such as removing special characters and standardizing encoding.

[0366] 3. Means for extracting features from the preprocessed content: The server analyzes the preprocessed content and extracts features such as sentence length and word diversity in the case of text, and melody lines and rhythmic patterns in the case of music.

[0367] 4. Calculating the probability of AI generation or human generation based on the extracted features: The server uses the extracted features to apply a trained machine learning model to calculate the probability that the content is AI-generated or human-generated. For example, the result may be "AI generation probability 65%, human generation probability 35%."

[0368] 5. Means for recognizing the user's emotional state: The server performs emotion recognition using the voice and image data provided by the user when uploading content. Using a model trained with OpenCV and Keras, the server determines the user's emotional state from their facial expressions and voice.

[0369] 6. Adjusting the content rating based on the calculated probability and the user's emotional state: The server determines the rating by taking into account the AI ​​generation rate and the user's emotional state, and adjusts the rating as necessary. For example, if the AI ​​generation rate is high and the user's emotion is positive, the rating will be adjusted to A.

[0370] 7. Means for providing the evaluation rank to the user: The final evaluation rank is sent to the user's terminal and displayed visually or numerically.

[0371] Specific examples

[0372] A user uploads a poem they have written to the server from their smartphone, and the text's characteristics are extracted. The system calculates the probability of AI generation at 65% and the probability of human generation at 35%. If the user's facial expression is recognized as "positive," the evaluation rank is adjusted to A, and the result is displayed on the user's device.

[0373] Prompt Sentence Examples

[0374] "Please enter the content of the poem you would like to rate."

[0375] "Please rate this music track."

[0376] In this way, the system of the present invention not only identifies the creator of the content and provides the result to the user, but also analyzes the user's emotional state and adjusts the evaluation result to provide a more personalized evaluation.

[0377] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0378] Step 1:

[0379] A user inputs text or music content from their own terminal. Specifically, they select a file through the terminal interface and press the upload button. The input is the text data or music file, which the terminal sends to the server. The output is the content data sent to the server.

[0380] Step 2:

[0381] The server preprocesses the received content. Specifically, for text content, it removes special characters and standardizes encoding, and for music content, it standardizes the sample rate. The input is the content data received from the user, and the output is the preprocessed, standardized content data.

[0382] Step 3:

[0383] The server extracts features from the preprocessed content. Specifically, in the case of text, features such as sentence length, word diversity, and frequent word list generation are extracted, while in the case of music, features such as melody line, beat pattern, and harmonic spectrum are analyzed. The input is the preprocessed content data, and the output is the extracted feature data.

[0384] Step 4:

[0385] The server calculates the probability of AI generation and human generation based on the extracted features. Specifically, it uses a machine learning model (e.g., a logistic regression model) to calculate the probability that the content was generated by AI and human generation. The input is feature data, and the output is the probability of AI generation and human generation.

[0386] Step 5:

[0387] When uploading content, users provide voice input or facial expression data via a camera. The device collects this data and sends it to the server. The input is voice data or image data, and the output is emotional data sent to the server.

[0388] Step 6:

[0389] The server recognizes the user's emotional state. Specifically, it uses a model trained with OpenCV and Keras to analyze and classify the user's emotions from audio and image data. The input is the emotional data received from the user, and the output is the recognized emotional state (e.g., positive, negative, neutral).

[0390] Step 7:

[0391] The server adjusts the content's rating based on the calculated AI generation rate and the recognized emotional state. For example, if the AI ​​generation rate is high and the user's emotion is positive, the server will raise the rating. The inputs are the AI ​​generation rate, the human generation rate, and the recognized emotional state, and the output is the adjusted rating.

[0392] Step 8:

[0393] The server provides the evaluation rank to the user. Specifically, it transmits the evaluation result to the user's terminal and displays the evaluation rank visually or numerically. The input is the adjusted evaluation rank, and the output is the evaluation rank displayed on the user's terminal.

[0394] 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.

[0395] 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.

[0396] 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.

[0397] [Second embodiment]

[0398] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0399] 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.

[0400] 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).

[0401] 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.

[0402] 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.

[0403] 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).

[0404] 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.

[0405] 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.

[0406] 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.

[0407] 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.

[0408] 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.

[0409] 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."

[0410] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of the program for this system will be described below.

[0411] Overall overview

[0412] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. The process involves the following main steps:

[0413] Uploading content

[0414] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0415] Content Preprocessing

[0416] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0417] Feature Extraction

[0418] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0419] Judgment and probability calculation

[0420] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0421] Rating ranking

[0422] Based on the calculated probability, the server determines the content's rating. This rating is based on pre-set criteria and is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0423] Providing and displaying results

[0424] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is B."

[0425] Specific examples

[0426] Example 1: Text content

[0427] 1. The user uploads their novel from their device to the server.

[0428] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0429] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0430] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0431] 5. The server determines that there is a high probability of AI generation and assigns a rank of C.

[0432] 6. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0433] Example 2: Music content

[0434] 1. The user uploads the music file they composed from their device to the server.

[0435] 2. The server standardizes the sample rate and format of music files.

[0436] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0437] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0438] 5. The server determines that there is a high probability of generating a human and assigns a rank of A.

[0439] 6. The server sends the results to the device, and the user confirms that "There is a 60% chance that this music was created by a human, and it has an evaluation rank of A."

[0440] In this way, the system of the present invention accurately identifies the creator of content and provides the result to the user, thereby providing an environment in which content can be used with peace of mind.

[0441] The processing flow will be explained below.

[0442] Step 1:

[0443] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[0444] Step 2:

[0445] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[0446] Step 3:

[0447] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, etc. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[0448] Step 4:

[0449] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[0450] Step 5:

[0451] The server assigns an evaluation rank based on the calculated probability. Based on pre-set criteria, it determines an evaluation rank such as "A," "B," "C," or "D" for each piece of content. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0452] Step 6:

[0453] The server sends the evaluation results and rank to the device. It then creates a message containing the generated results and evaluation rank and sends it back to the device. A specific example would be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank is B."

[0454] Step 7:

[0455] The user checks the results on their device. The device displays the received message, allowing the user to see whether their content was generated by AI or a human, as well as its rating. For example, it might say, "There's a 70% chance this text was created by AI, and it has a rating of B."

[0456] Example 1

[0457] 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."

[0458] Conventional content evaluation systems lack an effective method for determining whether content is generated by AI or humans. This makes it difficult for users to judge the reliability of content. While identifying the creator is particularly important for creative works such as text and music, current technology is unable to adequately address this issue.

[0459] 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.

[0460] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, and means for providing the evaluation rank to the user. This makes it possible to determine with high accuracy whether content was generated by AI or human and provide the result to the user.

[0461] "User" refers to an entity that uploads content for evaluation using the system.

[0462] "Content" refers to data such as text or music files that users upload to a server and wish to be evaluated.

[0463] "Terminal" refers to devices used by users, such as computers, tablets, and smartphones.

[0464] "Server" refers to the core part of the system that receives content, pre-processes it, extracts features, judges it, determines its rating, and provides the results.

[0465] "Preprocessing" refers to processing such as standardizing the format and deleting unnecessary information that the server performs on the content it receives.

[0466] "Features" refer to the data characteristics used by machine learning models to determine, for example, sentence length, word diversity, or melody lines extracted from preprocessed content.

[0467] "Extraction" refers to the operation of extracting features from preprocessed content.

[0468] "Probability" refers to the percentage of likelihood that the content was generated by AI or a human, calculated by the server based on the extracted features.

[0469] The "rating rank" refers to a grade indicating the level of the content's rating, determined by the server based on the calculated probability.

[0470] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of this system will be described below.

[0471] Uploading content

[0472] Users use their own devices (PCs, tablets, smartphones, etc.) to select and upload the content they want to rate. For example, a user might select a file such as "my_novel.txt" on their PC. The device then imports the selected content and sends it to the server.

[0473] Content Preprocessing

[0474] The server pre-processes the received content. This pre-processing includes the following specific steps:

[0475] Text format conversion: The server standardizes the format to UTF-8. For example, convert "my_novel.txt" to UTF-8.

[0476] Remove unnecessary information: Remove special characters and HTML tags. For example, 」「 Remove tags such as ".

[0477] Standardize data: Standardize line breaks and spaces. For example, standardize multiple spaces into one.

[0478] Feature Extraction

[0479] The server then uses the machine learning model to extract features from the pre-processed content. This process includes the following specific steps:

[0480] Text analysis: The server extracts sentence length, word variety, and word frequency. For example, from "my_novel.txt" we measure the average sentence length and the number of unique words.

[0481] Music analysis: The server analyzes the melody line and beat pattern of the music file. For example, it obtains the melody pitch from "my_music.mp3".

[0482] Judgment and probability calculation

[0483] The server uses machine learning models to calculate the decision and probability based on the extracted features. This process includes the following specific steps:

[0484] Calculating AI generation rate: The server uses the trained model to calculate the probability of AI generation and the probability of human generation. For example, analyzing "my_novel.txt" determines that 65% of the text is AI-generated.

[0485] Probability determination: The server determines the probability as a result. For example, "65% probability of AI generation, 35% probability of human generation."

[0486] Rating ranking

[0487] The server determines the reputation rank based on the calculated probability. This process includes the following specific steps:

[0488] Criteria-based ranking: The server assigns a ranking based on pre-defined criteria. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if it is 70% or higher, it will be ranked A.

[0489] Generating evaluation results: The server generates evaluation results such as "There is a 65% chance that this content was AI-generated, and the evaluation rank is C."

[0490] Providing and displaying results

[0491] The server then sends the final judgment result and evaluation rank to the user's device. This process includes the following specific steps:

[0492] Sending the result: The server sends the evaluation result to the user's terminal. For example, it sends the evaluation result of "my_novel.txt".

[0493] Display of results: The user checks the evaluation results on their device. For example, the user's screen may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0494] Examples of prompt statements

[0495] "There is a 60% chance that this sentence was created by an AI. Please rate it."

[0496] "Please tell me who created this music file and show me the probability."

[0497] As described above, this system can analyze content uploaded by users with high accuracy and determine whether the content was generated by AI or humans. Based on these results, it determines an evaluation rank and provides it to users to help them evaluate the reliability of the content.

[0498] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0499] Step 1:

[0500] A user selects the content they want to rate on their device and uploads it. For example, a user selects a text file called "my_novel.txt" on their computer and performs the upload. The device sends the selected file to the server. The input is the content selected by the user, and the output is the content sent to the server.

[0501] Step 2:

[0502] The server preprocesses the received content. This preprocessing includes converting text formats, removing unnecessary information, and standardizing data. For example, converting "my_novel.txt" to UTF-8, removing special characters and HTML tags, and standardizing line breaks and spaces. The input is the uploaded content, and the output is the preprocessed content.

[0503] Step 3:

[0504] The server extracts features from the preprocessed content. In this step, a machine learning model is used to analyze text such as sentence length, word diversity, and word frequency. For example, the average sentence length and number of unique words are measured from "my_novel.txt." The input is the preprocessed content, and the output is the extracted feature data.

[0505] Step 4:

[0506] The server calculates the probability of AI-generated versus human-generated sentences based on the extracted feature data. In this step, a pre-trained machine learning model is used to generate results such as "This sentence has a 65% probability of being AI-generated and a 35% probability of being human-generated." The input is the extracted feature data, and the output is the probability of AI-generated versus human-generated sentences.

[0507] Step 5:

[0508] The server determines the evaluation rank based on the calculated probability. In this step, a rank is assigned according to the probability based on pre-set criteria. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if it is 70% or higher, it is ranked A. The input is the probability of AI generation and human generation, and the output is the determined evaluation rank.

[0509] Step 6:

[0510] The server sends the final judgment result and evaluation rank to the user's device. In this step, the evaluation result is sent to the user's device so that the user can check the result. For example, it may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C." The input is the evaluation rank and judgment result, and the output is the result displayed on the user's device.

[0511] (Application example 1)

[0512] 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."

[0513] In recent years, advances in generative AI models have made it difficult to distinguish between human-generated and AI-generated content. This has made it difficult for users to find reliable information sources. In particular, there is a demand for mechanisms to evaluate the reliability of content in content distribution services. The objective of this invention is to solve this problem and provide an environment in which users can use content with peace of mind.

[0514] 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.

[0515] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, means for providing the evaluation rank to the user, and means for displaying the result as an application installed on the smart device. This makes it possible to accurately determine whether content uploaded by a user was generated by AI or a human, and visualize the result on the smart device.

[0516] A "user" is an operator of a terminal that generates content or uploads content for evaluation.

[0517] "Content" refers to information expressions such as text, music, and images.

[0518] A "receiving means" is a device or system that has the function of transmitting content from a user to a server.

[0519] A "preprocessing means" is a system that has processing capabilities to standardize received content and remove unnecessary information.

[0520] A "means for extracting features" is a system that has the function of extracting specific patterns or characteristics from content.

[0521] The "means for calculating probability" is a system that calculates the probability of AI-generated or human-generated based on the extracted features.

[0522] The "means for determining an evaluation rank" is a system having a function of assigning an evaluation rank to content based on probability.

[0523] The "means for displaying results" is an application that has the function of visualizing the evaluation rank and the creator's judgment results to the user on a smart device.

[0524] A "smart device" is a mobile device that can connect to the Internet, such as a smartphone, tablet, or smart glasses.

