System

Abstract

A system is provided.SOLUTION: The system of claim 1, wherein the analysis engine is configured to compare and analyze the plurality of generated AI models and the subject, and wherein the analysis engine is configured to compare and analyze the plurality of generated data models and the subject.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] In recent years, with the rapid advancement of generative AI technology, texts and images created using generative AI are increasingly being used in assignments and papers in educational settings. As a result, there are concerns that this will lead to a decline in students' writing and thinking skills. While educational institutions have taken measures to strictly restrict or prohibit the use of generative AI, there is still the problem of difficulty in determining whether or not generative AI is actually being used. It is difficult for ordinary humans to visually determine whether or not something has been created using generative AI, which places a significant burden on educational institutions. Therefore, there is a need for technology that can quickly and accurately identify works created by generative AI. [Means for solving the problem]

[0005] The present invention provides a system that acquires target digital data from an input device and transmits the acquired data to an analytical engine. The analytical engine compares and analyzes the target data with multiple generative AI models and returns the results in percentage format to the receiving device. The receiving device presents the analysis results to the user, instantly determining whether the data was created by generative AI. This allows educational institutions to efficiently and accurately determine the likelihood that student assignments or papers were created by generative AI, preventing a decline in students' writing and thinking skills and improving the efficiency of teachers' evaluation work.

[0006] "Target digital data" refers to text, images, or similar digital information that a user inputs into a system.

[0007] "Input device" refers to a device or means for a user to input digital data, and specifically includes PCs, smartphones, tablets, etc.

[0008] "Means for acquiring" refers to the process of collecting target digital data from an input device and converting it into a format that can be processed within the system.

[0009] "Transmitting means" refers to the communication protocol or method for transferring the acquired digital data to the analysis engine.

[0010] "Analytical Engine" means a software or hardware component that compares and analyzes received digital data with multiple generative AI models.

[0011] "Generative AI models" refer to algorithms or models that use AI technology to generate text or images, and specifically include GPT, BERT, T5, etc.

[0012] "Means for comparison and analysis" refers to methods and technologies for comparing the generated AI model with the target digital data and calculating the degree of match.

[0013] "Matchability" refers to the probability value indicating the possibility that the target digital data was generated by a generative AI model.

[0014] "Percentage display format" refers to a method of expressing analysis results as percentages so that users can intuitively understand them.

[0015] "Receiving device" refers to a device or means for receiving analysis results, and specifically includes PCs, smartphones, tablets, etc. that function as terminals.

[0016] "Presenting means" refers to the interface or method for visually displaying the analysis results to the user on the receiving device. [Brief explanation of the drawings]

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

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

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

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

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

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

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

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

[0025] [First embodiment]

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

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

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

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

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

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

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

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

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

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

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

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

[0038] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. Specific examples of the present invention are described in detail below.

[0039] System Configuration

[0040] This system consists of the following main components:

[0041] 1. Input Device (Terminal)

[0042] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[0043] 2. Server

[0044] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0045] 3. Analysis Engine

[0046] It is software or an application located on a server that compares target data with multiple generative AI models and calculates the degree of match. The analytical engine uses natural language processing technology and machine learning algorithms.

[0047] 4. Receiving device (terminal)

[0048] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0049] Program processing overview

[0050] 1. Enter and submit data

[0051] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[0052] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[0053] 2. Data Receipt and Analysis

[0054] The server receives data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[0055] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[0056] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[0057] 3. Sending and displaying analysis results

[0058] The server formats the analysis results in JSON format or similar and sends them to the receiving device.

[0059] The device visually presents the analysis results to the user, for example, displaying a message such as "There is an 85% chance that this sentence was created by a generative AI."

[0060] Specific examples

[0061] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0062] 1. Enter and submit data

[0063] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[0064] The user clicks the "Start Analysis" button and the text is sent to the server.

[0065] 2. Data Receipt and Analysis

[0066] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[0067] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[0068] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[0069] 3. Sending and displaying analysis results

[0070] The server formats the "75% match" result in JSON format and sends it to the receiving device.

[0071] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI."

[0072] In this way, the invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing inappropriate use of generative AI.

[0073] The processing flow will be explained below.

[0074] Detailed processing steps of the program

[0075] Step 1:

[0076] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[0077] Step 2:

[0078] The terminal converts the input digital data into JSON format, which makes it easier for the server to parse.

[0079] Step 3:

[0080] The device sends the converted JSON data as an HTTP POST request to the server's API endpoint.

[0081] Step 4:

[0082] The server receives the POST request, converts the received data from JSON back to an internal parsing format, and temporarily stores it in a database.

[0083] Step 5:

[0084] The server adds new analysis jobs to the queue, which contain the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[0085] Step 6:

[0086] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[0087] Step 7:

[0088] The analytics engine compares the incoming data with multiple generative AI models, specifically analyzing the following:

[0089] Contextual consistency

[0090] Grammatical accuracy

[0091] Vocabulary frequency and patterns

[0092] Prevalence of similar text or images

[0093] Step 8:

[0094] The server obtains individual probability values ​​from each generative AI model and combines them to calculate the final match rate. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, they will be combined to calculate a match rate of 75%.

[0095] Step 9:

[0096] The server converts the analysis results, including the final degree of match, into JSON format and sends it to the terminal as an HTTP response.

[0097] Step 10:

[0098] The device then analyzes the analysis results and displays them in a format that is easy for the user to understand. For example, the device might display something like, "There is a 75% chance that this sentence was created by a generative AI."

[0099] Step 11:

[0100] The user checks the analysis results and takes the next action if necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[0101] Example 1

[0102] 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."

[0103] In recent years, digital data created by generative AI has become widely used in educational institutions and companies, raising concerns about its misuse and unauthorized use. The present invention aims to solve these problems and promote the proper use of generative AI by providing a system that can accurately determine whether digital data was created by generative AI.

[0104] 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.

[0105] In this invention, the server includes a means for transmitting target digital data to an analytical engine, a means for the analytical engine to compare and analyze the target data with multiple generative AI models, and a means for calculating a final degree of match by combining the degrees of match obtained from each generative AI model, thereby making it possible to determine with high accuracy whether digital data was created by a generative AI.

[0106] "Target digital data" refers to electronic text or image data acquired from an input device as the target for analysis.

[0107] An "input device" is a device such as a computer, smartphone, or tablet that allows a user to input digital data.

[0108] "Analysis Engine" means a program or software that compares and analyzes acquired digital data with a generated AI model.

[0109] A "generative AI model" is an artificial intelligence model trained to generate text and images using natural language processing and machine learning algorithms.

[0110] "Matchability" is a numerical value that indicates the likelihood that the target digital data was created by a generative AI model, and is calculated as an analysis result.

[0111] The "receiving device" is a device that receives the analysis results sent from the server and presents them to the user.

[0112] The "percentage display format" is a method of expressing the degree of agreement of the analysis results as a percentage.

[0113] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. A specific embodiment of this system is described in detail below.

[0114] System Configuration

[0115] This system consists of the following main components:

[0116] 1. Input Device (Terminal)

[0117] A device that allows a user to input digital data (text or images). This device includes personal computers, smartphones, tablets, etc.

[0118] 2. Server

[0119] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0120] 3. Analysis Engine

[0121] It is a software or application located on a server that compares target data with multiple generative AI models and calculates the degree of agreement. The analysis engine uses natural language processing technology and machine learning algorithms, specifically models such as GPT, BERT, and T5.

[0122] 4. Receiving device (terminal)

[0123] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0124] Entering and Submitting Data

[0125] The user uses an input device to input the digital data to be analyzed (e.g., text or images) into the input form. Specifically, the user types text using a keyboard or selects an image from a file selection dialog.

[0126] The user clicks the "Start Analysis" button in the input form. This action sends the input data to the server. Specifically, an HTTP request is generated and a payload containing the data is sent to the server.

[0127] Receiving and analyzing data

[0128] The server receives the data sent by the user. Specifically, the server's web server captures the HTTP request and retrieves the body data.

[0129] The server converts the received data into a standard format such as JSON and formats it for analysis. Specifically, a data parser analyzes the raw data, extracts the necessary information, and formats it.

[0130] The server's analysis engine compares and analyzes the received data with multiple generative AI models (e.g., GPT, BERT, T5). Specifically, the server calls the analysis engine, inputs the data into the generative AI model, and analyzes the context, grammar, and vocabulary usage patterns.

[0131] The server then combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm to combine the match scores from different models and calculates the final score using methods such as a weighted average.

[0132] Sending and displaying analysis results

[0133] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results and serializes them into a JSON-formatted string.

[0134] The server then sends the formatted analysis results to the receiving device. Specifically, it generates an HTTP response and sends it to the device.

[0135] The device visually presents the analysis results to the user. Specifically, the device's web application parses the JSON data and displays a message saying, "There is an 85% chance that this sentence was created by generative AI."

[0136] Specific examples

[0137] For example, consider a case where a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0138] Prompt Sentence Examples

[0139] Below are some example prompts to input to the generative AI model:

[0140] "Please help me write the report that's due today."

[0141] "Write a data analysis explanation for a school project."

[0142] "Please translate the following text into an English fan fiction story."

[0143] This prompt statement can be used to check the system's operation and to demonstrate it.

[0144] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0145] Step 1:

[0146] A user inputs digital data using an input device. The user enters text or images to be analyzed into an input form. In this process, the user types text content using a keyboard or uploads an image file using a file selection dialog. The input data is temporarily stored in local storage on the input device. Input = digital data to be analyzed, Output = raw data stored on the input device.

[0147] Step 2:

[0148] The user clicks the "Start Analysis" button, which sends the input data to the server. Specifically, the input device generates an HTTP POST request, which includes the input data as the payload. Input = raw data stored on the input device, Output = data payload sent to the server.

[0149] Step 3:

[0150] The server receives data sent from the input device. The server's web server captures the HTTP request and obtains the payload data. Input = data payload sent from the input device, Output = formatted data to pass to the analysis engine.

[0151] Step 4:

[0152] The server converts the data it receives into a standard format such as JSON and formats it for analysis. Specifically, the server's data parser analyzes the raw data, extracts the necessary information, and serializes it. Input = data payload received by the server, output = formatted data in JSON format.

[0153] Step 5:

[0154] The server's analysis engine compares and analyzes the formatted data with multiple generative AI models. Specifically, the server calls the analysis engine, inputs the formatted data into generative AI models such as GPT, BERT, and T5, and analyzes the context, grammar, and vocabulary usage patterns. Input = formatted data in JSON format, output = match score for each generative AI model.

[0155] Step 6:

[0156] The server combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm (e.g., weighted average) to combine the match scores from different models to calculate the overall score. Input = match score for each generative AI model, output = final match score.

[0157] Step 7:

[0158] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results, serializes it into a JSON-formatted string, and sends it as an HTTP response. Input = final match score, output = JSON-formatted analysis results sent to the receiving device.

[0159] Step 8:

[0160] The device visually presents the analysis results received from the server to the user. Specifically, the device's web application parses the JSON data and displays the message, "There is an 85% chance that this sentence was created by generative AI." Input = JSON-formatted analysis results received from the server, Output = visual analysis results presented to the user.