[0525] The present invention relates to a system that determines whether user-generated content is created by AI or a human, and provides the result as an evaluation rank. The following describes an embodiment of the present invention.

[0526] System Overview

[0527] The system consists of the following main components:

[0528] 1. User's device: The smart device (smartphone, tablet, PC, etc.) used to upload content.

[0529] 2. Server: Performs content preprocessing, feature extraction, judgment, and evaluation ranking.

[0530] 3. Smart device application: An interface that displays the evaluation results to the user.

[0531] Specific hardware and software

[0532] This system uses the following hardware and software:

[0533] Server: A server with high-performance computing resources. For example, you can use a cloud server such as Amazon Web Services (AWS) or your own server.

[0534] User device: Any smart device that can connect to the Internet, including iPhones, Android smartphones, iPads, and Windows PCs.

[0535] Smart device applications: Implemented as iOS apps, Android apps, web applications, etc.

[0536] Program processing

[0537] In this system, the following processing is performed.

[0538] 1. Uploading Content

[0539] Content is uploaded from the user's terminal to the server through a smart device application. Examples include short stories, music files, and image files.

[0540] 2. Content Preprocessing

[0541] The server standardizes the received content and removes unnecessary information, including standardizing text encoding and removing special characters.

[0542] 3. Feature Extraction

[0543] Features are extracted from the preprocessed content: for text, this includes sentence length, word diversity, and word frequency list generation; for music files, this involves analysis of melody line, beat pattern, and harmonic spectrum.

[0544] 4. Judgment and evaluation ranking

[0545] Based on the extracted features, the server calculates the probability of whether the sentence is AI-generated or human-generated. To do this, a pre-trained generative AI model is used. For example, for a particular sentence, the result may be "60% probability of AI generation" or "40% probability of human generation." Based on the probability, the evaluation rank is determined as "A," "B," "C," or "D."

[0546] 5. Providing and displaying results

[0547] The final evaluation rank and judgment results are provided to the user via a smart device application. Users can view specific feedback on their devices, such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0548] Examples and Prompts

[0549] For example, a user uploads the following short story:

[0550] There stands an old tree in the forest. For hundreds of years, many birds have gathered around this tree, watching over the passage of time.

[0551] This prompt sentence is sent to the server, and after going through the steps above, a result such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C" is obtained. This result is displayed in the application on the smart device.

[0552] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0553] Step 1:

[0554] A user uploads content.

[0555] Users use their devices (smartphones, tablets, PCs, etc.) to select the content they want to rate and send it to the server via an upload form.

[0556] Input: User-uploaded content (text, music, images, etc.)

[0557] Output: Uploaded content is saved on the server

[0558] Step 2:

[0559] The server pre-processes the content.

[0560] The server standardizes the received content and removes unnecessary information, for example, standardizing text encoding and removing special characters.

[0561] Input: User-uploaded content

[0562] Output: Standardized and cleaned content

[0563] Step 3:

[0564] The server extracts the features.

[0565] The server extracts certain patterns and characteristics from the pre-processed content: for text, this includes sentence length, word diversity, and generating a list of frequent words, while for music files, it analyzes melody lines, beat patterns, and harmonic spectrum.

[0566] Input: Preprocessed content

[0567] Output: Patterns and characteristics extracted as content features

[0568] Step 4:

[0569] The server calculates the probability.

[0570] Based on the extracted features, the server calculates the probability of AI or human generation using a pre-trained generative AI model. For example, the server can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0571] Input: Extracted features

[0572] Output: AI-generated and human-generated probabilities

[0573] Step 5:

[0574] The server determines the evaluation rank.

[0575] Based on the calculated probability, the content is assigned an evaluation rank. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if the probability of human generation is 70% or higher, it is ranked A.

[0576] Input: AI-generated and human-generated probabilities

[0577] Output: Evaluation rank (A, B, C, D, etc.)

[0578] Step 6:

[0579] The server provides the results.

[0580] The final judgment result and evaluation rank are sent to the user's device, where the user can check the results.

[0581] Input: Evaluation rank and generator probability

[0582] Output: Display of results on the user's device (judgment result and evaluation rank)

[0583] Step 7:

[0584] The user checks the results.

[0585] The user can check the results on their device's application or web page, and receive feedback such as, "There is a 65% chance that this sentence was created by AI, and it has been given a rating of C."

[0586] Input: Final result sent from the server

[0587] Output: Evaluation rank and judgment result for user confirmation

[0588] 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.

[0589] This invention combines an emotion engine with a system that determines whether the creator of content is an AI or a human and provides the result to the user. Specific implementation methods for the program of this system are described below.

[0590] Overall overview

[0591] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. It also incorporates an emotion engine that recognizes the user's emotions. In addition to the traditional evaluation process, the emotion engine analyzes the user's emotional state and adjusts the evaluation results accordingly.

[0592] Uploading content

[0593] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0594] Content Preprocessing

[0595] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0596] Feature Extraction

[0597] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0598] Judgment and probability calculation

[0599] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0600] Emotional engine regulation

[0601] Before determining the rating rank, the server uses an emotion engine to recognize the user's emotions. When users upload content, they provide voice input or facial expression data via a camera. The server analyzes this data and classifies the user's emotional state (positive, negative, neutral, etc.). The server adjusts the rating rank based on the output of the emotion engine.

[0602] Rating ranking

[0603] Based on the adjusted results, the server determines the final evaluation rank, which is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, it will be ranked B, and if it is negative, it will be ranked C.

[0604] Providing and displaying results

[0605] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is C based on the user's sentiment."

[0606] Specific examples

[0607] Example 1: Text content

[0608] 1. The user uploads their novel from their device to the server.

[0609] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0610] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0611] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0612] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[0613] 6. The server adjusts the evaluation rank from B to A based on the output of the emotion engine.

[0614] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[0615] Example 2: Music content

[0616] 1. The user uploads the music file they composed from their device to the server.

[0617] 2. The server standardizes the sample rate and format of music files.

[0618] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0619] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0620] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[0621] 6. The server adjusts the evaluation rank from A to B based on the output of the emotion engine.

[0622] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[0623] In this way, the system of the present invention not only accurately identifies the creator of the content and provides the result to the user, but also analyzes the user's emotions and adjusts the evaluation results, thereby providing a more personalized evaluation.

[0624] The processing flow will be explained below.

[0625] Step 1:

[0626] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[0627] Step 2:

[0628] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[0629] Step 3:

[0630] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, and frequency word lists. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[0631] Step 4:

[0632] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[0633] Step 5:

[0634] The user provides emotional data to the emotion engine. When uploading content, the user uses a device (microphone, camera) to capture voice and facial expressions. The device then sends the collected voice and image data to the server.

[0635] Step 6:

[0636] The server uses an emotion engine to analyze the user's emotional state. For voice data, an emotion analysis algorithm is used to identify emotional categories such as positive, negative, or neutral from the tone and patterns of the voice. For image data, an expression recognition algorithm is used to recognize the emotional state from facial expressions and subtle movements.

[0637] Step 7:

[0638] The server adjusts the evaluation rank based on the output of the emotion engine. For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, the evaluation rank is adjusted from B to A. If the user's emotion is negative, the evaluation rank is adjusted from B to C.

[0639] Step 8:

[0640] The server sends the final evaluation result and rank to the device. A message containing the generated result and evaluation rank is created and sent back to the device. For example, it could be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank based on the user's sentiment is A."

[0641] Step 9:

[0642] The user checks the results on the device. The device displays the received message, allowing the user to check whether their content was generated by AI or a human, as well as its rating. For example, the device may display a message such as, "There is a 70% chance that this text was created by AI, and the rating based on user sentiment is A."

[0643] Example 2

[0644] 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."

[0645] Conventional content rating systems only distinguish whether the creator is an AI or a human, and are unable to adjust the rating taking into account the user's emotional state. As a result, ratings that do not reflect the individual needs of users based on their emotions are not provided, resulting in problems with reduced accuracy and reliability of the ratings. Furthermore, the user experience is uniform, making it difficult to provide individual responses.

[0646] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0647] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation and human generation based on the extracted features, means for acquiring user emotional data as input, means for analyzing the acquired emotional data and determining the user's emotional state, means for determining an evaluation rank based on the determined emotional state and the calculated probability of AI generation and human generation, and means for providing the evaluation rank to the user. This enables evaluation that takes user emotions into consideration and provides evaluation results that can meet individual requests.

[0648] A "server" is a computer that processes data over a network and provides services and resources to other computer systems.

[0649] A "terminal" is a device operated by a user, and is a hardware device such as a computer, tablet, or smartphone that inputs and displays content.

[0650] A "user" is a person who uses the system, uploads content, and receives the results.

[0651] "Content" refers to information resources such as text and music, and is the data or files that are the subject of evaluation.

[0652] "Preprocessing" is a general term for the process of standardizing received content and removing unnecessary information.

[0653] "Feature extraction" is the process of extracting characteristics and patterns from preprocessed content that are necessary for evaluation.

[0654] "Probability of AI vs. human generation" is a number that indicates the probability that content was generated by artificial intelligence or by a human.

[0655] "Emotion data" is information that indicates the user's emotional state, and is composed of voice data, facial expression data, and the like.

[0656] "Emotion analysis" is a process of analyzing acquired emotion data and determining the user's emotional state (positive, negative, neutral, etc.).

[0657] The "evaluation rank" is a rank that quantitatively indicates the value and quality of content based on the extracted features and the results of sentiment analysis.

[0658] "Providing an evaluation" is the act of reporting the determined evaluation rank to the user.

[0659] This invention is a system that determines whether a content creator is an AI or a human, and provides the result to the user. This system has the function of analyzing the user's emotions and adjusting the evaluation results. Specifically, the system is configured using the following hardware and software.

[0660] Hardware and software used

[0661] The devices used by users include PCs, tablets, smartphones, etc. These devices have the ability to upload content. The devices communicate with the server via a web browser or dedicated application.

[0662] The server is the central computer that processes and analyzes the content it receives. The server has the following software and libraries installed:

[0663] Text processing libraries (e.g. NLTK, spaCy)

[0664] Machine learning libraries (e.g., Scikit-learn, TensorFlow)

[0665] Music informatics library (e.g., LibROSA)

[0666] Sentiment analysis API (e.g., general sentiment analysis services)

[0667] Uploading content

[0668] Users select and upload the content they want to rate from their device. For example, if a user wants to upload a novel manuscript file from their smartphone, they click the file upload button and then press the "Upload" button, which transfers the file to the server.

[0669] Content Preprocessing

[0670] The server preprocesses the received content. This preprocessing includes format conversion, removal of unnecessary information, and standardization. Specifically, it uses the Python libraries NLTK and BeautifulSoup to remove special characters and standardize encoding. For example, in the case of a text file, the server removes special characters (e.g., $, &,) from the text and converts it to UTF-8 encoding.

[0671] Feature Extraction

[0672] After preprocessing is complete, the server uses a machine learning model to extract features from the content. For text content, feature extraction involves extracting characteristics and patterns such as sentence length, word diversity, and word frequency. Specifically, libraries such as Scikit-learn and TensorFlow are used. For music content, music informatics libraries such as LibROSA are used to extract melody lines and rhythmic patterns.

[0673] Judgment and probability calculation

[0674] Based on the extracted features, the server uses a pre-trained machine learning model to calculate the probability of AI generation versus human generation. For example, it can produce a result such as "There is a 60% probability that this sentence was generated by an AI." This can be achieved using models such as neural networks and support vector machines (SVMs).

[0675] Emotion analysis

[0676] Users provide voice data and facial expression data via a camera. The server receives this data and uses an emotion analysis API to classify the user's emotional state. For example, when a user provides voice input, the microphone is used to collect emotion data, which is then sent to the server.

[0677] Determining the evaluation rank

[0678] Based on the results of the emotion analysis, the server determines the evaluation rank. The evaluation rank is dynamically determined based on the calculated probability of AI generation and the user's emotional state. For example, the evaluation rank may be adjusted as follows: "The probability of AI generation is 65% and the user's emotion is positive, so the evaluation rank is A."

[0679] Providing and displaying results

[0680] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, a web browser might display "AI generation probability is 65%, evaluation rank is A."

[0681] Specific examples

[0682] Example 1: Text content

[0683] 1. The user uploads their novel from their device to the server.

[0684] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0685] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0686] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0687] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[0688] 6. The server adjusts the rating from B to A based on the results of the sentiment analysis.

[0689] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[0690] Example 2: Music content

[0691] 1. The user uploads the music file they composed from their device to the server.

[0692] 2. The server standardizes the sample rate and format of music files.

[0693] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0694] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0695] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[0696] 6. The server adjusts the rating from A to B based on the results of the sentiment analysis.

[0697] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[0698] Prompt Sentence Examples

[0699] "Is it likely that this novel was generated by AI?"

[0700] "Please rate this music. Please take into account the user's feelings."

[0701] The above description demonstrates how the system judges content creators and provides ratings that take user sentiment into account.

[0702] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0703] Step 1:

[0704] A user uploads content. Using their device, the user selects a text file or music file they want to rate and presses the upload button. The device then sends the selected content to the server. The input is the content file sent from the user's device, and the output is the content file uploaded to the server. Specifically, this includes the user selecting a novel text file from their smartphone and pressing the "upload" button.

[0705] Step 2:

[0706] This preprocesses the content received by the server, standardizing the format of the received content and removing unnecessary information. The input is the content file uploaded to the server, and the output is the preprocessed content data. Specifically, this includes using Python's NLTK and BeautifulSoup to remove special characters from the text and standardize the encoding to UTF-8.