[0161] Specific examples

[0162] Consider a case in which a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0163] Prompt Sentence Examples

[0164] Below are some example prompts to input to the generative AI model:

[0165] "Please help me write the report that's due today."

[0166] "Write a data analysis explanation for a school project."

[0167] "Please translate the following text into an English fan fiction story."

[0168] This prompt statement can be used to check the system's operation and to demonstrate it.

[0169] (Application example 1)

[0170] 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."

[0171] In recent years, the generation of digital data using generative AI has rapidly spread, resulting in a need to determine whether it was created by generative AI. As this technology becomes more widespread, the risk of unauthorized use of text and images created by generative AI also increases. It is particularly important for educational institutions and companies to properly monitor the use of generative AI and take necessary measures. However, current systems face the challenge of making this determination efficiently and accurately.

[0172] 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.

[0173] In this invention, the server includes means for transmitting digital data input from a user terminal to an analysis server, means for comparing and analyzing the target data using multiple generative AI models, and means for returning the analysis results in percentage format to the terminal, thereby enabling accurate real-time determination of whether the digital data was created by a generative AI.

[0174] A "terminal" is a device through which a user inputs digital data, and specifically refers to a personal computer, smartphone, tablet, etc.

[0175] An "analysis server" is a computer system that receives digital data sent from a terminal and analyzes the data using a generative AI model.

[0176] A "generative AI model" is an algorithm or model that uses artificial intelligence technology to generate digital data, and examples include GPT, BERT, and T5.

[0177] "Comparison and analysis" is the process in which the analysis server uses a generative AI model to compare input data with known data sets and analyze the data's features and patterns.

[0178] The "percentage display format" is a display format in which the analysis results are quantified and the degree of agreement or reliability is expressed as a percentage.

[0179] "Visual presentation means" refers to an interface or method for displaying analysis results in a way that is easy for users to understand, such as using graphs, text, charts, etc.

[0180] "Synthesis" refers to the process of integrating the results obtained from multiple generative AI models to derive the final analysis results.

[0181] In order to implement the present invention, the following system and means are utilized: Each component of the system is designed based on the claims.

[0182] System Configuration

[0183] This system consists of the following main components:

[0184] 1. Device:

[0185] A terminal is a device through which a user inputs digital data. Specifically, a personal computer, smartphone, tablet, etc. is used. The user inputs digital data into the terminal and sends it to an analysis server.

[0186] 2. Analysis server:

[0187] The analysis server is a computer system that receives digital data sent from the terminal and analyzes the data using a generative AI model. The server has generative AI models such as GPT, BERT, and T5 installed, and processes the data for analysis.

[0188] 3. Generative AI Model:

[0189] A generative AI model is an algorithm or model that uses generative AI to generate digital data. Examples include GPT, BERT, and T5. This model analyzes input digital data by comparing it with existing datasets.

[0190] 4. Data Receipt and Analysis:

[0191] The server receives the digital data sent from the device and converts it into a standard format such as JSON. The converted data is formatted for analysis and compared and analyzed using a generative AI model. The analysis results are calculated in the form of a confidence level (%).

[0192] 5. Presentation of results:

[0193] The analysis results are returned to the terminal in the form of a percentage. The terminal provides an interface to visually present the analysis results to the user, allowing the user to easily understand whether the digital data was created by generative AI.

[0194] Process flow:

[0195] As a concrete example, consider the analysis of the sentence "He finished writing his thesis."

[0196] 1. Data entry and transmission via terminal:

[0197] The user enters the sentence "He finished writing his thesis" into the input form on the terminal. By pressing the send button, this data is sent to the analysis server.

[0198] 2. Data reception and analysis by the server:

[0199] The analysis server formats the received digital data for analysis and performs the analysis using a generative AI model, such as GPT, BERT, or T5, to analyze the grammar and structure of the sentence.

[0200] 3. Sending and displaying analysis results:

[0201] The server calculates the analysis results and returns them to the device in the form of a percentage, which then displays a message to the user, such as "There is an 85% chance that this sentence was created by a generative AI."

[0202] The system uses Python hoformat to send and receive data between the server and the device. Specific analysis is performed based on a generative AI model, and the reliability of the analysis results is calculated.

[0203] The above process and components used enable the present invention to accurately determine in real time whether digital data entered by a user was created by generative AI.

[0204] Example prompt sentence:

[0205] "He finished writing his thesis."

[0206] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0207] Step 1:

[0208] A user uses a terminal to input digital data. Specifically, the user enters text data such as "He has finished writing his thesis" into the terminal's input form and clicks the "Submit" button. This input data is later formatted into JSON format to be sent to an analysis server. At this stage, the input is the text data entered by the user, and the output is JSON format data.

[0209] Step 2:

[0210] The terminal sends the digital data entered by the user to the analysis server. The terminal uses an HTTP POST request to send JSON-formatted data to the server. This request allows the data sent from the terminal to reach the analysis server, allowing the server to begin processing the received data. The input is JSON-formatted text data entered by the user, and the output is an HTTP request to the server.

[0211] Step 3:

[0212] The server receives the digital data sent from the device and formats it for analysis. Specifically, it converts the received JSON data into an appropriate data format that can be analyzed by the generative AI model. This conversion process includes a processing step that converts JSON into an internal data structure that can be processed by the analysis engine. The input is JSON-formatted data, and the output is internal data for the analysis engine.

[0213] Step 4:

[0214] The server's analysis engine analyzes the data using generative AI models. This analysis process uses multiple generative AI models, such as GPT, BERT, and T5, to compare the input digital data with existing datasets. Specifically, it performs grammar checks, contextual analysis, and the detection of specific lexical patterns. The input for this step is the internal data for the analysis engine, and the output is a confidence score analyzed by each generative AI model.

[0215] Step 5:

[0216] The server aggregates the analysis results from each generative AI model and calculates the final analysis result. Specifically, it combines the confidence scores obtained from each model to calculate the final match and confidence. In this step, the analysis results are consolidated using average and maximum values ​​based on context and patterns. The input is the confidence scores from each generative AI model, and the output is the overall analysis result.

[0217] Step 6:

[0218] The server formats the final analysis results in JSON format and sends them to the terminal. Specifically, the analysis results generated internally on the server are converted into JSON data format and sent back to the terminal as an HTTP response. The input is the comprehensive analysis results, and the output is JSON format data.

[0219] Step 7:

[0220] The device visually presents the analysis results received from the server to the user. For example, a message such as "There is an 85% chance that this sentence was created by a generative AI" is displayed on the screen. The input is the analysis results in JSON format received from the server, and the output is a visual display for the user.

[0221] These steps allow the system to determine in real time whether digital data was created by generative AI.

[0222] 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.

[0223] The present invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed. Specific examples of the present invention are described in detail below.

[0224] System Configuration

[0225] This system consists of the following main components:

[0226] 1. Input Device (Terminal)

[0227] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[0228] 2. Server

[0229] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[0230] 3. Analysis Engine

[0231] This software or application resides on a server and compares target data with multiple generative AI models to determine the results of generative AI creations.

[0232] 4. Emotion Engine

[0233] Software or application located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[0234] 5. Receiving Device (Terminal)

[0235] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user.

[0236] Program processing overview

[0237] 1. Enter and submit data

[0238] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[0239] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[0240] 2. Data Receipt and Analysis

[0241] The server receives the data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[0242] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[0243] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[0244] 3. Emotion Recognition Processing

[0245] The emotion engine in the server recognizes the user's emotions based on the analysis results. For text, it classifies emotions as negative, positive, neutral, etc. For images, it uses facial expression recognition technology to analyze emotions.

[0246] The server adds the emotion recognition results to the analysis results to create comprehensive feedback.

[0247] 4. Transmission and display of analysis results and emotion recognition results

[0248] The server formats the analysis results and emotion recognition results in JSON format or similar and sends them to the receiving device.

[0249] The device displays the received results in a user-friendly format, such as a message like, "There is an 85% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[0250] Specific examples

[0251] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0252] 1. Enter and submit data

[0253] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[0254] The user clicks the "Start Analysis" button and the text is sent to the server.

[0255] 2. Data Receipt and Analysis

[0256] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[0257] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[0258] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[0259] 3. Emotion Recognition Processing

[0260] The emotion engine in the server recognizes the user's emotion as positive based on the content of the text.

[0261] 4. Transmission and display of analysis results and emotion recognition results

[0262] The server formats the result of "75% match" and the recognition result of "emotion is positive" in JSON format and sends them to the receiving device.

[0263] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI. Also, the user's sentiment is positive."

[0264] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[0265] The processing flow will be explained below.

[0266] Specific processing steps of the invention that combines the emotion engine

[0267] Step 1:

[0268] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[0269] Step 2:

[0270] The terminal converts the input digital data into JSON format, which is then sent to the server.

[0271] Step 3:

[0272] The server receives the JSON data, which is then formatted for parsing and temporarily stored in a database.

[0273] Step 4:

[0274] The server adds a new analysis job to the queue, which contains the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[0275] Step 5:

[0276] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[0277] Step 6:

[0278] The analytics engine compares incoming data with multiple generative AI models, analyzing:

[0279] Contextual consistency

[0280] Grammatical accuracy

[0281] Vocabulary frequency and patterns

[0282] Prevalence of similar text or images

[0283] Step 7:

[0284] The server obtains the individual probability values ​​from each generative AI model. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, then the server combines these to calculate a match rate of 75%.

[0285] Step 8:

[0286] The server passes the results of the match analysis to the emotion engine, which then recognizes the user's emotion from the content of the received data. For text, it classifies emotions as negative, positive, neutral, etc., and for images, it uses facial expression recognition technology to analyze the emotion.

[0287] Step 9:

[0288] The server adds the emotion recognition results to the analysis results and generates comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[0289] Step 10:

[0290] The server converts the analysis results and emotion recognition results into JSON format and sends them to the receiving device as an HTTP response.

[0291] Step 11:

[0292] The device analyzes the received analysis results and emotion recognition results and displays them in a format that is easy for the user to understand. For example, a message such as "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive" may be displayed on the screen.

[0293] Step 12:

[0294] The user checks the analysis results and emotion recognition results and takes the following actions as necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[0295] Example 2

[0296] 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."

[0297] In recent years, technological advances in generative AI have made it easier to automatically generate text and images, increasing the risk of this being misused by educational institutions and businesses. Furthermore, there is a demand for feedback that takes into account the user's emotions, which has been difficult with conventional systems. Therefore, a system is needed that can simultaneously identify digital data automatically generated by generative AI, recognize the user's emotions, and provide feedback that reflects those emotions.

[0298] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for acquiring target digital data from an information processing device, means for transmitting the acquired data to an analysis device, means for the analysis device to compare and analyze the target data with multiple generative AI models, means for recognizing the user's emotions based on the analysis results, and means for returning the analysis results and emotion recognition results in a percentage format to the receiving device. This enables the generation AI to identify automatically generated data and provide feedback that reflects the user's emotions.

[0299] An "information processing device" is a device that allows a user to input digital data, and is a wide range of terminals including PCs, smartphones, tablets, etc.

[0300] The "analysis device" is a device for analyzing acquired digital data, and is a module within the server that uses multiple generative AI models to compare and analyze data.

[0301] A "generative AI model" is an algorithm that automatically generates digital data using generative AI technology, and specifically includes GPT, BERT, T5, etc.