[0707] Step 3:

[0708] The server extracts features from the preprocessed content. Based on the preprocessed data, features such as sentence length, word diversity, word frequency, melody line, and rhythmic patterns are analyzed. The input is the preprocessed content data, and the output is the extracted feature data. Specifically, this includes operations such as tokenizing text and generating feature vectors using Scikit-learn, and analyzing the frequency spectrum of music files using LibROSA.

[0709] Step 4:

[0710] The server calculates the probability of AI-generated or human-generated content based on the extracted features. It uses a machine learning model to calculate the probability that the content is AI-generated or human-generated. The input is the extracted feature data, and the output is the probability of AI-generated content and the probability of human-generated content. Specifically, this involves using a trained neural network model or support vector machine to calculate the probability.

[0711] Step 5:

[0712] The user provides emotion data. The user provides facial expression data using voice input or a camera. The input is the user's voice data and facial expression data, and the output is emotion data. Specifically, this includes operations such as recording voice with a microphone and sending it to a server.

[0713] Step 6:

[0714] The server analyzes the acquired emotional data and determines the user's emotional state. Using the emotion analysis API, the emotional state is classified as positive, negative, or neutral. The input is the provided emotional data, and the output is the determined emotional state. Specifically, this includes analyzing voice and facial expressions through the emotion analysis API and determining a "positive" state.

[0715] Step 7:

[0716] The server determines the evaluation rank based on the determined emotional state and the probability of AI generation and human generation. The evaluation is given as a rank from A to D. The inputs are the probability of AI generation and human generation, and the determined emotional state, and the output is the determined evaluation rank. Specifically, it includes an action to determine rank A when the probability of AI generation is high with positive emotions.

[0717] Step 8:

[0718] The server provides the evaluation rank to the user. The final judgment result and evaluation rank are sent to the user's device and displayed. The input is the determined evaluation rank, and the output is the evaluation result displayed on the user's device. Specifically, this includes generating an HTTP response and displaying in the browser, "There is a 65% probability that this content was generated by AI, and the evaluation rank is A."

[0719] (Application example 2)

[0720] 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."

[0721] Current content distribution services lack a mechanism for determining whether the content being viewed is generated by AI or humans, and for providing appropriate ratings based on the results. Furthermore, viewers' emotional state is not reflected in content ratings, and personalized ratings are not provided. This makes it difficult to accurately convey the reliability and value of content to viewers.

[0722] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI-generated or human-generated content based on the extracted features, means for recognizing the user's emotional state, means for adjusting the evaluation rank of the content based on the calculated probability and the recognized emotional state, and means for providing the evaluation rank to the user. This allows the user to not only determine whether the content they are viewing is AI-generated or human-generated, but also to receive a personalized evaluation based on their own emotional state.

[0723] The "means for receiving content from a user" is a mechanism for a user to send content created by the user using his / her own terminal to a server.

[0724] The "means for preprocessing received content" is a mechanism for standardizing the content received by the server into a specific format and performing processing to delete unnecessary information.

[0725] The "means for extracting features from preprocessed content" is a mechanism for analyzing the internal information of the preprocessed content and extracting its features as data.

[0726] "Means for calculating the probability of AI-generated or human-generated content based on extracted features" refers to algorithms and devices for calculating the probability that content is AI-generated or human-generated based on extracted features.

[0727] The "means for recognizing the user's emotional state" refers to a device and algorithm that analyzes the audio data and image data provided by the user when uploading content and determines the user's emotional state.

[0728] The "means for adjusting the evaluation rank of content based on the calculated probability and the recognized emotional state" refers to a mechanism that comprehensively evaluates the AI-generated probability of content and the user's emotional state, and determines and adjusts a visual or numerical evaluation rank.

[0729] The "means for providing the evaluation rank to the user" refers to a communication and display device for transmitting the finally determined evaluation rank to the user's terminal and displaying it.

[0730] This invention provides a system that determines whether the content a user is viewing is generated by AI or a human, and provides a rating based on the result. The system also has the ability to recognize the user's emotional state and adjust the rating based on that emotional state.

[0731] Hardware and software used

[0732] Hardware: Smartphone, webcam

[0733] Software: Python, OpenCV, Keras, scikit-learn

[0734] System configuration

[0735] 1. User's device: A device used by a user to upload content they create, such as a smartphone or tablet.

[0736] 2. Server: Performs key processing such as content preprocessing, feature extraction, probability calculation, emotion recognition, and rating adjustment.

[0737] Program processing

[0738] 1. A method for receiving content from users: Users upload text or music content to the server using devices such as smartphones or tablets. An example of user input is a prompt such as "Please rate this poem."

[0739] 2. A means of pre-processing the received content: The server standardizes the received content to a specific format and removes unnecessary information, such as removing special characters and standardizing encoding.

[0740] 3. Means for extracting features from the preprocessed content: The server analyzes the preprocessed content and extracts features such as sentence length and word diversity in the case of text, and melody lines and rhythmic patterns in the case of music.

[0741] 4. Calculating the probability of AI generation or human generation based on the extracted features: The server uses the extracted features to apply a trained machine learning model to calculate the probability that the content is AI-generated or human-generated. For example, the result may be "AI generation probability 65%, human generation probability 35%."

[0742] 5. Means for recognizing the user's emotional state: The server performs emotion recognition using the voice and image data provided by the user when uploading content. Using a model trained with OpenCV and Keras, the server determines the user's emotional state from their facial expressions and voice.

[0743] 6. Adjusting the content rating based on the calculated probability and the user's emotional state: The server determines the rating by taking into account the AI ​​generation rate and the user's emotional state, and adjusts the rating as necessary. For example, if the AI ​​generation rate is high and the user's emotion is positive, the rating will be adjusted to A.

[0744] 7. Means for providing the evaluation rank to the user: The final evaluation rank is sent to the user's terminal and displayed visually or numerically.

[0745] Specific examples

[0746] A user uploads a poem they have written to the server from their smartphone, and the text's characteristics are extracted. The system calculates the probability of AI generation at 65% and the probability of human generation at 35%. If the user's facial expression is recognized as "positive," the evaluation rank is adjusted to A, and the result is displayed on the user's device.

[0747] Prompt Sentence Examples

[0748] "Please enter the content of the poem you would like to rate."

[0749] "Please rate this music track."

[0750] In this way, the system of the present invention not only identifies the creator of the content and provides the result to the user, but also analyzes the user's emotional state and adjusts the evaluation result to provide a more personalized evaluation.

[0751] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0752] Step 1:

[0753] A user inputs text or music content from their own terminal. Specifically, they select a file through the terminal interface and press the upload button. The input is the text data or music file, which the terminal sends to the server. The output is the content data sent to the server.

[0754] Step 2:

[0755] The server preprocesses the received content. Specifically, for text content, it removes special characters and standardizes encoding, and for music content, it standardizes the sample rate. The input is the content data received from the user, and the output is the preprocessed, standardized content data.

[0756] Step 3:

[0757] The server extracts features from the preprocessed content. Specifically, in the case of text, features such as sentence length, word diversity, and frequent word list generation are extracted, while in the case of music, features such as melody line, beat pattern, and harmonic spectrum are analyzed. The input is the preprocessed content data, and the output is the extracted feature data.

[0758] Step 4:

[0759] The server calculates the probability of AI generation and human generation based on the extracted features. Specifically, it uses a machine learning model (e.g., a logistic regression model) to calculate the probability that the content was generated by AI and human generation. The input is feature data, and the output is the probability of AI generation and human generation.

[0760] Step 5:

[0761] When uploading content, users provide voice input or facial expression data via a camera. The device collects this data and sends it to the server. The input is voice data or image data, and the output is emotional data sent to the server.

[0762] Step 6:

[0763] The server recognizes the user's emotional state. Specifically, it uses a model trained with OpenCV and Keras to analyze and classify the user's emotions from audio and image data. The input is the emotional data received from the user, and the output is the recognized emotional state (e.g., positive, negative, neutral).

[0764] Step 7:

[0765] The server adjusts the content's rating based on the calculated AI generation rate and the recognized emotional state. For example, if the AI ​​generation rate is high and the user's emotion is positive, the server will raise the rating. The inputs are the AI ​​generation rate, the human generation rate, and the recognized emotional state, and the output is the adjusted rating.

[0766] Step 8:

[0767] The server provides the evaluation rank to the user. Specifically, it transmits the evaluation result to the user's terminal and displays the evaluation rank visually or numerically. The input is the adjusted evaluation rank, and the output is the evaluation rank displayed on the user's terminal.

[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 determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of the program for this system will be described below.

[0785] Overall overview

[0786] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. The process involves the following main steps:

[0787] Uploading content

[0788] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0789] Content Preprocessing

[0790] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0791] Feature Extraction

[0792] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0793] Judgment and probability calculation

[0794] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0795] Rating ranking

[0796] Based on the calculated probability, the server determines the content's rating. This rating is based on pre-set criteria and is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0797] Providing and displaying results

[0798] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is B."

[0799] Specific examples

[0800] Example 1: Text content

[0801] 1. The user uploads their novel from their device to the server.

[0802] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0803] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0804] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0805] 5. The server determines that there is a high probability of AI generation and assigns a rank of C.

[0806] 6. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0807] Example 2: Music content

[0808] 1. The user uploads the music file they composed from their device to the server.

[0809] 2. The server standardizes the sample rate and format of music files.

[0810] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0811] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0812] 5. The server determines that there is a high probability of generating a human and assigns a rank of A.

[0813] 6. The server sends the results to the device, and the user confirms that "There is a 60% chance that this music was created by a human, and it has an evaluation rank of A."

[0814] In this way, the system of the present invention accurately identifies the creator of content and provides the result to the user, thereby providing an environment in which content can be used with peace of mind.

[0815] The processing flow will be explained below.

[0816] Step 1:

[0817] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[0818] Step 2:

[0819] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[0820] Step 3:

[0821] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, etc. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[0822] Step 4:

[0823] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[0824] Step 5:

[0825] The server assigns an evaluation rank based on the calculated probability. Based on pre-set criteria, it determines an evaluation rank such as "A," "B," "C," or "D" for each piece of content. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[0826] Step 6:

[0827] The server sends the evaluation results and rank to the device. It then creates a message containing the generated results and evaluation rank and sends it back to the device. A specific example would be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank is B."

[0828] Step 7:

[0829] The user checks the results on their device. The device displays the received message, allowing the user to see whether their content was generated by AI or a human, as well as its rating. For example, it might say, "There's a 70% chance this text was created by AI, and it has a rating of B."

[0830] Example 1

[0831] 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."

[0832] Conventional content evaluation systems lack an effective method for determining whether content is generated by AI or humans. This makes it difficult for users to judge the reliability of content. While identifying the creator is particularly important for creative works such as text and music, current technology is unable to adequately address this issue.

[0833] 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.

[0834] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, and means for providing the evaluation rank to the user. This makes it possible to determine with high accuracy whether content was generated by AI or human and provide the result to the user.

[0835] "User" refers to an entity that uploads content for evaluation using the system.

[0836] "Content" refers to data such as text or music files that users upload to a server and wish to be evaluated.

[0837] "Terminal" refers to devices used by users, such as computers, tablets, and smartphones.

[0838] "Server" refers to the core part of the system that receives content, pre-processes it, extracts features, judges it, determines its rating, and provides the results.

[0839] "Preprocessing" refers to processing such as standardizing the format and deleting unnecessary information that the server performs on the content it receives.

[0840] "Features" refer to the data characteristics used by machine learning models to determine, for example, sentence length, word diversity, or melody lines extracted from preprocessed content.

[0841] "Extraction" refers to the operation of extracting features from preprocessed content.

[0842] "Probability" refers to the percentage of likelihood that the content was generated by AI or a human, calculated by the server based on the extracted features.

[0843] The "rating rank" refers to a grade indicating the level of the content's rating, determined by the server based on the calculated probability.

[0844] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of this system will be described below.

[0845] Uploading content

[0846] Users use their own devices (PCs, tablets, smartphones, etc.) to select and upload the content they want to rate. For example, a user might select a file such as "my_novel.txt" on their PC. The device then imports the selected content and sends it to the server.

[0847] Content Preprocessing

[0848] The server pre-processes the received content. This pre-processing includes the following specific steps:

[0849] Text format conversion: The server standardizes the format to UTF-8. For example, convert "my_novel.txt" to UTF-8.

[0850] Remove unnecessary information: Remove special characters and HTML tags. For example, 」「 Remove tags such as ".

[0851] Standardize data: Standardize line breaks and spaces. For example, standardize multiple spaces into one.

[0852] Feature Extraction

[0853] The server then uses the machine learning model to extract features from the pre-processed content. This process includes the following specific steps:

[0854] Text analysis: The server extracts sentence length, word variety, and word frequency. For example, from "my_novel.txt" we measure the average sentence length and the number of unique words.

[0855] Music analysis: The server analyzes the melody line and beat pattern of the music file. For example, it obtains the melody pitch from "my_music.mp3".

[0856] Judgment and probability calculation

[0857] The server uses machine learning models to calculate the decision and probability based on the extracted features. This process includes the following specific steps:

[0858] Calculating AI generation rate: The server uses the trained model to calculate the probability of AI generation and the probability of human generation. For example, analyzing "my_novel.txt" determines that 65% of the text is AI-generated.

[0859] Probability determination: The server determines the probability as a result. For example, "65% probability of AI generation, 35% probability of human generation."

[0860] Rating ranking

[0861] The server determines the reputation rank based on the calculated probability. This process includes the following specific steps:

[0862] Criteria-based ranking: The server assigns a ranking based on pre-defined criteria. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if it is 70% or higher, it will be ranked A.

[0863] Generating evaluation results: The server generates evaluation results such as "There is a 65% chance that this content was AI-generated, and the evaluation rank is C."

[0864] Providing and displaying results

[0865] The server then sends the final judgment result and evaluation rank to the user's device. This process includes the following specific steps:

[0866] Sending the result: The server sends the evaluation result to the user's terminal. For example, it sends the evaluation result of "my_novel.txt".

[0867] Display of results: The user checks the evaluation results on their device. For example, the user's screen may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0868] Examples of prompt statements

[0869] "There is a 60% chance that this sentence was created by an AI. Please rate it."

[0870] "Please tell me who created this music file and show me the probability."

[0871] As described above, this system can analyze content uploaded by users with high accuracy and determine whether the content was generated by AI or humans. Based on these results, it determines an evaluation rank and provides it to users to help them evaluate the reliability of the content.

[0872] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0873] Step 1:

[0874] A user selects the content they want to rate on their device and uploads it. For example, a user selects a text file called "my_novel.txt" on their computer and performs the upload. The device sends the selected file to the server. The input is the content selected by the user, and the output is the content sent to the server.

[0875] Step 2:

[0876] The server preprocesses the received content. This preprocessing includes converting text formats, removing unnecessary information, and standardizing data. For example, converting "my_novel.txt" to UTF-8, removing special characters and HTML tags, and standardizing line breaks and spaces. The input is the uploaded content, and the output is the preprocessed content.

[0877] Step 3:

[0878] The server extracts features from the preprocessed content. In this step, a machine learning model is used to analyze text such as sentence length, word diversity, and word frequency. For example, the average sentence length and number of unique words are measured from "my_novel.txt." The input is the preprocessed content, and the output is the extracted feature data.

[0879] Step 4:

[0880] The server calculates the probability of AI-generated versus human-generated sentences based on the extracted feature data. In this step, a pre-trained machine learning model is used to generate results such as "This sentence has a 65% probability of being AI-generated and a 35% probability of being human-generated." The input is the extracted feature data, and the output is the probability of AI-generated versus human-generated sentences.

[0881] Step 5:

[0882] The server determines the evaluation rank based on the calculated probability. In this step, a rank is assigned according to the probability based on pre-set criteria. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if it is 70% or higher, it is ranked A. The input is the probability of AI generation and human generation, and the output is the determined evaluation rank.

[0883] Step 6:

[0884] The server sends the final judgment result and evaluation rank to the user's device. In this step, the evaluation result is sent to the user's device so that the user can check the result. For example, it may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C." The input is the evaluation rank and judgment result, and the output is the result displayed on the user's device.

[0885] (Application example 1)

[0886] 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."

[0887] In recent years, advances in generative AI models have made it difficult to distinguish between human-generated and AI-generated content. This has made it difficult for users to find reliable information sources. In particular, there is a demand for mechanisms to evaluate the reliability of content in content distribution services. The objective of this invention is to solve this problem and provide an environment in which users can use content with peace of mind.

[0888] 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.

[0889] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, means for providing the evaluation rank to the user, and means for displaying the result as an application installed on the smart device. This makes it possible to accurately determine whether content uploaded by a user was generated by AI or a human, and visualize the result on the smart device.

[0890] A "user" is an operator of a terminal that generates content or uploads content for evaluation.

[0891] "Content" refers to information expressions such as text, music, and images.

[0892] A "receiving means" is a device or system that has the function of transmitting content from a user to a server.

[0893] A "preprocessing means" is a system that has processing capabilities to standardize received content and remove unnecessary information.

[0894] A "means for extracting features" is a system that has the function of extracting specific patterns or characteristics from content.

[0895] The "means for calculating probability" is a system that calculates the probability of AI-generated or human-generated based on the extracted features.

[0896] The "means for determining an evaluation rank" is a system having a function of assigning an evaluation rank to content based on probability.

[0897] The "means for displaying results" is an application that has the function of visualizing the evaluation rank and the creator's judgment results to the user on a smart device.

[0898] A "smart device" is a mobile device that can connect to the Internet, such as a smartphone, tablet, or smart glasses.

[0899] The present invention relates to a system that determines whether user-generated content is created by AI or a human, and provides the result as an evaluation rank. The following describes an embodiment of the present invention.

[0900] System Overview

[0901] The system consists of the following main components:

[0902] 1. User's device: The smart device (smartphone, tablet, PC, etc.) used to upload content.

[0903] 2. Server: Performs content preprocessing, feature extraction, judgment, and evaluation ranking.

[0904] 3. Smart device application: An interface that displays the evaluation results to the user.

[0905] Specific hardware and software

[0906] This system uses the following hardware and software:

[0907] Server: A server with high-performance computing resources. For example, you can use a cloud server such as Amazon Web Services (AWS) or your own server.

[0908] User device: Any smart device that can connect to the Internet, including iPhones, Android smartphones, iPads, and Windows PCs.

[0909] Smart device applications: Implemented as iOS apps, Android apps, web applications, etc.

[0910] Program processing

[0911] In this system, the following processing is performed.

[0912] 1. Uploading Content

[0913] Content is uploaded from the user's terminal to the server through a smart device application. Examples include short stories, music files, and image files.

[0914] 2. Content Preprocessing

[0915] The server standardizes the received content and removes unnecessary information, including standardizing text encoding and removing special characters.

[0916] 3. Feature Extraction

[0917] Features are extracted from the preprocessed content: for text, this includes sentence length, word diversity, and word frequency list generation; for music files, this involves analysis of melody line, beat pattern, and harmonic spectrum.

[0918] 4. Judgment and evaluation ranking

[0919] Based on the extracted features, the server calculates the probability of whether the sentence is AI-generated or human-generated. To do this, a pre-trained generative AI model is used. For example, for a particular sentence, the result may be "60% probability of AI generation" or "40% probability of human generation." Based on the probability, the evaluation rank is determined as "A," "B," "C," or "D."

[0920] 5. Providing and displaying results

[0921] The final evaluation rank and judgment results are provided to the user via a smart device application. Users can view specific feedback on their devices, such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[0922] Examples and Prompts

[0923] For example, a user uploads the following short story:

[0924] There stands an old tree in the forest. For hundreds of years, many birds have gathered around this tree, watching over the passage of time.

[0925] This prompt sentence is sent to the server, and after going through the steps above, a result such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C" is obtained. This result is displayed in the application on the smart device.

[0926] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0927] Step 1:

[0928] A user uploads content.

[0929] Users use their devices (smartphones, tablets, PCs, etc.) to select the content they want to rate and send it to the server via an upload form.

[0930] Input: User-uploaded content (text, music, images, etc.)

[0931] Output: Uploaded content is saved on the server

[0932] Step 2:

[0933] The server pre-processes the content.

[0934] The server standardizes the received content and removes unnecessary information, for example, standardizing text encoding and removing special characters.

[0935] Input: User-uploaded content

[0936] Output: Standardized and cleaned content

[0937] Step 3:

[0938] The server extracts the features.

[0939] The server extracts certain patterns and characteristics from the pre-processed content: for text, this includes sentence length, word diversity, and generating a list of frequent words, while for music files, it analyzes melody lines, beat patterns, and harmonic spectrum.

[0940] Input: Preprocessed content

[0941] Output: Patterns and characteristics extracted as content features

[0942] Step 4:

[0943] The server calculates the probability.

[0944] Based on the extracted features, the server calculates the probability of AI or human generation using a pre-trained generative AI model. For example, the server can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0945] Input: Extracted features

[0946] Output: AI-generated and human-generated probabilities

[0947] Step 5:

[0948] The server determines the evaluation rank.

[0949] Based on the calculated probability, the content is assigned an evaluation rank. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if the probability of human generation is 70% or higher, it is ranked A.

[0950] Input: AI-generated and human-generated probabilities

[0951] Output: Evaluation rank (A, B, C, D, etc.)

[0952] Step 6:

[0953] The server provides the results.

[0954] The final judgment result and evaluation rank are sent to the user's device, where the user can check the results.

[0955] Input: Evaluation rank and generator probability

[0956] Output: Display of results on the user's device (judgment result and evaluation rank)

[0957] Step 7:

[0958] The user checks the results.

[0959] The user can check the results on their device's application or web page, and receive feedback such as, "There is a 65% chance that this sentence was created by AI, and it has been given a rating of C."

[0960] Input: Final result sent from the server

[0961] Output: Evaluation rank and judgment result for user confirmation

[0962] 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.

[0963] This invention combines an emotion engine with a system that determines whether the creator of content is an AI or a human and provides the result to the user. Specific implementation methods for the program of this system are described below.

[0964] Overall overview

[0965] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. It also incorporates an emotion engine that recognizes the user's emotions. In addition to the traditional evaluation process, the emotion engine analyzes the user's emotional state and adjusts the evaluation results accordingly.

[0966] Uploading content

[0967] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[0968] Content Preprocessing

[0969] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[0970] Feature Extraction

[0971] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[0972] Judgment and probability calculation

[0973] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[0974] Emotional engine regulation

[0975] Before determining the rating rank, the server uses an emotion engine to recognize the user's emotions. When users upload content, they provide voice input or facial expression data via a camera. The server analyzes this data and classifies the user's emotional state (positive, negative, neutral, etc.). The server adjusts the rating rank based on the output of the emotion engine.

[0976] Rating ranking

[0977] Based on the adjusted results, the server determines the final evaluation rank, which is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, it will be ranked B, and if it is negative, it will be ranked C.

[0978] Providing and displaying results

[0979] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is C based on the user's sentiment."

[0980] Specific examples

[0981] Example 1: Text content

[0982] 1. The user uploads their novel from their device to the server.

[0983] 2. The server normalizes the text (removes special characters and standardizes encoding).

[0984] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[0985] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[0986] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[0987] 6. The server adjusts the evaluation rank from B to A based on the output of the emotion engine.

[0988] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[0989] Example 2: Music content

[0990] 1. The user uploads the music file they composed from their device to the server.

[0991] 2. The server standardizes the sample rate and format of music files.

[0992] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[0993] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[0994] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[0995] 6. The server adjusts the evaluation rank from A to B based on the output of the emotion engine.

[0996] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[0997] In this way, the system of the present invention not only accurately identifies the creator of the content and provides the result to the user, but also analyzes the user's emotions and adjusts the evaluation results, thereby providing a more personalized evaluation.

[0998] The processing flow will be explained below.

[0999] Step 1:

[1000] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[1001] Step 2:

[1002] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[1003] Step 3:

[1004] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, and frequency word lists. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[1005] Step 4:

[1006] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[1007] Step 5:

[1008] The user provides emotional data to the emotion engine. When uploading content, the user uses a device (microphone, camera) to capture voice and facial expressions. The device then sends the collected voice and image data to the server.

[1009] Step 6:

[1010] The server uses an emotion engine to analyze the user's emotional state. For voice data, an emotion analysis algorithm is used to identify emotional categories such as positive, negative, or neutral from the tone and patterns of the voice. For image data, an expression recognition algorithm is used to recognize the emotional state from facial expressions and subtle movements.

[1011] Step 7:

[1012] The server adjusts the evaluation rank based on the output of the emotion engine. For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, the evaluation rank is adjusted from B to A. If the user's emotion is negative, the evaluation rank is adjusted from B to C.

[1013] Step 8:

[1014] The server sends the final evaluation result and rank to the device. A message containing the generated result and evaluation rank is created and sent back to the device. For example, it could be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank based on the user's sentiment is A."

[1015] Step 9:

[1016] The user checks the results on the device. The device displays the received message, allowing the user to check whether their content was generated by AI or a human, as well as its rating. For example, the device may display a message such as, "There is a 70% chance that this text was created by AI, and the rating based on user sentiment is A."

[1017] Example 2

[1018] 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."

[1019] Conventional content rating systems only distinguish whether the creator is an AI or a human, and are unable to adjust the rating taking into account the user's emotional state. As a result, ratings that do not reflect the individual needs of users based on their emotions are not provided, resulting in problems with reduced accuracy and reliability of the ratings. Furthermore, the user experience is uniform, making it difficult to provide individual responses.

[1020] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1021] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation and human generation based on the extracted features, means for acquiring user emotional data as input, means for analyzing the acquired emotional data and determining the user's emotional state, means for determining an evaluation rank based on the determined emotional state and the calculated probability of AI generation and human generation, and means for providing the evaluation rank to the user. This enables evaluation that takes user emotions into consideration and provides evaluation results that can meet individual requests.

[1022] A "server" is a computer that processes data over a network and provides services and resources to other computer systems.

[1023] A "terminal" is a device operated by a user, and is a hardware device such as a computer, tablet, or smartphone that inputs and displays content.

[1024] A "user" is a person who uses the system, uploads content, and receives the results.

[1025] "Content" refers to information resources such as text and music, and is the data or files that are the subject of evaluation.

[1026] "Preprocessing" is a general term for the process of standardizing received content and removing unnecessary information.

[1027] "Feature extraction" is the process of extracting characteristics and patterns from preprocessed content that are necessary for evaluation.

[1028] "Probability of AI vs. human generation" is a number that indicates the probability that content was generated by artificial intelligence or by a human.