[0302] "Digital data" refers to data provided by a user via an input device, and includes text data, image data, and the like.

[0303] "Means for recognizing emotions" refers to software or algorithms for classifying and recognizing a user's emotions based on the analysis results, including text analysis and facial expression recognition technologies.

[0304] A "receiving device" is a device that receives the analysis results and emotion recognition results and presents them to the user, and includes smartphones, PCs, tablets, etc.

[0305] The "means for transmitting data" refers to a protocol or interface for transmitting digital data from the information processing device to the analysis device.

[0306] This invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed.

[0307] This system consists of the following main components:

[0308] 1. Input devices (information processing devices)

[0309] A device that allows users to input digital data (text and images). Examples of this device include PCs, smartphones, and tablets.

[0310] 2. Server (analysis device)

[0311] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[0312] 3. Analysis Engine

[0313] It is software or a program located on a server that compares target data with multiple generative AI models to determine the results of generative AI creations. Specifically, models such as GPT, BERT, and T5 are used.

[0314] 4. Emotion Engine

[0315] Software or a program located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[0316] 5. Receiving Device

[0317] This is a device that receives analysis results and emotion recognition results sent from the server and presents them to the user.

[0318] Specific processing of the program

[0319] Digital data entered by a user through an information processing device is first sent to a server. The server's analysis engine compares the received data with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. Next, an emotion engine recognizes the user's emotions. The analysis results and emotion recognition results are integrated and sent to the receiving device, where they are finally presented to the user.

[0320] Specific examples

[0321] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0322] The user enters the following sentence into the form on their PC: "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." When the user clicks the "Start Analysis" button, the sentence is sent to the server.

[0323] The server formats the text data and passes it to the analysis engine. The analysis engine compares the text with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. For example, if GPT reports a match of 70%, BERT reports a match of 75%, and T5 reports a match of 80%, these will be combined to arrive at a final match of 75%.

[0324] Next, the emotion engine on the server recognizes the user's emotion as positive based on the content of the sentence. This integrates the analysis results with the emotion recognition results, and displays on the receiving device the message, "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive."

[0325] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[0326] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0327] Step 1:

[0328] Users input text or images into a form on an input device such as a PC, smartphone, or tablet. Specifically, they enter text into a text area or upload an image file. This input data (e.g., the text "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." or an image file) is converted to JSON format on the device.

[0329] Input: Text or image data entered by a user into an input device

[0330] Output: Digital data formatted in JSON format

[0331] Step 2:

[0332] The user clicks the "Start Analysis" button on the input device. This operation sends the digital data to the server. When the send button is clicked, the device starts communication to send the data to the server.

[0333] Input: Digital data formatted in JSON format

[0334] Output: Server receives data

[0335] Step 3:

[0336] The server receives the digital data sent from the input device, where it is reconfirmed in JSON format and prepared to be passed to the analysis engine.

[0337] Input: JSON formatted data sent from the input device

[0338] Output: Formatted data to be passed to the analysis engine

[0339] Step 4:

[0340] The server's analytics engine analyzes the incoming digital data. This engine then invokes generative AI models such as GPT, BERT, and T5 to match the data. A match is calculated for each model, comparing context, grammar, and vocabulary usage patterns.

[0341] Input: Formatted digital data

[0342] Output: Match score by each generative AI model (e.g., 70% for GPT, 75% for BERT, 80% for T5)

[0343] Step 5:

[0344] The server then combines the match scores from each generative AI model to calculate the overall match score. For example, the match scores (70% for GPT, 75% for BERT, and 80% for T5) are averaged to calculate the final match score (e.g., 75%).

[0345] Input: Match scores from each generative AI model

[0346] Output: Overall match (e.g. 75%)

[0347] Step 6:

[0348] The emotion engine in the server recognizes the user's emotion based on the analysis results. For text, it classifies it as negative, positive, or neutral. For example, the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." would be classified as positive. For images, emotion is analyzed using facial expression recognition technology.

[0349] Input: Formatted data and analysis results

[0350] Output: User's emotion category (e.g., positive)

[0351] Step 7:

[0352] The server combines the emotion recognition results with the analysis results to create comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[0353] Input: Overall agreement and emotion category

[0354] Output: General feedback message

[0355] Step 8:

[0356] The server formats the aggregate feedback message in JSON format and sends it to the receiving device.

[0357] Input: General feedback message

[0358] Output: Feedback data in JSON format

[0359] Step 9:

[0360] The device displays the feedback data received from the server to the user, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[0361] Input: Feedback data in JSON format

[0362] Output: Feedback displayed in a format that the user can understand

[0363] (Application example 2)

[0364] 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."

[0365] Currently, it is difficult for brick-and-mortar stores to determine the authenticity and sentiment of customers' text messages in real time. This makes it difficult for customer service staff to respond appropriately, potentially hindering the improvement of the customer experience. It is also challenging to respond appropriately to false information and messages created by generative AI. These issues need to be resolved.

[0366] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0367] In this invention, the server includes a means for acquiring target digital data from an input device, a means for transmitting the acquired data to an analysis device, a means for the analysis device to compare and analyze the target data with multiple generative AI models, and a means for analyzing emotions and displaying the recognition results on a receiving device. This allows customer service staff to confirm the authenticity of customer messages and simultaneously recognize customer emotions in real time. This enables more appropriate and prompt customer service, and is expected to improve the customer experience.

[0368] "Target digital data" refers to any form of digital data that is the subject of analysis, including text, images, etc.

[0369] An "input device" is a device that allows a user to input digital data, and includes PCs, smartphones, tablets, smart glasses, etc.

[0370] An "analysis device" is a computer system that acquires target digital data and analyzes it using an analysis engine and an emotion recognition engine.

[0371] A "generative AI model" is a model generated using artificial intelligence, such as a natural language processing model such as GPT or BERT.

[0372] The "receiving device" is a device that receives the analysis results and emotion recognition results sent from the analyzing device and presents them to the user.

[0373] The "means for analyzing emotions and displaying the recognition results on the receiving device" is a means having the function of analyzing the user's emotions from the target digital data and displaying the results on the receiving device in real time.

[0374] This invention relates to a system that determines whether target digital data was created by generative AI and simultaneously recognizes the user's emotions. This system is intended to be used primarily in customer service situations in brick-and-mortar stores.

[0375] System Configuration

[0376] The system includes the following major components:

[0377] 1. Input Devices

[0378] A device that allows customers to input digital data (text and images). This includes wearable devices such as smart glasses.

[0379] 2. Analysis device

[0380] This is a computer system that receives and analyzes data sent from an input device. This device has a built-in analysis engine and emotion engine.

[0381] 3. Analysis Engine

[0382] This software is located on a server and compares the target data with multiple generative AI models to determine the results of generative AI creations. GPT and BERT are used as generative AI models.

[0383] 4. Emotion Engine

[0384] This software is located on the server and analyzes target data to recognize user emotions. It uses text analysis and image analysis techniques.

[0385] 5. Receiving Device

[0386] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user. Smart glasses are used as this receiving device.

[0387] Hardware and software used

[0388] Hardware

[0389] Smart glasses (e.g., Google Glass, Vuzix Blade)

[0390] Server (e.g. AWS, Google Cloud)

[0391] software

[0392] Web framework: Flask

[0393] Generative AI decision engine: AIAnalyzer (library of hypotheses)

[0394] Emotion recognition engine: SentimentAnalyzer (hypothetical library)

[0395] Data processing and calculation methods

[0396] The analysis device (server) performs the following processes.

[0397] 1. Receiving and formatting data

[0398] The data sent from the input device is received by the server, where it is converted to JSON format or similar and formatted for analysis.

[0399] 2. Discrimination by Generative AI

[0400] Using a generative AI model (e.g., GPT, BERT, etc.), the probability that the target data is generated by generative AI is analyzed. The analysis result is expressed in percentage form.

[0401] 3. Emotion recognition

[0402] The emotion engine analyzes the user's emotions from the target data and classifies them as positive, negative, neutral, etc.

[0403] 4. Sending the results

[0404] The analysis results and emotion recognition results are formatted and transmitted to a receiving device.

[0405] Specific examples

[0406] Let's take the example of a customer in a brick-and-mortar clothing store who sends a message asking, "Is this clothing popular on social media?" This message is sent to the server as a prompt as follows:

[0407] Prompt Sentence Examples

[0408] Is this outfit creating a buzz on social media?

[0409] The analysis results are displayed as follows: "Probability of message generated by AI: 85%, Emotion: Neutral." Based on this information, customer service staff can quickly respond appropriately.

[0410] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0411] Step 1:

[0412] A user inputs target digital data such as text or images into an input device such as smart glasses. Specifically, the user inputs a message such as "Is this outfit trending on social media?" using voice or text input. This input data is sent to the smart glasses.

[0413] Step 2:

[0414] The smart glasses send the input digital data to the server, where it is converted into JSON format and passed to an analysis device. If the input sentence is something like "Is this clothing popular on social media?", it will be sent as JSON data.

[0415] Step 3:

[0416] The analysis device receives the data received by the server and formats it. The formatted data is converted into a format that can be analyzed by a generative AI model (GPT, BERT, etc.). For example, text data is formatted according to a standard format.

[0417] Step 4:

[0418] The server's analysis engine compares and analyzes the formatted data with the generative AI model. Specifically, input data is fed into a generative AI model such as GPT or BERT, and the degree of match and probability are calculated to determine whether the data was generated by the generative AI. For example, for the data "Is this clothing item trending on social media?", the probability that it was generated by the generative AI is determined to be 85%.

[0419] Step 5:

[0420] The server's emotion engine recognizes the user's emotion based on the analysis results. Text analysis technology is used to classify emotions from the target data (text). For example, the emotion engine will recognize the sentence "Is this clothing popular on social media?" as "neutral."

[0421] Step 6:

[0422] The server formats the generated AI's judgment results and emotion recognition results into JSON format and sends them to the smart glasses, which act as a receiving device. The formatted data includes the generated AI's probability (85%) and emotion classification (neutral).

[0423] Step 7:

[0424] The smart glasses then present the analysis results and emotion recognition results to the user. The display shows "Probability of message generated by AI: 85%, Emotion: Neutral," and the customer service staff can use this information to respond appropriately.

[0425] 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.

[0426] 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.

[0427] 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.

[0428] [Second embodiment]

[0429] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0430] 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.

[0431] 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).

[0432] 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.

[0433] 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.

[0434] 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).

[0435] 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.

[0436] 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.

[0437] 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.

[0438] 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.

[0439] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0440] 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."

[0441] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. Specific examples of the present invention are described in detail below.

[0442] System Configuration

[0443] This system consists of the following main components:

[0444] 1. Input Device (Terminal)

[0445] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[0446] 2. Server

[0447] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0448] 3. Analysis Engine

[0449] It is software or an application located on a server that compares target data with multiple generative AI models and calculates the degree of match. The analytical engine uses natural language processing technology and machine learning algorithms.

[0450] 4. Receiving device (terminal)

[0451] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0452] Program processing overview

[0453] 1. Enter and submit data

[0454] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[0455] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[0456] 2. Data Receipt and Analysis

[0457] The server receives data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[0458] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[0459] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[0460] 3. Sending and displaying analysis results

[0461] The server formats the analysis results in JSON format or similar and sends them to the receiving device.

[0462] The device visually presents the analysis results to the user, for example, displaying a message such as "There is an 85% chance that this sentence was created by a generative AI."

[0463] Specific examples

[0464] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0465] 1. Enter and submit data

[0466] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[0467] The user clicks the "Start Analysis" button and the text is sent to the server.

[0468] 2. Data Receipt and Analysis

[0469] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[0470] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[0471] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[0472] 3. Sending and displaying analysis results

[0473] The server formats the "75% match" result in JSON format and sends it to the receiving device.

[0474] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI."

[0475] In this way, the invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing inappropriate use of generative AI.

[0476] The processing flow will be explained below.

[0477] Detailed processing steps of the program

[0478] Step 1:

[0479] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[0480] Step 2:

[0481] The terminal converts the input digital data into JSON format, which makes it easier for the server to parse.

[0482] Step 3:

[0483] The device sends the converted JSON data as an HTTP POST request to the server's API endpoint.

[0484] Step 4:

[0485] The server receives the POST request, converts the received data from JSON back to an internal parsing format, and temporarily stores it in a database.

[0486] Step 5:

[0487] The server adds new analysis jobs to the queue, which contain the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[0488] Step 6:

[0489] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[0490] Step 7:

[0491] The analytics engine compares the incoming data with multiple generative AI models, specifically analyzing the following:

[0492] Contextual consistency

[0493] Grammatical accuracy

[0494] Vocabulary frequency and patterns

[0495] Prevalence of similar text or images

[0496] Step 8:

[0497] The server obtains individual probability values ​​from each generative AI model and combines them to calculate the final match rate. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, they will be combined to calculate a match rate of 75%.

[0498] Step 9:

[0499] The server converts the analysis results, including the final degree of match, into JSON format and sends it to the terminal as an HTTP response.

[0500] Step 10:

[0501] The device then analyzes the analysis results and displays them in a format that is easy for the user to understand. For example, the device might display something like, "There is a 75% chance that this sentence was created by a generative AI."

[0502] Step 11:

[0503] The user checks the analysis results and takes the next action if necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[0504] Example 1

[0505] 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."

[0506] In recent years, digital data created by generative AI has become widely used in educational institutions and companies, raising concerns about its misuse and unauthorized use. The present invention aims to solve these problems and promote the proper use of generative AI by providing a system that can accurately determine whether digital data was created by generative AI.

[0507] 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.

[0508] In this invention, the server includes a means for transmitting target digital data to an analytical engine, a means for the analytical engine to compare and analyze the target data with multiple generative AI models, and a means for calculating a final degree of match by combining the degrees of match obtained from each generative AI model, thereby making it possible to determine with high accuracy whether digital data was created by a generative AI.

[0509] "Target digital data" refers to electronic text or image data acquired from an input device as the target for analysis.

[0510] An "input device" is a device such as a computer, smartphone, or tablet that allows a user to input digital data.

[0511] "Analysis Engine" means a program or software that compares and analyzes acquired digital data with a generated AI model.

[0512] A "generative AI model" is an artificial intelligence model trained to generate text and images using natural language processing and machine learning algorithms.

[0513] "Matchability" is a numerical value that indicates the likelihood that the target digital data was created by a generative AI model, and is calculated as an analysis result.

[0514] The "receiving device" is a device that receives the analysis results sent from the server and presents them to the user.

[0515] The "percentage display format" is a method of expressing the degree of agreement of the analysis results as a percentage.

[0516] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. A specific embodiment of this system is described in detail below.

[0517] System Configuration

[0518] This system consists of the following main components:

[0519] 1. Input Device (Terminal)

[0520] A device that allows a user to input digital data (text or images). This device includes personal computers, smartphones, tablets, etc.

[0521] 2. Server

[0522] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0523] 3. Analysis Engine

[0524] It is a software or application located on a server that compares target data with multiple generative AI models and calculates the degree of agreement. The analysis engine uses natural language processing technology and machine learning algorithms, specifically models such as GPT, BERT, and T5.

[0525] 4. Receiving device (terminal)

[0526] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0527] Entering and Submitting Data

[0528] The user uses an input device to input the digital data to be analyzed (e.g., text or images) into the input form. Specifically, the user types text using a keyboard or selects an image from a file selection dialog.

[0529] The user clicks the "Start Analysis" button in the input form. This action sends the input data to the server. Specifically, an HTTP request is generated and a payload containing the data is sent to the server.

[0530] Receiving and analyzing data

[0531] The server receives the data sent by the user. Specifically, the server's web server captures the HTTP request and retrieves the body data.

[0532] The server converts the received data into a standard format such as JSON and formats it for analysis. Specifically, a data parser analyzes the raw data, extracts the necessary information, and formats it.

[0533] The server's analysis engine compares and analyzes the received data with multiple generative AI models (e.g., GPT, BERT, T5). Specifically, the server calls the analysis engine, inputs the data into the generative AI model, and analyzes the context, grammar, and vocabulary usage patterns.

[0534] The server then combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm to combine the match scores from different models and calculates the final score using methods such as a weighted average.

[0535] Sending and displaying analysis results

[0536] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results and serializes them into a JSON-formatted string.

[0537] The server then sends the formatted analysis results to the receiving device. Specifically, it generates an HTTP response and sends it to the device.

[0538] The device visually presents the analysis results to the user. Specifically, the device's web application parses the JSON data and displays a message saying, "There is an 85% chance that this sentence was created by generative AI."

[0539] Specific examples

[0540] For example, consider a case where a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0541] Prompt Sentence Examples

[0542] Below are some example prompts to input to the generative AI model:

[0543] "Please help me write the report that's due today."

[0544] "Write a data analysis explanation for a school project."

[0545] "Please translate the following text into an English fan fiction story."

[0546] This prompt statement can be used to check the system's operation and to demonstrate it.

[0547] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0548] Step 1:

[0549] A user inputs digital data using an input device. The user enters text or images to be analyzed into an input form. In this process, the user types text content using a keyboard or uploads an image file using a file selection dialog. The input data is temporarily stored in local storage on the input device. Input = digital data to be analyzed, Output = raw data stored on the input device.

[0550] Step 2:

[0551] The user clicks the "Start Analysis" button, which sends the input data to the server. Specifically, the input device generates an HTTP POST request, which includes the input data as the payload. Input = raw data stored on the input device, Output = data payload sent to the server.

[0552] Step 3:

[0553] The server receives data sent from the input device. The server's web server captures the HTTP request and obtains the payload data. Input = data payload sent from the input device, Output = formatted data to pass to the analysis engine.

[0554] Step 4:

[0555] The server converts the data it receives into a standard format such as JSON and formats it for analysis. Specifically, the server's data parser analyzes the raw data, extracts the necessary information, and serializes it. Input = data payload received by the server, output = formatted data in JSON format.

[0556] Step 5:

[0557] The server's analysis engine compares and analyzes the formatted data with multiple generative AI models. Specifically, the server calls the analysis engine, inputs the formatted data into generative AI models such as GPT, BERT, and T5, and analyzes the context, grammar, and vocabulary usage patterns. Input = formatted data in JSON format, output = match score for each generative AI model.

[0558] Step 6:

[0559] The server combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm (e.g., weighted average) to combine the match scores from different models to calculate the overall score. Input = match score for each generative AI model, output = final match score.

[0560] Step 7:

[0561] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results, serializes it into a JSON-formatted string, and sends it as an HTTP response. Input = final match score, output = JSON-formatted analysis results sent to the receiving device.

[0562] Step 8:

[0563] The device visually presents the analysis results received from the server to the user. Specifically, the device's web application parses the JSON data and displays the message, "There is an 85% chance that this sentence was created by generative AI." Input = JSON-formatted analysis results received from the server, Output = visual analysis results presented to the user.

[0564] Specific examples

[0565] Consider a case in which a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0566] Prompt Sentence Examples

[0567] Below are some example prompts to input to the generative AI model:

[0568] "Please help me write the report that's due today."

[0569] "Write a data analysis explanation for a school project."

[0570] "Please translate the following text into an English fan fiction story."

[0571] This prompt statement can be used to check the system's operation and to demonstrate it.

[0572] (Application example 1)

[0573] 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."

[0574] In recent years, the generation of digital data using generative AI has rapidly spread, resulting in a need to determine whether it was created by generative AI. As this technology becomes more widespread, the risk of unauthorized use of text and images created by generative AI also increases. It is particularly important for educational institutions and companies to properly monitor the use of generative AI and take necessary measures. However, current systems face the challenge of making this determination efficiently and accurately.

[0575] 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.

[0576] In this invention, the server includes means for transmitting digital data input from a user terminal to an analysis server, means for comparing and analyzing the target data using multiple generative AI models, and means for returning the analysis results in percentage format to the terminal, thereby enabling accurate real-time determination of whether the digital data was created by a generative AI.

[0577] A "terminal" is a device through which a user inputs digital data, and specifically refers to a personal computer, smartphone, tablet, etc.

[0578] An "analysis server" is a computer system that receives digital data sent from a terminal and analyzes the data using a generative AI model.

[0579] A "generative AI model" is an algorithm or model that uses artificial intelligence technology to generate digital data, and examples include GPT, BERT, and T5.

[0580] "Comparison and analysis" is the process in which the analysis server uses a generative AI model to compare input data with known data sets and analyze the data's features and patterns.

[0581] The "percentage display format" is a display format in which the analysis results are quantified and the degree of agreement or reliability is expressed as a percentage.

[0582] "Visual presentation means" refers to an interface or method for displaying analysis results in a way that is easy for users to understand, such as using graphs, text, charts, etc.

[0583] "Synthesis" refers to the process of integrating the results obtained from multiple generative AI models to derive the final analysis results.

[0584] In order to implement the present invention, the following system and means are utilized: Each component of the system is designed based on the claims.

[0585] System Configuration

[0586] This system consists of the following main components:

[0587] 1. Device:

[0588] A terminal is a device through which a user inputs digital data. Specifically, a personal computer, smartphone, tablet, etc. is used. The user inputs digital data into the terminal and sends it to an analysis server.

[0589] 2. Analysis server:

[0590] The analysis server is a computer system that receives digital data sent from the terminal and analyzes the data using a generative AI model. The server has generative AI models such as GPT, BERT, and T5 installed, and processes the data for analysis.

[0591] 3. Generative AI Model:

[0592] A generative AI model is an algorithm or model that uses generative AI to generate digital data. Examples include GPT, BERT, and T5. This model analyzes input digital data by comparing it with existing datasets.

[0593] 4. Data Receipt and Analysis:

[0594] The server receives the digital data sent from the device and converts it into a standard format such as JSON. The converted data is formatted for analysis and compared and analyzed using a generative AI model. The analysis results are calculated in the form of a confidence level (%).

[0595] 5. Presentation of results:

[0596] The analysis results are returned to the terminal in the form of a percentage. The terminal provides an interface to visually present the analysis results to the user, allowing the user to easily understand whether the digital data was created by generative AI.

[0597] Process flow:

[0598] As a concrete example, consider the analysis of the sentence "He finished writing his thesis."