[1029] "Emotion data" is information that indicates the user's emotional state, and is composed of voice data, facial expression data, and the like.

[1030] "Emotion analysis" is a process of analyzing acquired emotion data and determining the user's emotional state (positive, negative, neutral, etc.).

[1031] The "evaluation rank" is a rank that quantitatively indicates the value and quality of content based on the extracted features and the results of sentiment analysis.

[1032] "Providing an evaluation" is the act of reporting the determined evaluation rank to the user.

[1033] This invention is a system that determines whether a content creator is an AI or a human, and provides the result to the user. This system has the function of analyzing the user's emotions and adjusting the evaluation results. Specifically, the system is configured using the following hardware and software.

[1034] Hardware and software used

[1035] The devices used by users include PCs, tablets, smartphones, etc. These devices have the ability to upload content. The devices communicate with the server via a web browser or dedicated application.

[1036] The server is the central computer that processes and analyzes the content it receives. The server has the following software and libraries installed:

[1037] Text processing libraries (e.g. NLTK, spaCy)

[1038] Machine learning libraries (e.g., Scikit-learn, TensorFlow)

[1039] Music informatics library (e.g., LibROSA)

[1040] Sentiment analysis API (e.g., general sentiment analysis services)

[1041] Uploading content

[1042] Users select and upload the content they want to rate from their device. For example, if a user wants to upload a novel manuscript file from their smartphone, they click the file upload button and then press the "Upload" button, which transfers the file to the server.

[1043] Content Preprocessing

[1044] The server preprocesses the received content. This preprocessing includes format conversion, removal of unnecessary information, and standardization. Specifically, it uses the Python libraries NLTK and BeautifulSoup to remove special characters and standardize encoding. For example, in the case of a text file, the server removes special characters (e.g., $, &,) from the text and converts it to UTF-8 encoding.

[1045] Feature Extraction

[1046] After preprocessing is complete, the server uses a machine learning model to extract features from the content. For text content, feature extraction involves extracting characteristics and patterns such as sentence length, word diversity, and word frequency. Specifically, libraries such as Scikit-learn and TensorFlow are used. For music content, music informatics libraries such as LibROSA are used to extract melody lines and rhythmic patterns.

[1047] Judgment and probability calculation

[1048] Based on the extracted features, the server uses a pre-trained machine learning model to calculate the probability of AI generation versus human generation. For example, it can produce a result such as "There is a 60% probability that this sentence was generated by an AI." This can be achieved using models such as neural networks and support vector machines (SVMs).

[1049] Emotion analysis

[1050] Users provide voice data and facial expression data via a camera. The server receives this data and uses an emotion analysis API to classify the user's emotional state. For example, when a user provides voice input, the microphone is used to collect emotion data, which is then sent to the server.

[1051] Determining the evaluation rank

[1052] Based on the results of the emotion analysis, the server determines the evaluation rank. The evaluation rank is dynamically determined based on the calculated probability of AI generation and the user's emotional state. For example, the evaluation rank may be adjusted as follows: "The probability of AI generation is 65% and the user's emotion is positive, so the evaluation rank is A."

[1053] Providing and displaying results

[1054] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, a web browser might display "AI generation probability is 65%, evaluation rank is A."

[1055] Specific examples

[1056] Example 1: Text content

[1057] 1. The user uploads their novel from their device to the server.

[1058] 2. The server normalizes the text (removes special characters and standardizes encoding).

[1059] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[1060] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[1061] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[1062] 6. The server adjusts the rating from B to A based on the results of the sentiment analysis.

[1063] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[1064] Example 2: Music content

[1065] 1. The user uploads the music file they composed from their device to the server.

[1066] 2. The server standardizes the sample rate and format of music files.

[1067] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[1068] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[1069] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[1070] 6. The server adjusts the rating from A to B based on the results of the sentiment analysis.

[1071] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[1072] Prompt Sentence Examples

[1073] "Is it likely that this novel was generated by AI?"

[1074] "Please rate this music. Please take into account the user's feelings."

[1075] The above description demonstrates how the system judges content creators and provides ratings that take user sentiment into account.

[1076] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1077] Step 1:

[1078] A user uploads content. Using their device, the user selects a text file or music file they want to rate and presses the upload button. The device then sends the selected content to the server. The input is the content file sent from the user's device, and the output is the content file uploaded to the server. Specifically, this includes the user selecting a novel text file from their smartphone and pressing the "upload" button.

[1079] Step 2:

[1080] This preprocesses the content received by the server, standardizing the format of the received content and removing unnecessary information. The input is the content file uploaded to the server, and the output is the preprocessed content data. Specifically, this includes using Python's NLTK and BeautifulSoup to remove special characters from the text and standardize the encoding to UTF-8.

[1081] Step 3:

[1082] The server extracts features from the preprocessed content. Based on the preprocessed data, features such as sentence length, word diversity, word frequency, melody line, and rhythmic patterns are analyzed. The input is the preprocessed content data, and the output is the extracted feature data. Specifically, this includes operations such as tokenizing text and generating feature vectors using Scikit-learn, and analyzing the frequency spectrum of music files using LibROSA.

[1083] Step 4:

[1084] The server calculates the probability of AI-generated or human-generated content based on the extracted features. It uses a machine learning model to calculate the probability that the content is AI-generated or human-generated. The input is the extracted feature data, and the output is the probability of AI-generated content and the probability of human-generated content. Specifically, this involves using a trained neural network model or support vector machine to calculate the probability.

[1085] Step 5:

[1086] The user provides emotion data. The user provides facial expression data using voice input or a camera. The input is the user's voice data and facial expression data, and the output is emotion data. Specifically, this includes operations such as recording voice with a microphone and sending it to a server.

[1087] Step 6:

[1088] The server analyzes the acquired emotional data and determines the user's emotional state. Using the emotion analysis API, the emotional state is classified as positive, negative, or neutral. The input is the provided emotional data, and the output is the determined emotional state. Specifically, this includes analyzing voice and facial expressions through the emotion analysis API and determining a "positive" state.

[1089] Step 7:

[1090] The server determines the evaluation rank based on the determined emotional state and the probability of AI generation and human generation. The evaluation is given as a rank from A to D. The inputs are the probability of AI generation and human generation, and the determined emotional state, and the output is the determined evaluation rank. Specifically, it includes an action to determine rank A when the probability of AI generation is high with positive emotions.

[1091] Step 8:

[1092] The server provides the evaluation rank to the user. The final judgment result and evaluation rank are sent to the user's device and displayed. The input is the determined evaluation rank, and the output is the evaluation result displayed on the user's device. Specifically, this includes generating an HTTP response and displaying in the browser, "There is a 65% probability that this content was generated by AI, and the evaluation rank is A."

[1093] (Application example 2)

[1094] 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."

[1095] Current content distribution services lack a mechanism for determining whether the content being viewed is generated by AI or humans, and for providing appropriate ratings based on the results. Furthermore, viewers' emotional state is not reflected in content ratings, and personalized ratings are not provided. This makes it difficult to accurately convey the reliability and value of content to viewers.

[1096] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI-generated or human-generated content based on the extracted features, means for recognizing the user's emotional state, means for adjusting the evaluation rank of the content based on the calculated probability and the recognized emotional state, and means for providing the evaluation rank to the user. This allows the user to not only determine whether the content they are viewing is AI-generated or human-generated, but also to receive a personalized evaluation based on their own emotional state.

[1097] The "means for receiving content from a user" is a mechanism for a user to send content created by the user using his / her own terminal to a server.

[1098] The "means for preprocessing received content" is a mechanism for standardizing the content received by the server into a specific format and performing processing to delete unnecessary information.

[1099] The "means for extracting features from preprocessed content" is a mechanism for analyzing the internal information of the preprocessed content and extracting its features as data.

[1100] "Means for calculating the probability of AI-generated or human-generated content based on extracted features" refers to algorithms and devices for calculating the probability that content is AI-generated or human-generated based on extracted features.

[1101] The "means for recognizing the user's emotional state" refers to a device and algorithm that analyzes the audio data and image data provided by the user when uploading content and determines the user's emotional state.

[1102] The "means for adjusting the evaluation rank of content based on the calculated probability and the recognized emotional state" refers to a mechanism that comprehensively evaluates the AI-generated probability of content and the user's emotional state, and determines and adjusts a visual or numerical evaluation rank.

[1103] The "means for providing the evaluation rank to the user" refers to a communication and display device for transmitting the finally determined evaluation rank to the user's terminal and displaying it.

[1104] This invention provides a system that determines whether the content a user is viewing is generated by AI or a human, and provides a rating based on the result. The system also has the ability to recognize the user's emotional state and adjust the rating based on that emotional state.

[1105] Hardware and software used

[1106] Hardware: Smartphone, webcam

[1107] Software: Python, OpenCV, Keras, scikit-learn

[1108] System configuration

[1109] 1. User's device: A device used by a user to upload content they create, such as a smartphone or tablet.

[1110] 2. Server: Performs key processing such as content preprocessing, feature extraction, probability calculation, emotion recognition, and rating adjustment.

[1111] Program processing

[1112] 1. A method for receiving content from users: Users upload text or music content to the server using devices such as smartphones or tablets. An example of user input is a prompt such as "Please rate this poem."

[1113] 2. A means of pre-processing the received content: The server standardizes the received content to a specific format and removes unnecessary information, such as removing special characters and standardizing encoding.

[1114] 3. Means for extracting features from the preprocessed content: The server analyzes the preprocessed content and extracts features such as sentence length and word diversity in the case of text, and melody lines and rhythmic patterns in the case of music.

[1115] 4. Calculating the probability of AI generation or human generation based on the extracted features: The server uses the extracted features to apply a trained machine learning model to calculate the probability that the content is AI-generated or human-generated. For example, the result may be "AI generation probability 65%, human generation probability 35%."

[1116] 5. Means for recognizing the user's emotional state: The server performs emotion recognition using the voice and image data provided by the user when uploading content. Using a model trained with OpenCV and Keras, the server determines the user's emotional state from their facial expressions and voice.

[1117] 6. Adjusting the content rating based on the calculated probability and the user's emotional state: The server determines the rating by taking into account the AI ​​generation rate and the user's emotional state, and adjusts the rating as necessary. For example, if the AI ​​generation rate is high and the user's emotion is positive, the rating will be adjusted to A.

[1118] 7. Means for providing the evaluation rank to the user: The final evaluation rank is sent to the user's terminal and displayed visually or numerically.

[1119] Specific examples

[1120] A user uploads a poem they have written to the server from their smartphone, and the text's characteristics are extracted. The system calculates the probability of AI generation at 65% and the probability of human generation at 35%. If the user's facial expression is recognized as "positive," the evaluation rank is adjusted to A, and the result is displayed on the user's device.

[1121] Prompt Sentence Examples

[1122] "Please enter the content of the poem you would like to rate."

[1123] "Please rate this music track."

[1124] In this way, the system of the present invention not only identifies the creator of the content and provides the result to the user, but also analyzes the user's emotional state and adjusts the evaluation result to provide a more personalized evaluation.

[1125] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1126] Step 1:

[1127] A user inputs text or music content from their own terminal. Specifically, they select a file through the terminal interface and press the upload button. The input is the text data or music file, which the terminal sends to the server. The output is the content data sent to the server.

[1128] Step 2:

[1129] The server preprocesses the received content. Specifically, for text content, it removes special characters and standardizes encoding, and for music content, it standardizes the sample rate. The input is the content data received from the user, and the output is the preprocessed, standardized content data.

[1130] Step 3:

[1131] The server extracts features from the preprocessed content. Specifically, in the case of text, features such as sentence length, word diversity, and frequent word list generation are extracted, while in the case of music, features such as melody line, beat pattern, and harmonic spectrum are analyzed. The input is the preprocessed content data, and the output is the extracted feature data.

[1132] Step 4:

[1133] The server calculates the probability of AI generation and human generation based on the extracted features. Specifically, it uses a machine learning model (e.g., a logistic regression model) to calculate the probability that the content was generated by AI and human generation. The input is feature data, and the output is the probability of AI generation and human generation.

[1134] Step 5:

[1135] When uploading content, users provide voice input or facial expression data via a camera. The device collects this data and sends it to the server. The input is voice data or image data, and the output is emotional data sent to the server.

[1136] Step 6:

[1137] The server recognizes the user's emotional state. Specifically, it uses a model trained with OpenCV and Keras to analyze and classify the user's emotions from audio and image data. The input is the emotional data received from the user, and the output is the recognized emotional state (e.g., positive, negative, neutral).

[1138] Step 7:

[1139] The server adjusts the content's rating based on the calculated AI generation rate and the recognized emotional state. For example, if the AI ​​generation rate is high and the user's emotion is positive, the server will raise the rating. The inputs are the AI ​​generation rate, the human generation rate, and the recognized emotional state, and the output is the adjusted rating.

[1140] Step 8:

[1141] The server provides the evaluation rank to the user. Specifically, it transmits the evaluation result to the user's terminal and displays the evaluation rank visually or numerically. The input is the adjusted evaluation rank, and the output is the evaluation rank displayed on the user's terminal.

[1142] 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.

[1143] 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.

[1144] 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.

[1145] [Fourth embodiment]

[1146] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1147] 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.

[1148] 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).

[1149] 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.

[1150] 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.

[1151] 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).

[1152] 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.

[1153] 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.

[1154] 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.

[1155] 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.