[0599] 1. Data entry and transmission via terminal:

[0600] The user enters the sentence "He finished writing his thesis" into the input form on the terminal. By pressing the send button, this data is sent to the analysis server.

[0601] 2. Data reception and analysis by the server:

[0602] The analysis server formats the received digital data for analysis and performs the analysis using a generative AI model, such as GPT, BERT, or T5, to analyze the grammar and structure of the sentence.

[0603] 3. Sending and displaying analysis results:

[0604] The server calculates the analysis results and returns them to the device in the form of a percentage, which then displays a message to the user, such as "There is an 85% chance that this sentence was created by a generative AI."

[0605] The system uses Python hoformat to send and receive data between the server and the device. Specific analysis is performed based on a generative AI model, and the reliability of the analysis results is calculated.

[0606] The above process and components used enable the present invention to accurately determine in real time whether digital data entered by a user was created by generative AI.

[0607] Example prompt sentence:

[0608] "He finished writing his thesis."

[0609] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0610] Step 1:

[0611] A user uses a terminal to input digital data. Specifically, the user enters text data such as "He has finished writing his thesis" into the terminal's input form and clicks the "Submit" button. This input data is later formatted into JSON format to be sent to an analysis server. At this stage, the input is the text data entered by the user, and the output is JSON format data.

[0612] Step 2:

[0613] The terminal sends the digital data entered by the user to the analysis server. The terminal uses an HTTP POST request to send JSON-formatted data to the server. This request allows the data sent from the terminal to reach the analysis server, allowing the server to begin processing the received data. The input is JSON-formatted text data entered by the user, and the output is an HTTP request to the server.

[0614] Step 3:

[0615] The server receives the digital data sent from the device and formats it for analysis. Specifically, it converts the received JSON data into an appropriate data format that can be analyzed by the generative AI model. This conversion process includes a processing step that converts JSON into an internal data structure that can be processed by the analysis engine. The input is JSON-formatted data, and the output is internal data for the analysis engine.

[0616] Step 4:

[0617] The server's analysis engine analyzes the data using generative AI models. This analysis process uses multiple generative AI models, such as GPT, BERT, and T5, to compare the input digital data with existing datasets. Specifically, it performs grammar checks, contextual analysis, and the detection of specific lexical patterns. The input for this step is the internal data for the analysis engine, and the output is a confidence score analyzed by each generative AI model.

[0618] Step 5:

[0619] The server aggregates the analysis results from each generative AI model and calculates the final analysis result. Specifically, it combines the confidence scores obtained from each model to calculate the final match and confidence. In this step, the analysis results are consolidated using average and maximum values ​​based on context and patterns. The input is the confidence scores from each generative AI model, and the output is the overall analysis result.

[0620] Step 6:

[0621] The server formats the final analysis results in JSON format and sends them to the terminal. Specifically, the analysis results generated internally on the server are converted into JSON data format and sent back to the terminal as an HTTP response. The input is the comprehensive analysis results, and the output is JSON format data.

[0622] Step 7:

[0623] The device visually presents the analysis results received from the server to the user. For example, a message such as "There is an 85% chance that this sentence was created by a generative AI" is displayed on the screen. The input is the analysis results in JSON format received from the server, and the output is a visual display for the user.

[0624] These steps allow the system to determine in real time whether digital data was created by generative AI.

[0625] 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.

[0626] The present invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed. Specific examples of the present invention are described in detail below.

[0627] System Configuration

[0628] This system consists of the following main components:

[0629] 1. Input Device (Terminal)

[0630] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[0631] 2. Server

[0632] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[0633] 3. Analysis Engine

[0634] This software or application resides on a server and compares target data with multiple generative AI models to determine the results of generative AI creations.

[0635] 4. Emotion Engine

[0636] Software or application located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[0637] 5. Receiving Device (Terminal)

[0638] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user.

[0639] Program processing overview

[0640] 1. Enter and submit data

[0641] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[0642] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[0643] 2. Data Receipt and Analysis

[0644] The server receives the data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[0645] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[0646] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[0647] 3. Emotion Recognition Processing

[0648] The emotion engine in the server recognizes the user's emotions based on the analysis results. For text, it classifies emotions as negative, positive, neutral, etc. For images, it uses facial expression recognition technology to analyze emotions.

[0649] The server adds the emotion recognition results to the analysis results to create comprehensive feedback.

[0650] 4. Transmission and display of analysis results and emotion recognition results

[0651] The server formats the analysis results and emotion recognition results in JSON format or similar and sends them to the receiving device.

[0652] The device displays the received results in a user-friendly format, such as a message like, "There is an 85% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[0653] Specific examples

[0654] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0655] 1. Enter and submit data

[0656] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[0657] The user clicks the "Start Analysis" button and the text is sent to the server.

[0658] 2. Data Receipt and Analysis

[0659] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[0660] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[0661] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[0662] 3. Emotion Recognition Processing

[0663] The emotion engine in the server recognizes the user's emotion as positive based on the content of the text.

[0664] 4. Transmission and display of analysis results and emotion recognition results

[0665] The server formats the result of "75% match" and the recognition result of "emotion is positive" in JSON format and sends them to the receiving device.

[0666] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI. Also, the user's sentiment is positive."

[0667] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[0668] The processing flow will be explained below.

[0669] Specific processing steps of the invention that combines the emotion engine

[0670] Step 1:

[0671] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[0672] Step 2:

[0673] The terminal converts the input digital data into JSON format, which is then sent to the server.

[0674] Step 3:

[0675] The server receives the JSON data, which is then formatted for parsing and temporarily stored in a database.

[0676] Step 4:

[0677] The server adds a new analysis job to the queue, which contains the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[0678] Step 5:

[0679] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[0680] Step 6:

[0681] The analytics engine compares incoming data with multiple generative AI models, analyzing:

[0682] Contextual consistency

[0683] Grammatical accuracy

[0684] Vocabulary frequency and patterns

[0685] Prevalence of similar text or images

[0686] Step 7:

[0687] The server obtains the individual probability values ​​from each generative AI model. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, then the server combines these to calculate a match rate of 75%.

[0688] Step 8:

[0689] The server passes the results of the match analysis to the emotion engine, which then recognizes the user's emotion from the content of the received data. For text, it classifies emotions as negative, positive, neutral, etc., and for images, it uses facial expression recognition technology to analyze the emotion.

[0690] Step 9:

[0691] The server adds the emotion recognition results to the analysis results and generates comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[0692] Step 10:

[0693] The server converts the analysis results and emotion recognition results into JSON format and sends them to the receiving device as an HTTP response.

[0694] Step 11:

[0695] The device analyzes the received analysis results and emotion recognition results and displays them in a format that is easy for the user to understand. For example, a message such as "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive" may be displayed on the screen.

[0696] Step 12:

[0697] The user checks the analysis results and emotion recognition results and takes the following actions as necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[0698] Example 2

[0699] 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."

[0700] In recent years, technological advances in generative AI have made it easier to automatically generate text and images, increasing the risk of this being misused by educational institutions and businesses. Furthermore, there is a demand for feedback that takes into account the user's emotions, which has been difficult with conventional systems. Therefore, a system is needed that can simultaneously identify digital data automatically generated by generative AI, recognize the user's emotions, and provide feedback that reflects those emotions.

[0701] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for acquiring target digital data from an information processing device, means for transmitting the acquired data to an analysis device, means for the analysis device to compare and analyze the target data with multiple generative AI models, means for recognizing the user's emotions based on the analysis results, and means for returning the analysis results and emotion recognition results in a percentage format to the receiving device. This enables the generation AI to identify automatically generated data and provide feedback that reflects the user's emotions.

[0702] An "information processing device" is a device that allows a user to input digital data, and is a wide range of terminals including PCs, smartphones, tablets, etc.

[0703] The "analysis device" is a device for analyzing acquired digital data, and is a module within the server that uses multiple generative AI models to compare and analyze data.

[0704] A "generative AI model" is an algorithm that automatically generates digital data using generative AI technology, and specifically includes GPT, BERT, T5, etc.

[0705] "Digital data" refers to data provided by a user via an input device, and includes text data, image data, and the like.

[0706] "Means for recognizing emotions" refers to software or algorithms for classifying and recognizing a user's emotions based on the analysis results, including text analysis and facial expression recognition technologies.

[0707] A "receiving device" is a device that receives the analysis results and emotion recognition results and presents them to the user, and includes smartphones, PCs, tablets, etc.

[0708] The "means for transmitting data" refers to a protocol or interface for transmitting digital data from the information processing device to the analysis device.

[0709] This invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed.

[0710] This system consists of the following main components:

[0711] 1. Input devices (information processing devices)

[0712] A device that allows users to input digital data (text and images). Examples of this device include PCs, smartphones, and tablets.

[0713] 2. Server (analysis device)

[0714] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[0715] 3. Analysis Engine

[0716] It is software or a program located on a server that compares target data with multiple generative AI models to determine the results of generative AI creations. Specifically, models such as GPT, BERT, and T5 are used.

[0717] 4. Emotion Engine

[0718] Software or a program located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[0719] 5. Receiving Device

[0720] This is a device that receives analysis results and emotion recognition results sent from the server and presents them to the user.

[0721] Specific processing of the program

[0722] Digital data entered by a user through an information processing device is first sent to a server. The server's analysis engine compares the received data with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. Next, an emotion engine recognizes the user's emotions. The analysis results and emotion recognition results are integrated and sent to the receiving device, where they are finally presented to the user.

[0723] Specific examples

[0724] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0725] The user enters the following sentence into the form on their PC: "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." When the user clicks the "Start Analysis" button, the sentence is sent to the server.

[0726] The server formats the text data and passes it to the analysis engine. The analysis engine compares the text with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. For example, if GPT reports a match of 70%, BERT reports a match of 75%, and T5 reports a match of 80%, these will be combined to arrive at a final match of 75%.

[0727] Next, the emotion engine on the server recognizes the user's emotion as positive based on the content of the sentence. This integrates the analysis results with the emotion recognition results, and displays on the receiving device the message, "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive."

[0728] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[0729] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0730] Step 1:

[0731] Users input text or images into a form on an input device such as a PC, smartphone, or tablet. Specifically, they enter text into a text area or upload an image file. This input data (e.g., the text "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." or an image file) is converted to JSON format on the device.

[0732] Input: Text or image data entered by a user into an input device

[0733] Output: Digital data formatted in JSON format

[0734] Step 2:

[0735] The user clicks the "Start Analysis" button on the input device. This operation sends the digital data to the server. When the send button is clicked, the device starts communication to send the data to the server.

[0736] Input: Digital data formatted in JSON format

[0737] Output: Server receives data

[0738] Step 3:

[0739] The server receives the digital data sent from the input device, where it is reconfirmed in JSON format and prepared to be passed to the analysis engine.

[0740] Input: JSON formatted data sent from the input device

[0741] Output: Formatted data to be passed to the analysis engine

[0742] Step 4:

[0743] The server's analytics engine analyzes the incoming digital data. This engine then invokes generative AI models such as GPT, BERT, and T5 to match the data. A match is calculated for each model, comparing context, grammar, and vocabulary usage patterns.

[0744] Input: Formatted digital data

[0745] Output: Match score by each generative AI model (e.g., 70% for GPT, 75% for BERT, 80% for T5)

[0746] Step 5:

[0747] The server then combines the match scores from each generative AI model to calculate the overall match score. For example, the match scores (70% for GPT, 75% for BERT, and 80% for T5) are averaged to calculate the final match score (e.g., 75%).

[0748] Input: Match scores from each generative AI model

[0749] Output: Overall match (e.g. 75%)

[0750] Step 6:

[0751] The emotion engine in the server recognizes the user's emotion based on the analysis results. For text, it classifies it as negative, positive, or neutral. For example, the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." would be classified as positive. For images, emotion is analyzed using facial expression recognition technology.

[0752] Input: Formatted data and analysis results

[0753] Output: User's emotion category (e.g., positive)

[0754] Step 7:

[0755] The server combines the emotion recognition results with the analysis results to create comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[0756] Input: Overall agreement and emotion category

[0757] Output: General feedback message

[0758] Step 8:

[0759] The server formats the aggregate feedback message in JSON format and sends it to the receiving device.

[0760] Input: General feedback message

[0761] Output: Feedback data in JSON format

[0762] Step 9:

[0763] The device displays the feedback data received from the server to the user, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[0764] Input: Feedback data in JSON format

[0765] Output: Feedback displayed in a format that the user can understand

[0766] (Application example 2)

[0767] 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."

[0768] Currently, it is difficult for brick-and-mortar stores to determine the authenticity and sentiment of customers' text messages in real time. This makes it difficult for customer service staff to respond appropriately, potentially hindering the improvement of the customer experience. It is also challenging to respond appropriately to false information and messages created by generative AI. These issues need to be resolved.

[0769] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0770] In this invention, the server includes a means for acquiring target digital data from an input device, a means for transmitting the acquired data to an analysis device, a means for the analysis device to compare and analyze the target data with multiple generative AI models, and a means for analyzing emotions and displaying the recognition results on a receiving device. This allows customer service staff to confirm the authenticity of customer messages and simultaneously recognize customer emotions in real time. This enables more appropriate and prompt customer service, and is expected to improve the customer experience.

[0771] "Target digital data" refers to any form of digital data that is the subject of analysis, including text, images, etc.

[0772] An "input device" is a device that allows a user to input digital data, and includes PCs, smartphones, tablets, smart glasses, etc.

[0773] An "analysis device" is a computer system that acquires target digital data and analyzes it using an analysis engine and an emotion recognition engine.

[0774] A "generative AI model" is a model generated using artificial intelligence, such as a natural language processing model such as GPT or BERT.

[0775] The "receiving device" is a device that receives the analysis results and emotion recognition results sent from the analyzing device and presents them to the user.

[0776] The "means for analyzing emotions and displaying the recognition results on the receiving device" is a means having the function of analyzing the user's emotions from the target digital data and displaying the results on the receiving device in real time.

[0777] This invention relates to a system that determines whether target digital data was created by generative AI and simultaneously recognizes the user's emotions. This system is intended to be used primarily in customer service situations in brick-and-mortar stores.

[0778] System Configuration

[0779] The system includes the following major components:

[0780] 1. Input Devices

[0781] A device that allows customers to input digital data (text and images). This includes wearable devices such as smart glasses.

[0782] 2. Analysis device

[0783] This is a computer system that receives and analyzes data sent from an input device. This device has a built-in analysis engine and emotion engine.

[0784] 3. Analysis Engine

[0785] This software is located on a server and compares the target data with multiple generative AI models to determine the results of generative AI creations. GPT and BERT are used as generative AI models.

[0786] 4. Emotion Engine

[0787] This software is located on the server and analyzes target data to recognize user emotions. It uses text analysis and image analysis techniques.

[0788] 5. Receiving Device

[0789] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user. Smart glasses are used as this receiving device.

[0790] Hardware and software used

[0791] Hardware

[0792] Smart glasses (e.g., Google Glass, Vuzix Blade)

[0793] Server (e.g. AWS, Google Cloud)

[0794] software

[0795] Web framework: Flask

[0796] Generative AI decision engine: AIAnalyzer (library of hypotheses)

[0797] Emotion recognition engine: SentimentAnalyzer (hypothetical library)

[0798] Data processing and calculation methods

[0799] The analysis device (server) performs the following processes.

[0800] 1. Receiving and formatting data

[0801] The data sent from the input device is received by the server, where it is converted to JSON format or similar and formatted for analysis.

[0802] 2. Discrimination by Generative AI

[0803] Using a generative AI model (e.g., GPT, BERT, etc.), the probability that the target data is generated by generative AI is analyzed. The analysis result is expressed in percentage form.

[0804] 3. Emotion recognition

[0805] The emotion engine analyzes the user's emotions from the target data and classifies them as positive, negative, neutral, etc.

[0806] 4. Sending the results

[0807] The analysis results and emotion recognition results are formatted and transmitted to a receiving device.

[0808] Specific examples

[0809] Let's take the example of a customer in a brick-and-mortar clothing store who sends a message asking, "Is this clothing popular on social media?" This message is sent to the server as a prompt as follows:

[0810] Prompt Sentence Examples

[0811] Is this outfit creating a buzz on social media?

[0812] The analysis results are displayed as follows: "Probability of message generated by AI: 85%, Emotion: Neutral." Based on this information, customer service staff can quickly respond appropriately.

[0813] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0814] Step 1:

[0815] A user inputs target digital data such as text or images into an input device such as smart glasses. Specifically, the user inputs a message such as "Is this outfit trending on social media?" using voice or text input. This input data is sent to the smart glasses.

[0816] Step 2:

[0817] The smart glasses send the input digital data to the server, where it is converted into JSON format and passed to an analysis device. If the input sentence is something like "Is this clothing popular on social media?", it will be sent as JSON data.

[0818] Step 3:

[0819] The analysis device receives the data received by the server and formats it. The formatted data is converted into a format that can be analyzed by a generative AI model (GPT, BERT, etc.). For example, text data is formatted according to a standard format.

[0820] Step 4:

[0821] The server's analysis engine compares and analyzes the formatted data with the generative AI model. Specifically, input data is fed into a generative AI model such as GPT or BERT, and the degree of match and probability are calculated to determine whether the data was generated by the generative AI. For example, for the data "Is this clothing item trending on social media?", the probability that it was generated by the generative AI is determined to be 85%.

[0822] Step 5:

[0823] The server's emotion engine recognizes the user's emotion based on the analysis results. Text analysis technology is used to classify emotions from the target data (text). For example, the emotion engine will recognize the sentence "Is this clothing popular on social media?" as "neutral."

[0824] Step 6:

[0825] The server formats the generated AI's judgment results and emotion recognition results into JSON format and sends them to the smart glasses, which act as a receiving device. The formatted data includes the generated AI's probability (85%) and emotion classification (neutral).

[0826] Step 7:

[0827] The smart glasses then present the analysis results and emotion recognition results to the user. The display shows "Probability of message generated by AI: 85%, Emotion: Neutral," and the customer service staff can use this information to respond appropriately.

[0828] 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.

[0829] 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.

[0830] 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.

[0831] [Third embodiment]

[0832] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0833] 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.

[0834] 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).

[0835] 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.

[0836] 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.

[0837] 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).

[0838] 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.

[0839] 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.

[0840] 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.

[0841] 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.

[0842] 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.

[0843] 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."

[0844] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. Specific examples of the present invention are described in detail below.

[0845] System Configuration

[0846] This system consists of the following main components:

[0847] 1. Input Device (Terminal)

[0848] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[0849] 2. Server

[0850] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0851] 3. Analysis Engine

[0852] It is software or an application located on a server that compares target data with multiple generative AI models and calculates the degree of match. The analytical engine uses natural language processing technology and machine learning algorithms.

[0853] 4. Receiving device (terminal)

[0854] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0855] Program processing overview

[0856] 1. Enter and submit data

[0857] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[0858] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[0859] 2. Data Receipt and Analysis

[0860] The server receives data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[0861] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[0862] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[0863] 3. Sending and displaying analysis results

[0864] The server formats the analysis results in JSON format or similar and sends them to the receiving device.

[0865] The device visually presents the analysis results to the user, for example, displaying a message such as "There is an 85% chance that this sentence was created by a generative AI."

[0866] Specific examples

[0867] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[0868] 1. Enter and submit data

[0869] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[0870] The user clicks the "Start Analysis" button and the text is sent to the server.

[0871] 2. Data Receipt and Analysis

[0872] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[0873] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[0874] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[0875] 3. Sending and displaying analysis results

[0876] The server formats the "75% match" result in JSON format and sends it to the receiving device.

[0877] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI."

[0878] In this way, the invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing inappropriate use of generative AI.

[0879] The processing flow will be explained below.

[0880] Detailed processing steps of the program

[0881] Step 1:

[0882] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[0883] Step 2:

[0884] The terminal converts the input digital data into JSON format, which makes it easier for the server to parse.

[0885] Step 3:

[0886] The device sends the converted JSON data as an HTTP POST request to the server's API endpoint.

[0887] Step 4:

[0888] The server receives the POST request, converts the received data from JSON back to an internal parsing format, and temporarily stores it in a database.

[0889] Step 5:

[0890] The server adds new analysis jobs to the queue, which contain the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[0891] Step 6:

[0892] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[0893] Step 7:

[0894] The analytics engine compares the incoming data with multiple generative AI models, specifically analyzing the following:

[0895] Contextual consistency

[0896] Grammatical accuracy

[0897] Vocabulary frequency and patterns

[0898] Prevalence of similar text or images

[0899] Step 8:

[0900] The server obtains individual probability values ​​from each generative AI model and combines them to calculate the final match rate. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, they will be combined to calculate a match rate of 75%.

[0901] Step 9:

[0902] The server converts the analysis results, including the final degree of match, into JSON format and sends it to the terminal as an HTTP response.

[0903] Step 10:

[0904] The device then analyzes the analysis results and displays them in a format that is easy for the user to understand. For example, the device might display something like, "There is a 75% chance that this sentence was created by a generative AI."

[0905] Step 11:

[0906] The user checks the analysis results and takes the next action if necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[0907] Example 1

[0908] 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."

[0909] In recent years, digital data created by generative AI has become widely used in educational institutions and companies, raising concerns about its misuse and unauthorized use. The present invention aims to solve these problems and promote the proper use of generative AI by providing a system that can accurately determine whether digital data was created by generative AI.

[0910] 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.

[0911] In this invention, the server includes a means for transmitting target digital data to an analytical engine, a means for the analytical engine to compare and analyze the target data with multiple generative AI models, and a means for calculating a final degree of match by combining the degrees of match obtained from each generative AI model, thereby making it possible to determine with high accuracy whether digital data was created by a generative AI.

[0912] "Target digital data" refers to electronic text or image data acquired from an input device as the target for analysis.

[0913] An "input device" is a device such as a computer, smartphone, or tablet that allows a user to input digital data.

[0914] "Analysis Engine" means a program or software that compares and analyzes acquired digital data with a generated AI model.

[0915] A "generative AI model" is an artificial intelligence model trained to generate text and images using natural language processing and machine learning algorithms.

[0916] "Matchability" is a numerical value that indicates the likelihood that the target digital data was created by a generative AI model, and is calculated as an analysis result.