[1156] 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.

[1157] 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.

[1158] 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."

[1159] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of the program for this system will be described below.

[1160] Overall overview

[1161] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. The process involves the following main steps:

[1162] Uploading content

[1163] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[1164] Content Preprocessing

[1165] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[1166] Feature Extraction

[1167] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[1168] Judgment and probability calculation

[1169] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[1170] Rating ranking

[1171] Based on the calculated probability, the server determines the content's rating. This rating is based on pre-set criteria and is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[1172] Providing and displaying results

[1173] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is B."

[1174] Specific examples

[1175] Example 1: Text content

[1176] 1. The user uploads their novel from their device to the server.

[1177] 2. The server normalizes the text (removes special characters and standardizes encoding).

[1178] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[1179] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[1180] 5. The server determines that there is a high probability of AI generation and assigns a rank of C.

[1181] 6. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[1182] Example 2: Music content

[1183] 1. The user uploads the music file they composed from their device to the server.

[1184] 2. The server standardizes the sample rate and format of music files.

[1185] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[1186] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[1187] 5. The server determines that there is a high probability of generating a human and assigns a rank of A.

[1188] 6. The server sends the results to the device, and the user confirms that "There is a 60% chance that this music was created by a human, and it has an evaluation rank of A."

[1189] In this way, the system of the present invention accurately identifies the creator of content and provides the result to the user, thereby providing an environment in which content can be used with peace of mind.

[1190] The processing flow will be explained below.

[1191] Step 1:

[1192] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[1193] Step 2:

[1194] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[1195] Step 3:

[1196] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, etc. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[1197] Step 4:

[1198] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[1199] Step 5:

[1200] The server assigns an evaluation rank based on the calculated probability. Based on pre-set criteria, it determines an evaluation rank such as "A," "B," "C," or "D" for each piece of content. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if the probability of human generation is 70% or higher, it will be ranked A.

[1201] Step 6:

[1202] The server sends the evaluation results and rank to the device. It then creates a message containing the generated results and evaluation rank and sends it back to the device. A specific example would be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank is B."

[1203] Step 7:

[1204] The user checks the results on their device. The device displays the received message, allowing the user to see whether their content was generated by AI or a human, as well as its rating. For example, it might say, "There's a 70% chance this text was created by AI, and it has a rating of B."

[1205] Example 1

[1206] 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."

[1207] Conventional content evaluation systems lack an effective method for determining whether content is generated by AI or humans. This makes it difficult for users to judge the reliability of content. While identifying the creator is particularly important for creative works such as text and music, current technology is unable to adequately address this issue.

[1208] 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.

[1209] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, and means for providing the evaluation rank to the user. This makes it possible to determine with high accuracy whether content was generated by AI or human and provide the result to the user.

[1210] "User" refers to an entity that uploads content for evaluation using the system.

[1211] "Content" refers to data such as text or music files that users upload to a server and wish to be evaluated.

[1212] "Terminal" refers to devices used by users, such as computers, tablets, and smartphones.

[1213] "Server" refers to the core part of the system that receives content, pre-processes it, extracts features, judges it, determines its rating, and provides the results.

[1214] "Preprocessing" refers to processing such as standardizing the format and deleting unnecessary information that the server performs on the content it receives.

[1215] "Features" refer to the data characteristics used by machine learning models to determine, for example, sentence length, word diversity, or melody lines extracted from preprocessed content.

[1216] "Extraction" refers to the operation of extracting features from preprocessed content.

[1217] "Probability" refers to the percentage of likelihood that the content was generated by AI or a human, calculated by the server based on the extracted features.

[1218] The "rating rank" refers to a grade indicating the level of the content's rating, determined by the server based on the calculated probability.

[1219] This invention relates to a system that determines whether a content creator is an AI or a human, and provides the result to the user. A specific implementation method of this system will be described below.

[1220] Uploading content

[1221] Users use their own devices (PCs, tablets, smartphones, etc.) to select and upload the content they want to rate. For example, a user might select a file such as "my_novel.txt" on their PC. The device then imports the selected content and sends it to the server.

[1222] Content Preprocessing

[1223] The server pre-processes the received content. This pre-processing includes the following specific steps:

[1224] Text format conversion: The server standardizes the format to UTF-8. For example, convert "my_novel.txt" to UTF-8.

[1225] Remove unnecessary information: Remove special characters and HTML tags. For example, 」「 Remove tags such as ".

[1226] Standardize data: Standardize line breaks and spaces. For example, standardize multiple spaces into one.

[1227] Feature Extraction

[1228] The server then uses the machine learning model to extract features from the pre-processed content. This process includes the following specific steps:

[1229] Text analysis: The server extracts sentence length, word variety, and word frequency. For example, from "my_novel.txt" we measure the average sentence length and the number of unique words.

[1230] Music analysis: The server analyzes the melody line and beat pattern of the music file. For example, it obtains the melody pitch from "my_music.mp3".

[1231] Judgment and probability calculation

[1232] The server uses machine learning models to calculate the decision and probability based on the extracted features. This process includes the following specific steps:

[1233] Calculating AI generation rate: The server uses the trained model to calculate the probability of AI generation and the probability of human generation. For example, analyzing "my_novel.txt" determines that 65% of the text is AI-generated.

[1234] Probability determination: The server determines the probability as a result. For example, "65% probability of AI generation, 35% probability of human generation."

[1235] Rating ranking

[1236] The server determines the reputation rank based on the calculated probability. This process includes the following specific steps:

[1237] Criteria-based ranking: The server assigns a ranking based on pre-defined criteria. For example, if the probability of AI generation is 50% or higher, it will be ranked B, and if it is 70% or higher, it will be ranked A.

[1238] Generating evaluation results: The server generates evaluation results such as "There is a 65% chance that this content was AI-generated, and the evaluation rank is C."

[1239] Providing and displaying results

[1240] The server then sends the final judgment result and evaluation rank to the user's device. This process includes the following specific steps:

[1241] Sending the result: The server sends the evaluation result to the user's terminal. For example, it sends the evaluation result of "my_novel.txt".

[1242] Display of results: The user checks the evaluation results on their device. For example, the user's screen may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[1243] Examples of prompt statements

[1244] "There is a 60% chance that this sentence was created by an AI. Please rate it."

[1245] "Please tell me who created this music file and show me the probability."

[1246] As described above, this system can analyze content uploaded by users with high accuracy and determine whether the content was generated by AI or humans. Based on these results, it determines an evaluation rank and provides it to users to help them evaluate the reliability of the content.

[1247] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1248] Step 1:

[1249] A user selects the content they want to rate on their device and uploads it. For example, a user selects a text file called "my_novel.txt" on their computer and performs the upload. The device sends the selected file to the server. The input is the content selected by the user, and the output is the content sent to the server.

[1250] Step 2:

[1251] The server preprocesses the received content. This preprocessing includes converting text formats, removing unnecessary information, and standardizing data. For example, converting "my_novel.txt" to UTF-8, removing special characters and HTML tags, and standardizing line breaks and spaces. The input is the uploaded content, and the output is the preprocessed content.

[1252] Step 3:

[1253] The server extracts features from the preprocessed content. In this step, a machine learning model is used to analyze text such as sentence length, word diversity, and word frequency. For example, the average sentence length and number of unique words are measured from "my_novel.txt." The input is the preprocessed content, and the output is the extracted feature data.

[1254] Step 4:

[1255] The server calculates the probability of AI-generated versus human-generated sentences based on the extracted feature data. In this step, a pre-trained machine learning model is used to generate results such as "This sentence has a 65% probability of being AI-generated and a 35% probability of being human-generated." The input is the extracted feature data, and the output is the probability of AI-generated versus human-generated sentences.

[1256] Step 5:

[1257] The server determines the evaluation rank based on the calculated probability. In this step, a rank is assigned according to the probability based on pre-set criteria. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if it is 70% or higher, it is ranked A. The input is the probability of AI generation and human generation, and the output is the determined evaluation rank.

[1258] Step 6:

[1259] The server sends the final judgment result and evaluation rank to the user's device. In this step, the evaluation result is sent to the user's device so that the user can check the result. For example, it may say, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C." The input is the evaluation rank and judgment result, and the output is the result displayed on the user's device.

[1260] (Application example 1)

[1261] 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."

[1262] In recent years, advances in generative AI models have made it difficult to distinguish between human-generated and AI-generated content. This has made it difficult for users to find reliable information sources. In particular, there is a demand for mechanisms to evaluate the reliability of content in content distribution services. The objective of this invention is to solve this problem and provide an environment in which users can use content with peace of mind.

[1263] 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.

[1264] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation or human generation based on the extracted features, means for determining an evaluation rank for the content based on the calculated probability, means for providing the evaluation rank to the user, and means for displaying the result as an application installed on the smart device. This makes it possible to accurately determine whether content uploaded by a user was generated by AI or a human, and visualize the result on the smart device.

[1265] A "user" is an operator of a terminal that generates content or uploads content for evaluation.

[1266] "Content" refers to information expressions such as text, music, and images.

[1267] A "receiving means" is a device or system that has the function of transmitting content from a user to a server.

[1268] A "preprocessing means" is a system that has processing capabilities to standardize received content and remove unnecessary information.

[1269] A "means for extracting features" is a system that has the function of extracting specific patterns or characteristics from content.

[1270] The "means for calculating probability" is a system that calculates the probability of AI-generated or human-generated based on the extracted features.

[1271] The "means for determining an evaluation rank" is a system having a function of assigning an evaluation rank to content based on probability.

[1272] The "means for displaying results" is an application that has the function of visualizing the evaluation rank and the creator's judgment results to the user on a smart device.

[1273] A "smart device" is a mobile device that can connect to the Internet, such as a smartphone, tablet, or smart glasses.

[1274] The present invention relates to a system that determines whether user-generated content is created by AI or a human, and provides the result as an evaluation rank. The following describes an embodiment of the present invention.

[1275] System Overview

[1276] The system consists of the following main components:

[1277] 1. User's device: The smart device (smartphone, tablet, PC, etc.) used to upload content.

[1278] 2. Server: Performs content preprocessing, feature extraction, judgment, and evaluation ranking.

[1279] 3. Smart device application: An interface that displays the evaluation results to the user.

[1280] Specific hardware and software

[1281] This system uses the following hardware and software:

[1282] Server: A server with high-performance computing resources. For example, you can use a cloud server such as Amazon Web Services (AWS) or your own server.

[1283] User device: Any smart device that can connect to the Internet, including iPhones, Android smartphones, iPads, and Windows PCs.

[1284] Smart device applications: Implemented as iOS apps, Android apps, web applications, etc.

[1285] Program processing

[1286] In this system, the following processing is performed.

[1287] 1. Uploading Content

[1288] Content is uploaded from the user's terminal to the server through a smart device application. Examples include short stories, music files, and image files.

[1289] 2. Content Preprocessing

[1290] The server standardizes the received content and removes unnecessary information, including standardizing text encoding and removing special characters.

[1291] 3. Feature Extraction

[1292] Features are extracted from the preprocessed content: for text, this includes sentence length, word diversity, and word frequency list generation; for music files, this involves analysis of melody line, beat pattern, and harmonic spectrum.

[1293] 4. Judgment and evaluation ranking

[1294] Based on the extracted features, the server calculates the probability of whether the sentence is AI-generated or human-generated. To do this, a pre-trained generative AI model is used. For example, for a particular sentence, the result may be "60% probability of AI generation" or "40% probability of human generation." Based on the probability, the evaluation rank is determined as "A," "B," "C," or "D."

[1295] 5. Providing and displaying results

[1296] The final evaluation rank and judgment results are provided to the user via a smart device application. Users can view specific feedback on their devices, such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C."

[1297] Examples and Prompts

[1298] For example, a user uploads the following short story:

[1299] There stands an old tree in the forest. For hundreds of years, many birds have gathered around this tree, watching over the passage of time.

[1300] This prompt sentence is sent to the server, and after going through the steps above, a result such as "There is a 65% chance that this sentence was created by AI, and the evaluation rank is C" is obtained. This result is displayed in the application on the smart device.

[1301] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1302] Step 1:

[1303] A user uploads content.

[1304] Users use their devices (smartphones, tablets, PCs, etc.) to select the content they want to rate and send it to the server via an upload form.

[1305] Input: User-uploaded content (text, music, images, etc.)

[1306] Output: Uploaded content is saved on the server

[1307] Step 2:

[1308] The server pre-processes the content.

[1309] The server standardizes the received content and removes unnecessary information, for example, standardizing text encoding and removing special characters.

[1310] Input: User-uploaded content

[1311] Output: Standardized and cleaned content

[1312] Step 3:

[1313] The server extracts the features.

[1314] The server extracts certain patterns and characteristics from the pre-processed content: for text, this includes sentence length, word diversity, and generating a list of frequent words, while for music files, it analyzes melody lines, beat patterns, and harmonic spectrum.

[1315] Input: Preprocessed content

[1316] Output: Patterns and characteristics extracted as content features

[1317] Step 4:

[1318] The server calculates the probability.

[1319] Based on the extracted features, the server calculates the probability of AI or human generation using a pre-trained generative AI model. For example, the server can give results such as "60% probability of AI generation" or "40% probability of human generation."

[1320] Input: Extracted features

[1321] Output: AI-generated and human-generated probabilities

[1322] Step 5:

[1323] The server determines the evaluation rank.