[0917] The "receiving device" is a device that receives the analysis results sent from the server and presents them to the user.

[0918] The "percentage display format" is a method of expressing the degree of agreement of the analysis results as a percentage.

[0919] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. A specific embodiment of this system is described in detail below.

[0920] System Configuration

[0921] This system consists of the following main components:

[0922] 1. Input Device (Terminal)

[0923] A device that allows a user to input digital data (text or images). This device includes personal computers, smartphones, tablets, etc.

[0924] 2. Server

[0925] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[0926] 3. Analysis Engine

[0927] It is a software or application located on a server that compares target data with multiple generative AI models and calculates the degree of agreement. The analysis engine uses natural language processing technology and machine learning algorithms, specifically models such as GPT, BERT, and T5.

[0928] 4. Receiving device (terminal)

[0929] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[0930] Entering and Submitting Data

[0931] The user uses an input device to input the digital data to be analyzed (e.g., text or images) into the input form. Specifically, the user types text using a keyboard or selects an image from a file selection dialog.

[0932] The user clicks the "Start Analysis" button in the input form. This action sends the input data to the server. Specifically, an HTTP request is generated and a payload containing the data is sent to the server.

[0933] Receiving and analyzing data

[0934] The server receives the data sent by the user. Specifically, the server's web server captures the HTTP request and retrieves the body data.

[0935] The server converts the received data into a standard format such as JSON and formats it for analysis. Specifically, a data parser analyzes the raw data, extracts the necessary information, and formats it.

[0936] The server's analysis engine compares and analyzes the received data with multiple generative AI models (e.g., GPT, BERT, T5). Specifically, the server calls the analysis engine, inputs the data into the generative AI model, and analyzes the context, grammar, and vocabulary usage patterns.

[0937] The server then combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm to combine the match scores from different models and calculates the final score using methods such as a weighted average.

[0938] Sending and displaying analysis results

[0939] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results and serializes them into a JSON-formatted string.

[0940] The server then sends the formatted analysis results to the receiving device. Specifically, it generates an HTTP response and sends it to the device.

[0941] The device visually presents the analysis results to the user. Specifically, the device's web application parses the JSON data and displays a message saying, "There is an 85% chance that this sentence was created by generative AI."

[0942] Specific examples

[0943] For example, consider a case where a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0944] Prompt Sentence Examples

[0945] Below are some example prompts to input to the generative AI model:

[0946] "Please help me write the report that's due today."

[0947] "Write a data analysis explanation for a school project."

[0948] "Please translate the following text into an English fan fiction story."

[0949] This prompt statement can be used to check the system's operation and to demonstrate it.

[0950] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0951] Step 1:

[0952] A user inputs digital data using an input device. The user enters text or images to be analyzed into an input form. In this process, the user types text content using a keyboard or uploads an image file using a file selection dialog. The input data is temporarily stored in local storage on the input device. Input = digital data to be analyzed, Output = raw data stored on the input device.

[0953] Step 2:

[0954] The user clicks the "Start Analysis" button, which sends the input data to the server. Specifically, the input device generates an HTTP POST request, which includes the input data as the payload. Input = raw data stored on the input device, Output = data payload sent to the server.

[0955] Step 3:

[0956] The server receives data sent from the input device. The server's web server captures the HTTP request and obtains the payload data. Input = data payload sent from the input device, Output = formatted data to pass to the analysis engine.

[0957] Step 4:

[0958] The server converts the data it receives into a standard format such as JSON and formats it for analysis. Specifically, the server's data parser analyzes the raw data, extracts the necessary information, and serializes it. Input = data payload received by the server, output = formatted data in JSON format.

[0959] Step 5:

[0960] The server's analysis engine compares and analyzes the formatted data with multiple generative AI models. Specifically, the server calls the analysis engine, inputs the formatted data into generative AI models such as GPT, BERT, and T5, and analyzes the context, grammar, and vocabulary usage patterns. Input = formatted data in JSON format, output = match score for each generative AI model.

[0961] Step 6:

[0962] The server combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm (e.g., weighted average) to combine the match scores from different models to calculate the overall score. Input = match score for each generative AI model, output = final match score.

[0963] Step 7:

[0964] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results, serializes it into a JSON-formatted string, and sends it as an HTTP response. Input = final match score, output = JSON-formatted analysis results sent to the receiving device.

[0965] Step 8:

[0966] The device visually presents the analysis results received from the server to the user. Specifically, the device's web application parses the JSON data and displays the message, "There is an 85% chance that this sentence was created by generative AI." Input = JSON-formatted analysis results received from the server, Output = visual analysis results presented to the user.

[0967] Specific examples

[0968] Consider a case in which a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[0969] Prompt Sentence Examples

[0970] Below are some example prompts to input to the generative AI model:

[0971] "Please help me write the report that's due today."

[0972] "Write a data analysis explanation for a school project."

[0973] "Please translate the following text into an English fan fiction story."

[0974] This prompt statement can be used to check the system's operation and to demonstrate it.

[0975] (Application example 1)

[0976] 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."

[0977] In recent years, the generation of digital data using generative AI has rapidly spread, resulting in a need to determine whether it was created by generative AI. As this technology becomes more widespread, the risk of unauthorized use of text and images created by generative AI also increases. It is particularly important for educational institutions and companies to properly monitor the use of generative AI and take necessary measures. However, current systems face the challenge of making this determination efficiently and accurately.

[0978] 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.

[0979] In this invention, the server includes means for transmitting digital data input from a user terminal to an analysis server, means for comparing and analyzing the target data using multiple generative AI models, and means for returning the analysis results in percentage format to the terminal, thereby enabling accurate real-time determination of whether the digital data was created by a generative AI.

[0980] A "terminal" is a device through which a user inputs digital data, and specifically refers to a personal computer, smartphone, tablet, etc.

[0981] An "analysis server" is a computer system that receives digital data sent from a terminal and analyzes the data using a generative AI model.

[0982] A "generative AI model" is an algorithm or model that uses artificial intelligence technology to generate digital data, and examples include GPT, BERT, and T5.

[0983] "Comparison and analysis" is the process in which the analysis server uses a generative AI model to compare input data with known data sets and analyze the data's features and patterns.

[0984] The "percentage display format" is a display format in which the analysis results are quantified and the degree of agreement or reliability is expressed as a percentage.

[0985] "Visual presentation means" refers to an interface or method for displaying analysis results in a way that is easy for users to understand, such as using graphs, text, charts, etc.

[0986] "Synthesis" refers to the process of integrating the results obtained from multiple generative AI models to derive the final analysis results.

[0987] In order to implement the present invention, the following system and means are utilized: Each component of the system is designed based on the claims.

[0988] System Configuration

[0989] This system consists of the following main components:

[0990] 1. Device:

[0991] A terminal is a device through which a user inputs digital data. Specifically, a personal computer, smartphone, tablet, etc. is used. The user inputs digital data into the terminal and sends it to an analysis server.

[0992] 2. Analysis server:

[0993] The analysis server is a computer system that receives digital data sent from the terminal and analyzes the data using a generative AI model. The server has generative AI models such as GPT, BERT, and T5 installed, and processes the data for analysis.

[0994] 3. Generative AI Model:

[0995] A generative AI model is an algorithm or model that uses generative AI to generate digital data. Examples include GPT, BERT, and T5. This model analyzes input digital data by comparing it with existing datasets.

[0996] 4. Data Receipt and Analysis:

[0997] The server receives the digital data sent from the device and converts it into a standard format such as JSON. The converted data is formatted for analysis and compared and analyzed using a generative AI model. The analysis results are calculated in the form of a confidence level (%).

[0998] 5. Presentation of results:

[0999] The analysis results are returned to the terminal in the form of a percentage. The terminal provides an interface to visually present the analysis results to the user, allowing the user to easily understand whether the digital data was created by generative AI.

[1000] Process flow:

[1001] As a concrete example, consider the analysis of the sentence "He finished writing his thesis."

[1002] 1. Data entry and transmission via terminal:

[1003] The user enters the sentence "He finished writing his thesis" into the input form on the terminal. By pressing the send button, this data is sent to the analysis server.

[1004] 2. Data reception and analysis by the server:

[1005] The analysis server formats the received digital data for analysis and performs the analysis using a generative AI model, such as GPT, BERT, or T5, to analyze the grammar and structure of the sentence.

[1006] 3. Sending and displaying analysis results:

[1007] The server calculates the analysis results and returns them to the device in the form of a percentage, which then displays a message to the user, such as "There is an 85% chance that this sentence was created by a generative AI."

[1008] The system uses Python hoformat to send and receive data between the server and the device. Specific analysis is performed based on a generative AI model, and the reliability of the analysis results is calculated.

[1009] The above process and components used enable the present invention to accurately determine in real time whether digital data entered by a user was created by generative AI.

[1010] Example prompt sentence:

[1011] "He finished writing his thesis."

[1012] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1013] Step 1:

[1014] A user uses a terminal to input digital data. Specifically, the user enters text data such as "He has finished writing his thesis" into the terminal's input form and clicks the "Submit" button. This input data is later formatted into JSON format to be sent to an analysis server. At this stage, the input is the text data entered by the user, and the output is JSON format data.

[1015] Step 2:

[1016] The terminal sends the digital data entered by the user to the analysis server. The terminal uses an HTTP POST request to send JSON-formatted data to the server. This request allows the data sent from the terminal to reach the analysis server, allowing the server to begin processing the received data. The input is JSON-formatted text data entered by the user, and the output is an HTTP request to the server.

[1017] Step 3:

[1018] The server receives the digital data sent from the device and formats it for analysis. Specifically, it converts the received JSON data into an appropriate data format that can be analyzed by the generative AI model. This conversion process includes a processing step that converts JSON into an internal data structure that can be processed by the analysis engine. The input is JSON-formatted data, and the output is internal data for the analysis engine.

[1019] Step 4:

[1020] The server's analysis engine analyzes the data using generative AI models. This analysis process uses multiple generative AI models, such as GPT, BERT, and T5, to compare the input digital data with existing datasets. Specifically, it performs grammar checks, contextual analysis, and the detection of specific lexical patterns. The input for this step is the internal data for the analysis engine, and the output is a confidence score analyzed by each generative AI model.

[1021] Step 5:

[1022] The server aggregates the analysis results from each generative AI model and calculates the final analysis result. Specifically, it combines the confidence scores obtained from each model to calculate the final match and confidence. In this step, the analysis results are consolidated using average and maximum values ​​based on context and patterns. The input is the confidence scores from each generative AI model, and the output is the overall analysis result.

[1023] Step 6:

[1024] The server formats the final analysis results in JSON format and sends them to the terminal. Specifically, the analysis results generated internally on the server are converted into JSON data format and sent back to the terminal as an HTTP response. The input is the comprehensive analysis results, and the output is JSON format data.

[1025] Step 7:

[1026] The device visually presents the analysis results received from the server to the user. For example, a message such as "There is an 85% chance that this sentence was created by a generative AI" is displayed on the screen. The input is the analysis results in JSON format received from the server, and the output is a visual display for the user.

[1027] These steps allow the system to determine in real time whether digital data was created by generative AI.

[1028] 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.

[1029] The present invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed. Specific examples of the present invention are described in detail below.