[1324] Based on the calculated probability, the content is assigned an evaluation rank. For example, if the probability of AI generation is 50% or higher, it is ranked B, and if the probability of human generation is 70% or higher, it is ranked A.

[1325] Input: AI-generated and human-generated probabilities

[1326] Output: Evaluation rank (A, B, C, D, etc.)

[1327] Step 6:

[1328] The server provides the results.

[1329] The final judgment result and evaluation rank are sent to the user's device, where the user can check the results.

[1330] Input: Evaluation rank and generator probability

[1331] Output: Display of results on the user's device (judgment result and evaluation rank)

[1332] Step 7:

[1333] The user checks the results.

[1334] The user can check the results on their device's application or web page, and receive feedback such as, "There is a 65% chance that this sentence was created by AI, and it has been given a rating of C."

[1335] Input: Final result sent from the server

[1336] Output: Evaluation rank and judgment result for user confirmation

[1337] 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.

[1338] This invention combines an emotion engine with a system that determines whether the creator of content is an AI or a human and provides the result to the user. Specific implementation methods for the program of this system are described below.

[1339] Overall overview

[1340] In this system, users upload content via their devices, and the server processes the content and provides evaluation results. It also incorporates an emotion engine that recognizes the user's emotions. In addition to the traditional evaluation process, the emotion engine analyzes the user's emotional state and adjusts the evaluation results accordingly.

[1341] Uploading content

[1342] Users select and upload the content they want to rate using their own devices (PC, tablet, smartphone, etc.). The device has the function to send the selected content to the server.

[1343] Content Preprocessing

[1344] The server pre-processes the received content, which may include formatting, removing unnecessary information, and standardizing the content (e.g., removing special characters and standardizing encoding for text content).

[1345] Feature Extraction

[1346] After preprocessing is complete, the server uses a machine learning model to extract features from the content. Feature extraction is performed differently depending on the type of content. For text, features extracted include sentence length, word diversity, and sentence structure analysis. For music, features extracted include melody lines, beat patterns, and harmonic spectrum.

[1347] Judgment and probability calculation

[1348] Based on the extracted features, the server calculates the probability that the content is AI-generated or human-generated using a pre-trained machine learning model. For example, for a given sentence, it can give results such as "60% probability of AI generation" or "40% probability of human generation."

[1349] Emotional engine regulation

[1350] Before determining the rating rank, the server uses an emotion engine to recognize the user's emotions. When users upload content, they provide voice input or facial expression data via a camera. The server analyzes this data and classifies the user's emotional state (positive, negative, neutral, etc.). The server adjusts the rating rank based on the output of the emotion engine.

[1351] Rating ranking

[1352] Based on the adjusted results, the server determines the final evaluation rank, which is expressed as a rank such as "A," "B," "C," or "D." For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, it will be ranked B, and if it is negative, it will be ranked C.

[1353] Providing and displaying results

[1354] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, it may say, "There is a 60% chance that this sentence was created by AI, and the evaluation rank is C based on the user's sentiment."

[1355] Specific examples

[1356] Example 1: Text content

[1357] 1. The user uploads their novel from their device to the server.

[1358] 2. The server normalizes the text (removes special characters and standardizes encoding).

[1359] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[1360] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[1361] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[1362] 6. The server adjusts the evaluation rank from B to A based on the output of the emotion engine.

[1363] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[1364] Example 2: Music content

[1365] 1. The user uploads the music file they composed from their device to the server.

[1366] 2. The server standardizes the sample rate and format of music files.

[1367] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[1368] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[1369] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[1370] 6. The server adjusts the evaluation rank from A to B based on the output of the emotion engine.

[1371] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[1372] In this way, the system of the present invention not only accurately identifies the creator of the content and provides the result to the user, but also analyzes the user's emotions and adjusts the evaluation results, thereby providing a more personalized evaluation.

[1373] The processing flow will be explained below.

[1374] Step 1:

[1375] Users upload the content they want to rate from their device to the server. Specifically, users select text, music, video files, etc. on the file selection screen on their device and press the upload button. This operation sends the selected files from the device to the server.

[1376] Step 2:

[1377] The server checks the type of content received and performs preprocessing. For text files, the encoding is standardized to UTF-8 and special characters and unnecessary tags are removed. For music files, the sample rate and bit rate are standardized. For video files, the resolution is adjusted and the frame rate is standardized.

[1378] Step 3:

[1379] The server extracts features from the pre-processed content. For text content, it analyzes sentence length, word diversity, grammatical structure, and frequency word lists. For music content, it analyzes melody lines, beat patterns, and harmonic spectrum. For video content, it extracts scene changes, color distribution, and movement patterns.

[1380] Step 4:

[1381] The server calculates the AI-generated and human-generated probabilities based on the extracted features. Using a pre-trained machine learning model, it inputs a feature vector and outputs the AI-generated and human-generated probabilities. For example, for text content, it can output something like "70% probability of AI generation" and "30% probability of human generation."

[1382] Step 5:

[1383] The user provides emotional data to the emotion engine. When uploading content, the user uses a device (microphone, camera) to capture voice and facial expressions. The device then sends the collected voice and image data to the server.

[1384] Step 6:

[1385] The server uses an emotion engine to analyze the user's emotional state. For voice data, an emotion analysis algorithm is used to identify emotional categories such as positive, negative, or neutral from the tone and patterns of the voice. For image data, an expression recognition algorithm is used to recognize the emotional state from facial expressions and subtle movements.

[1386] Step 7:

[1387] The server adjusts the evaluation rank based on the output of the emotion engine. For example, if the probability of AI generation is 50% or higher and the user's emotion is positive, the evaluation rank is adjusted from B to A. If the user's emotion is negative, the evaluation rank is adjusted from B to C.

[1388] Step 8:

[1389] The server sends the final evaluation result and rank to the device. A message containing the generated result and evaluation rank is created and sent back to the device. For example, it could be something like, "There is a 70% chance that this text was created by AI, and the evaluation rank based on the user's sentiment is A."

[1390] Step 9:

[1391] The user checks the results on the device. The device displays the received message, allowing the user to check whether their content was generated by AI or a human, as well as its rating. For example, the device may display a message such as, "There is a 70% chance that this text was created by AI, and the rating based on user sentiment is A."

[1392] Example 2

[1393] 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."

[1394] Conventional content rating systems only distinguish whether the creator is an AI or a human, and are unable to adjust the rating taking into account the user's emotional state. As a result, ratings that do not reflect the individual needs of users based on their emotions are not provided, resulting in problems with reduced accuracy and reliability of the ratings. Furthermore, the user experience is uniform, making it difficult to provide individual responses.

[1395] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1396] In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI generation and human generation based on the extracted features, means for acquiring user emotional data as input, means for analyzing the acquired emotional data and determining the user's emotional state, means for determining an evaluation rank based on the determined emotional state and the calculated probability of AI generation and human generation, and means for providing the evaluation rank to the user. This enables evaluation that takes user emotions into consideration and provides evaluation results that can meet individual requests.

[1397] A "server" is a computer that processes data over a network and provides services and resources to other computer systems.

[1398] A "terminal" is a device operated by a user, and is a hardware device such as a computer, tablet, or smartphone that inputs and displays content.

[1399] A "user" is a person who uses the system, uploads content, and receives the results.

[1400] "Content" refers to information resources such as text and music, and is the data or files that are the subject of evaluation.

[1401] "Preprocessing" is a general term for the process of standardizing received content and removing unnecessary information.

[1402] "Feature extraction" is the process of extracting characteristics and patterns from preprocessed content that are necessary for evaluation.

[1403] "Probability of AI vs. human generation" is a number that indicates the probability that content was generated by artificial intelligence or by a human.

[1404] "Emotion data" is information that indicates the user's emotional state, and is composed of voice data, facial expression data, and the like.

[1405] "Emotion analysis" is a process of analyzing acquired emotion data and determining the user's emotional state (positive, negative, neutral, etc.).

[1406] The "evaluation rank" is a rank that quantitatively indicates the value and quality of content based on the extracted features and the results of sentiment analysis.

[1407] "Providing an evaluation" is the act of reporting the determined evaluation rank to the user.

[1408] This invention is a system that determines whether a content creator is an AI or a human, and provides the result to the user. This system has the function of analyzing the user's emotions and adjusting the evaluation results. Specifically, the system is configured using the following hardware and software.

[1409] Hardware and software used

[1410] The devices used by users include PCs, tablets, smartphones, etc. These devices have the ability to upload content. The devices communicate with the server via a web browser or dedicated application.

[1411] The server is the central computer that processes and analyzes the content it receives. The server has the following software and libraries installed:

[1412] Text processing libraries (e.g. NLTK, spaCy)

[1413] Machine learning libraries (e.g., Scikit-learn, TensorFlow)

[1414] Music informatics library (e.g., LibROSA)

[1415] Sentiment analysis API (e.g., general sentiment analysis services)

[1416] Uploading content

[1417] Users select and upload the content they want to rate from their device. For example, if a user wants to upload a novel manuscript file from their smartphone, they click the file upload button and then press the "Upload" button, which transfers the file to the server.

[1418] Content Preprocessing

[1419] The server preprocesses the received content. This preprocessing includes format conversion, removal of unnecessary information, and standardization. Specifically, it uses the Python libraries NLTK and BeautifulSoup to remove special characters and standardize encoding. For example, in the case of a text file, the server removes special characters (e.g., $, &,) from the text and converts it to UTF-8 encoding.

[1420] Feature Extraction

[1421] After preprocessing is complete, the server uses a machine learning model to extract features from the content. For text content, feature extraction involves extracting characteristics and patterns such as sentence length, word diversity, and word frequency. Specifically, libraries such as Scikit-learn and TensorFlow are used. For music content, music informatics libraries such as LibROSA are used to extract melody lines and rhythmic patterns.

[1422] Judgment and probability calculation

[1423] Based on the extracted features, the server uses a pre-trained machine learning model to calculate the probability of AI generation versus human generation. For example, it can produce a result such as "There is a 60% probability that this sentence was generated by an AI." This can be achieved using models such as neural networks and support vector machines (SVMs).

[1424] Emotion analysis

[1425] Users provide voice data and facial expression data via a camera. The server receives this data and uses an emotion analysis API to classify the user's emotional state. For example, when a user provides voice input, the microphone is used to collect emotion data, which is then sent to the server.

[1426] Determining the evaluation rank

[1427] Based on the results of the emotion analysis, the server determines the evaluation rank. The evaluation rank is dynamically determined based on the calculated probability of AI generation and the user's emotional state. For example, the evaluation rank may be adjusted as follows: "The probability of AI generation is 65% and the user's emotion is positive, so the evaluation rank is A."

[1428] Providing and displaying results

[1429] The final judgment result and evaluation rank are sent from the server to the device. The user can check the results on their device. For example, a web browser might display "AI generation probability is 65%, evaluation rank is A."

[1430] Specific examples

[1431] Example 1: Text content

[1432] 1. The user uploads their novel from their device to the server.

[1433] 2. The server normalizes the text (removes special characters and standardizes encoding).

[1434] 3. The server analyzes features such as sentence length, word diversity, and extraction of frequently occurring words.

[1435] 4. The server uses a machine learning model to calculate the probability of generating an AI (65%) and a human (35%).

[1436] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is positive.

[1437] 6. The server adjusts the rating from B to A based on the results of the sentiment analysis.

[1438] 7. The server sends the results to the device, and the user confirms, "There is a 65% chance that this sentence was created by AI, and the evaluation rank is A."

[1439] Example 2: Music content

[1440] 1. The user uploads the music file they composed from their device to the server.

[1441] 2. The server standardizes the sample rate and format of music files.

[1442] 3. The server analyzes the melody line, rhythmic patterns, and harmonic characteristics.

[1443] 4. The server uses a machine learning model to calculate the probability of generating an AI (40%) and a human (60%).

[1444] 5. The user provides emotional data (voice and facial expressions), and the server determines that the emotion is negative.

[1445] 6. The server adjusts the rating from A to B based on the results of the sentiment analysis.

[1446] 7. The server sends the results to the device, and the user confirms, "There is a 60% chance that this music was created by a human, and it has a rating of B."

[1447] Prompt Sentence Examples

[1448] "Is it likely that this novel was generated by AI?"

[1449] "Please rate this music. Please take into account the user's feelings."

[1450] The above description demonstrates how the system judges content creators and provides ratings that take user sentiment into account.

[1451] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1452] Step 1:

[1453] A user uploads content. Using their device, the user selects a text file or music file they want to rate and presses the upload button. The device then sends the selected content to the server. The input is the content file sent from the user's device, and the output is the content file uploaded to the server. Specifically, this includes the user selecting a novel text file from their smartphone and pressing the "upload" button.

[1454] Step 2:

[1455] This preprocesses the content received by the server, standardizing the format of the received content and removing unnecessary information. The input is the content file uploaded to the server, and the output is the preprocessed content data. Specifically, this includes using Python's NLTK and BeautifulSoup to remove special characters from the text and standardize the encoding to UTF-8.

[1456] Step 3:

[1457] The server extracts features from the preprocessed content. Based on the preprocessed data, features such as sentence length, word diversity, word frequency, melody line, and rhythmic patterns are analyzed. The input is the preprocessed content data, and the output is the extracted feature data. Specifically, this includes operations such as tokenizing text and generating feature vectors using Scikit-learn, and analyzing the frequency spectrum of music files using LibROSA.

[1458] Step 4:

[1459] The server calculates the probability of AI-generated or human-generated content based on the extracted features. It uses a machine learning model to calculate the probability that the content is AI-generated or human-generated. The input is the extracted feature data, and the output is the probability of AI-generated content and the probability of human-generated content. Specifically, this involves using a trained neural network model or support vector machine to calculate the probability.

[1460] Step 5:

[1461] The user provides emotion data. The user provides facial expression data using voice input or a camera. The input is the user's voice data and facial expression data, and the output is emotion data. Specifically, this includes operations such as recording voice with a microphone and sending it to a server.

[1462] Step 6:

[1463] The server analyzes the acquired emotional data and determines the user's emotional state. Using the emotion analysis API, the emotional state is classified as positive, negative, or neutral. The input is the provided emotional data, and the output is the determined emotional state. Specifically, this includes analyzing voice and facial expressions through the emotion analysis API and determining a "positive" state.

[1464] Step 7:

[1465] The server determines the evaluation rank based on the determined emotional state and the probability of AI generation and human generation. The evaluation is given as a rank from A to D. The inputs are the probability of AI generation and human generation, and the determined emotional state, and the output is the determined evaluation rank. Specifically, it includes an action to determine rank A when the probability of AI generation is high with positive emotions.

[1466] Step 8:

[1467] The server provides the evaluation rank to the user. The final judgment result and evaluation rank are sent to the user's device and displayed. The input is the determined evaluation rank, and the output is the evaluation result displayed on the user's device. Specifically, this includes generating an HTTP response and displaying in the browser, "There is a 65% probability that this content was generated by AI, and the evaluation rank is A."

[1468] (Application example 2)

[1469] 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."

[1470] Current content distribution services lack a mechanism for determining whether the content being viewed is generated by AI or humans, and for providing appropriate ratings based on the results. Furthermore, viewers' emotional state is not reflected in content ratings, and personalized ratings are not provided. This makes it difficult to accurately convey the reliability and value of content to viewers.

[1471] The identification processing by the identification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for receiving content from a user, means for preprocessing the received content, means for extracting features from the preprocessed content, means for calculating the probability of AI-generated or human-generated content based on the extracted features, means for recognizing the user's emotional state, means for adjusting the evaluation rank of the content based on the calculated probability and the recognized emotional state, and means for providing the evaluation rank to the user. This allows the user to not only determine whether the content they are viewing is AI-generated or human-generated, but also to receive a personalized evaluation based on their own emotional state.

[1472] The "means for receiving content from a user" is a mechanism for a user to send content created by the user using his / her own terminal to a server.

[1473] The "means for preprocessing received content" is a mechanism for standardizing the content received by the server into a specific format and performing processing to delete unnecessary information.

[1474] The "means for extracting features from preprocessed content" is a mechanism for analyzing the internal information of the preprocessed content and extracting its features as data.

[1475] "Means for calculating the probability of AI-generated or human-generated content based on extracted features" refers to algorithms and devices for calculating the probability that content is AI-generated or human-generated based on extracted features.

[1476] The "means for recognizing the user's emotional state" refers to a device and algorithm that analyzes the audio data and image data provided by the user when uploading content and determines the user's emotional state.

[1477] The "means for adjusting the evaluation rank of content based on the calculated probability and the recognized emotional state" refers to a mechanism that comprehensively evaluates the AI-generated probability of content and the user's emotional state, and determines and adjusts a visual or numerical evaluation rank.

[1478] The "means for providing the evaluation rank to the user" refers to a communication and display device for transmitting the finally determined evaluation rank to the user's terminal and displaying it.

[1479] This invention provides a system that determines whether the content a user is viewing is generated by AI or a human, and provides a rating based on the result. The system also has the ability to recognize the user's emotional state and adjust the rating based on that emotional state.

[1480] Hardware and software used

[1481] Hardware: Smartphone, webcam

[1482] Software: Python, OpenCV, Keras, scikit-learn

[1483] System configuration

[1484] 1. User's device: A device used by a user to upload content they create, such as a smartphone or tablet.

[1485] 2. Server: Performs key processing such as content preprocessing, feature extraction, probability calculation, emotion recognition, and rating adjustment.

[1486] Program processing

[1487] 1. A method for receiving content from users: Users upload text or music content to the server using devices such as smartphones or tablets. An example of user input is a prompt such as "Please rate this poem."

[1488] 2. A means of pre-processing the received content: The server standardizes the received content to a specific format and removes unnecessary information, such as removing special characters and standardizing encoding.

[1489] 3. Means for extracting features from the preprocessed content: The server analyzes the preprocessed content and extracts features such as sentence length and word diversity in the case of text, and melody lines and rhythmic patterns in the case of music.

[1490] 4. Calculating the probability of AI generation or human generation based on the extracted features: The server uses the extracted features to apply a trained machine learning model to calculate the probability that the content is AI-generated or human-generated. For example, the result may be "AI generation probability 65%, human generation probability 35%."

[1491] 5. Means for recognizing the user's emotional state: The server performs emotion recognition using the voice and image data provided by the user when uploading content. Using a model trained with OpenCV and Keras, the server determines the user's emotional state from their facial expressions and voice.

[1492] 6. Adjusting the content rating based on the calculated probability and the user's emotional state: The server determines the rating by taking into account the AI ​​generation rate and the user's emotional state, and adjusts the rating as necessary. For example, if the AI ​​generation rate is high and the user's emotion is positive, the rating will be adjusted to A.

[1493] 7. Means for providing the evaluation rank to the user: The final evaluation rank is sent to the user's terminal and displayed visually or numerically.

[1494] Specific examples

[1495] A user uploads a poem they have written to the server from their smartphone, and the text's characteristics are extracted. The system calculates the probability of AI generation at 65% and the probability of human generation at 35%. If the user's facial expression is recognized as "positive," the evaluation rank is adjusted to A, and the result is displayed on the user's device.

[1496] Prompt Sentence Examples

[1497] "Please enter the content of the poem you would like to rate."

[1498] "Please rate this music track."

[1499] In this way, the system of the present invention not only identifies the creator of the content and provides the result to the user, but also analyzes the user's emotional state and adjusts the evaluation result to provide a more personalized evaluation.

[1500] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1501] Step 1:

[1502] A user inputs text or music content from their own terminal. Specifically, they select a file through the terminal interface and press the upload button. The input is the text data or music file, which the terminal sends to the server. The output is the content data sent to the server.

[1503] Step 2:

[1504] The server preprocesses the received content. Specifically, for text content, it removes special characters and standardizes encoding, and for music content, it standardizes the sample rate. The input is the content data received from the user, and the output is the preprocessed, standardized content data.

[1505] Step 3:

[1506] The server extracts features from the preprocessed content. Specifically, in the case of text, features such as sentence length, word diversity, and frequent word list generation are extracted, while in the case of music, features such as melody line, beat pattern, and harmonic spectrum are analyzed. The input is the preprocessed content data, and the output is the extracted feature data.

[1507] Step 4:

[1508] The server calculates the probability of AI generation and human generation based on the extracted features. Specifically, it uses a machine learning model (e.g., a logistic regression model) to calculate the probability that the content was generated by AI and human generation. The input is feature data, and the output is the probability of AI generation and human generation.

[1509] Step 5:

[1510] When uploading content, users provide voice input or facial expression data via a camera. The device collects this data and sends it to the server. The input is voice data or image data, and the output is emotional data sent to the server.

[1511] Step 6:

[1512] The server recognizes the user's emotional state. Specifically, it uses a model trained with OpenCV and Keras to analyze and classify the user's emotions from audio and image data. The input is the emotional data received from the user, and the output is the recognized emotional state (e.g., positive, negative, neutral).

[1513] Step 7:

[1514] The server adjusts the content's rating based on the calculated AI generation rate and the recognized emotional state. For example, if the AI ​​generation rate is high and the user's emotion is positive, the server will raise the rating. The inputs are the AI ​​generation rate, the human generation rate, and the recognized emotional state, and the output is the adjusted rating.

[1515] Step 8:

[1516] The server provides the evaluation rank to the user. Specifically, it transmits the evaluation result to the user's terminal and displays the evaluation rank visually or numerically. The input is the adjusted evaluation rank, and the output is the evaluation rank displayed on the user's terminal.

[1517] 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.

[1518] 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.

[1519] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[1520] 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.

[1521] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1522] 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.

[1523] 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).

[1524] 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.

[1525] 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."

[1526] 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.

[1527] 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).

[1528] 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.

[1529] 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.

[1530] 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.

[1531] 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.

[1532] 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.

[1533] 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.

[1534] 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.

[1535] 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.

[1536] 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.

[1537] 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.

[1538] The following is further disclosed regarding the above embodiment.

[1539] (Claim 1)

[1540] To identify the content creator,

[1541] means for receiving content from a user;

[1542] means for preprocessing the received content;

[1543] means for extracting features from the preprocessed content;

[1544] A means for calculating the probability of AI generation and human generation based on the extracted features;

[1545] a means for determining an evaluation rank of the content based on the calculated probability;

[1546] A means for providing an evaluation rank to a user;

[1547] A system including:

[1548] (Claim 2)

[1549] 10. The system of claim 1, wherein preprocessing the content includes standardizing a content format and removing unnecessary information.

[1550] (Claim 3)

[1551] 10. The system of claim 1, wherein feature extraction includes sentence length, word diversity, and frequent word list generation for text content, and analysis of melody line, beat pattern, and harmonic spectrum for music content.

[1552] "Example 1"

[1553] (Claim 1)

[1554] means for receiving content from a user;

[1555] means for preprocessing the received content;

[1556] means for extracting features from the preprocessed content;

[1557] A means for calculating the probability of AI generation and human generation based on the extracted features;

[1558] a means for determining an evaluation rank of the content based on the calculated probability;

[1559] A means for providing an evaluation rank to a user;

[1560] A system including:

[1561] (Claim 2)

[1562] 10. The system of claim 1, wherein preprocessing the content includes standardizing a content format and removing unnecessary information.

[1563] (Claim 3)

[1564] 10. The system of claim 1, wherein feature extraction includes sentence length, word diversity, and frequent word list generation for text content, and analysis of melody line, rhythmic patterns, and harmonic spectrum for musical content.

[1565] "Application Example 1"

[1566] (Claim 1)

[1567] means for receiving content from a user;

[1568] means for preprocessing the received content;

[1569] means for extracting features from the preprocessed content;

[1570] A means for calculating the probability of AI generation and human generation based on the extracted features;

[1571] a means for determining an evaluation rank of the content based on the calculated probability;

[1572] A means for providing an evaluation rank to a user;

[1573] a means for displaying the results as an application installed on a smart device;

[1574] A system including:

[1575] (Claim 2)

[1576] 10. The system of claim 1, wherein preprocessing the content includes standardizing a content format and removing unnecessary information.

[1577] (Claim 3)

[1578] The system of claim 1, wherein the feature extraction includes, in the case of text content, sentence length, word diversity, and generating a list of frequently occurring words, and in the case of music content, analysis of melody line, beat pattern, and harmonic spectrum, and includes a function to display the extracted features on a smart device.

[1579] "Example 2: Combining Emotion Engines"

[1580] (Claim 1)

[1581] means for receiving content from a user;

[1582] means for preprocessing the received content;

[1583] means for extracting features from the preprocessed content;

[1584] A means for calculating the probability of AI generation and human generation based on the extracted features;

[1585] means for acquiring user emotion data as input;

[1586] means for analyzing the acquired emotion data and determining the user's emotional state;

[1587] a means for determining an evaluation rank based on the determined emotional state and the calculated probabilities of AI generation and human generation;

[1588] A means for providing an evaluation rank to a user;

[1589] A system including:

[1590] (Claim 2)

[1591] 10. The system of claim 1, wherein preprocessing the content includes standardizing a content format and removing unnecessary information.

[1592] (Claim 3)

[1593] 10. The system of claim 1, wherein feature extraction includes sentence length, word diversity, and frequent word list generation for text content, and analysis of melody line, beat pattern, and harmonic spectrum for music content.

[1594] "Application example 2 when combining emotion engines"

[1595] (Claim 1)

[1596] means for receiving content from a user;

[1597] means for preprocessing the received content;

[1598] means for extracting features from the preprocessed content;

[1599] A means for calculating the probability of AI generation and human generation based on the extracted features;

[1600] means for recognizing the emotional state of a user;

[1601] means for adjusting a rating rank of the content based on the calculated probability and the perceived emotional state;

[1602] A means for providing an evaluation rank to a user;

[1603] A system including:

[1604] (Claim 2)

[1605] 10. The system of claim 1, wherein preprocessing the content includes standardizing a content format and removing unnecessary information.

[1606] (Claim 3)

[1607] 10. The system of claim 1, wherein feature extraction includes sentence length, word diversity, and frequent word list generation for text content, and analysis of melody line, beat pattern, and harmonic spectrum for music content. [Explanation of symbols]

[1608] 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. To identify the content creator, means for receiving content from a user; means for preprocessing the received content; means for extracting features from the preprocessed content; A means for calculating the probability of AI generation and human generation based on the extracted features; a means for determining an evaluation rank of the content based on the calculated probability; A means for providing an evaluation rank to a user; A system including:

2. The system of claim 1 , wherein pre-processing the content includes standardizing the format of the content and removing unnecessary information.

3. 10. The system of claim 1, wherein feature extraction includes sentence length, word diversity, and frequent word list generation for text content, and analysis of melody line, beat pattern, and harmonic spectrum for music content.

Images