[1030] System Configuration

[1031] This system consists of the following main components:

[1032] 1. Input Device (Terminal)

[1033] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[1034] 2. Server

[1035] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[1036] 3. Analysis Engine

[1037] This software or application resides on a server and compares target data with multiple generative AI models to determine the results of generative AI creations.

[1038] 4. Emotion Engine

[1039] Software or application located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[1040] 5. Receiving Device (Terminal)

[1041] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user.

[1042] Program processing overview

[1043] 1. Enter and submit data

[1044] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[1045] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[1046] 2. Data Receipt and Analysis

[1047] The server receives the data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[1048] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[1049] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[1050] 3. Emotion Recognition Processing

[1051] The emotion engine in the server recognizes the user's emotions based on the analysis results. For text, it classifies emotions as negative, positive, neutral, etc. For images, it uses facial expression recognition technology to analyze emotions.

[1052] The server adds the emotion recognition results to the analysis results to create comprehensive feedback.

[1053] 4. Transmission and display of analysis results and emotion recognition results

[1054] The server formats the analysis results and emotion recognition results in JSON format or similar and sends them to the receiving device.

[1055] The device displays the received results in a user-friendly format, such as a message like, "There is an 85% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[1056] Specific examples

[1057] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[1058] 1. Enter and submit data

[1059] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[1060] The user clicks the "Start Analysis" button and the text is sent to the server.

[1061] 2. Data Receipt and Analysis

[1062] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[1063] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[1064] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[1065] 3. Emotion Recognition Processing

[1066] The emotion engine in the server recognizes the user's emotion as positive based on the content of the text.

[1067] 4. Transmission and display of analysis results and emotion recognition results

[1068] The server formats the result of "75% match" and the recognition result of "emotion is positive" in JSON format and sends them to the receiving device.

[1069] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI. Also, the user's sentiment is positive."

[1070] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[1071] The processing flow will be explained below.

[1072] Specific processing steps of the invention that combines the emotion engine

[1073] Step 1:

[1074] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[1075] Step 2:

[1076] The terminal converts the input digital data into JSON format, which is then sent to the server.

[1077] Step 3:

[1078] The server receives the JSON data, which is then formatted for parsing and temporarily stored in a database.

[1079] Step 4:

[1080] The server adds a new analysis job to the queue, which contains the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[1081] Step 5:

[1082] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[1083] Step 6:

[1084] The analytics engine compares incoming data with multiple generative AI models, analyzing:

[1085] Contextual consistency

[1086] Grammatical accuracy

[1087] Vocabulary frequency and patterns

[1088] Prevalence of similar text or images

[1089] Step 7:

[1090] The server obtains the individual probability values ​​from each generative AI model. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, then the server combines these to calculate a match rate of 75%.

[1091] Step 8:

[1092] The server passes the results of the match analysis to the emotion engine, which then recognizes the user's emotion from the content of the received data. For text, it classifies emotions as negative, positive, neutral, etc., and for images, it uses facial expression recognition technology to analyze the emotion.

[1093] Step 9:

[1094] The server adds the emotion recognition results to the analysis results and generates comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[1095] Step 10:

[1096] The server converts the analysis results and emotion recognition results into JSON format and sends them to the receiving device as an HTTP response.

[1097] Step 11:

[1098] The device analyzes the received analysis results and emotion recognition results and displays them in a format that is easy for the user to understand. For example, a message such as "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive" may be displayed on the screen.

[1099] Step 12:

[1100] The user checks the analysis results and emotion recognition results and takes the following actions as necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[1101] Example 2

[1102] 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."

[1103] In recent years, technological advances in generative AI have made it easier to automatically generate text and images, increasing the risk of this being misused by educational institutions and businesses. Furthermore, there is a demand for feedback that takes into account the user's emotions, which has been difficult with conventional systems. Therefore, a system is needed that can simultaneously identify digital data automatically generated by generative AI, recognize the user's emotions, and provide feedback that reflects those emotions.

[1104] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for acquiring target digital data from an information processing device, means for transmitting the acquired data to an analysis device, means for the analysis device to compare and analyze the target data with multiple generative AI models, means for recognizing the user's emotions based on the analysis results, and means for returning the analysis results and emotion recognition results in a percentage format to the receiving device. This enables the generation AI to identify automatically generated data and provide feedback that reflects the user's emotions.

[1105] An "information processing device" is a device that allows a user to input digital data, and is a wide range of terminals including PCs, smartphones, tablets, etc.

[1106] The "analysis device" is a device for analyzing acquired digital data, and is a module within the server that uses multiple generative AI models to compare and analyze data.

[1107] A "generative AI model" is an algorithm that automatically generates digital data using generative AI technology, and specifically includes GPT, BERT, T5, etc.

[1108] "Digital data" refers to data provided by a user via an input device, and includes text data, image data, and the like.

[1109] "Means for recognizing emotions" refers to software or algorithms for classifying and recognizing a user's emotions based on the analysis results, including text analysis and facial expression recognition technologies.

[1110] A "receiving device" is a device that receives the analysis results and emotion recognition results and presents them to the user, and includes smartphones, PCs, tablets, etc.

[1111] The "means for transmitting data" refers to a protocol or interface for transmitting digital data from the information processing device to the analysis device.

[1112] This invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed.

[1113] This system consists of the following main components:

[1114] 1. Input devices (information processing devices)

[1115] A device that allows users to input digital data (text and images). Examples of this device include PCs, smartphones, and tablets.

[1116] 2. Server (analysis device)

[1117] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[1118] 3. Analysis Engine

[1119] It is software or a program located on a server that compares target data with multiple generative AI models to determine the results of generative AI creations. Specifically, models such as GPT, BERT, and T5 are used.

[1120] 4. Emotion Engine

[1121] Software or a program located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[1122] 5. Receiving Device

[1123] This is a device that receives analysis results and emotion recognition results sent from the server and presents them to the user.

[1124] Specific processing of the program

[1125] Digital data entered by a user through an information processing device is first sent to a server. The server's analysis engine compares the received data with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. Next, an emotion engine recognizes the user's emotions. The analysis results and emotion recognition results are integrated and sent to the receiving device, where they are finally presented to the user.

[1126] Specific examples

[1127] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[1128] The user enters the following sentence into the form on their PC: "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." When the user clicks the "Start Analysis" button, the sentence is sent to the server.

[1129] The server formats the text data and passes it to the analysis engine. The analysis engine compares the text with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. For example, if GPT reports a match of 70%, BERT reports a match of 75%, and T5 reports a match of 80%, these will be combined to arrive at a final match of 75%.

[1130] Next, the emotion engine on the server recognizes the user's emotion as positive based on the content of the sentence. This integrates the analysis results with the emotion recognition results, and displays on the receiving device the message, "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive."

[1131] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[1132] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1133] Step 1:

[1134] Users input text or images into a form on an input device such as a PC, smartphone, or tablet. Specifically, they enter text into a text area or upload an image file. This input data (e.g., the text "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." or an image file) is converted to JSON format on the device.

[1135] Input: Text or image data entered by a user into an input device

[1136] Output: Digital data formatted in JSON format

[1137] Step 2:

[1138] The user clicks the "Start Analysis" button on the input device. This operation sends the digital data to the server. When the send button is clicked, the device starts communication to send the data to the server.

[1139] Input: Digital data formatted in JSON format

[1140] Output: Server receives data

[1141] Step 3:

[1142] The server receives the digital data sent from the input device, where it is reconfirmed in JSON format and prepared to be passed to the analysis engine.

[1143] Input: JSON formatted data sent from the input device

[1144] Output: Formatted data to be passed to the analysis engine

[1145] Step 4:

[1146] The server's analytics engine analyzes the incoming digital data. This engine then invokes generative AI models such as GPT, BERT, and T5 to match the data. A match is calculated for each model, comparing context, grammar, and vocabulary usage patterns.

[1147] Input: Formatted digital data

[1148] Output: Match score by each generative AI model (e.g., 70% for GPT, 75% for BERT, 80% for T5)

[1149] Step 5:

[1150] The server then combines the match scores from each generative AI model to calculate the overall match score. For example, the match scores (70% for GPT, 75% for BERT, and 80% for T5) are averaged to calculate the final match score (e.g., 75%).

[1151] Input: Match scores from each generative AI model

[1152] Output: Overall match (e.g. 75%)

[1153] Step 6:

[1154] The emotion engine in the server recognizes the user's emotion based on the analysis results. For text, it classifies it as negative, positive, or neutral. For example, the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." would be classified as positive. For images, emotion is analyzed using facial expression recognition technology.

[1155] Input: Formatted data and analysis results

[1156] Output: User's emotion category (e.g., positive)

[1157] Step 7:

[1158] The server combines the emotion recognition results with the analysis results to create comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[1159] Input: Overall agreement and emotion category

[1160] Output: General feedback message

[1161] Step 8:

[1162] The server formats the aggregate feedback message in JSON format and sends it to the receiving device.

[1163] Input: General feedback message

[1164] Output: Feedback data in JSON format

[1165] Step 9:

[1166] The device displays the feedback data received from the server to the user, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[1167] Input: Feedback data in JSON format

[1168] Output: Feedback displayed in a format that the user can understand

[1169] (Application example 2)

[1170] 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."

[1171] Currently, it is difficult for brick-and-mortar stores to determine the authenticity and sentiment of customers' text messages in real time. This makes it difficult for customer service staff to respond appropriately, potentially hindering the improvement of the customer experience. It is also challenging to respond appropriately to false information and messages created by generative AI. These issues need to be resolved.

[1172] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1173] In this invention, the server includes a means for acquiring target digital data from an input device, a means for transmitting the acquired data to an analysis device, a means for the analysis device to compare and analyze the target data with multiple generative AI models, and a means for analyzing emotions and displaying the recognition results on a receiving device. This allows customer service staff to confirm the authenticity of customer messages and simultaneously recognize customer emotions in real time. This enables more appropriate and prompt customer service, and is expected to improve the customer experience.

[1174] "Target digital data" refers to any form of digital data that is the subject of analysis, including text, images, etc.

[1175] An "input device" is a device that allows a user to input digital data, and includes PCs, smartphones, tablets, smart glasses, etc.

[1176] An "analysis device" is a computer system that acquires target digital data and analyzes it using an analysis engine and an emotion recognition engine.

[1177] A "generative AI model" is a model generated using artificial intelligence, such as a natural language processing model such as GPT or BERT.

[1178] The "receiving device" is a device that receives the analysis results and emotion recognition results sent from the analyzing device and presents them to the user.

[1179] The "means for analyzing emotions and displaying the recognition results on the receiving device" is a means having the function of analyzing the user's emotions from the target digital data and displaying the results on the receiving device in real time.

[1180] This invention relates to a system that determines whether target digital data was created by generative AI and simultaneously recognizes the user's emotions. This system is intended to be used primarily in customer service situations in brick-and-mortar stores.

[1181] System Configuration

[1182] The system includes the following major components:

[1183] 1. Input Devices

[1184] A device that allows customers to input digital data (text and images). This includes wearable devices such as smart glasses.

[1185] 2. Analysis device

[1186] This is a computer system that receives and analyzes data sent from an input device. This device has a built-in analysis engine and emotion engine.

[1187] 3. Analysis Engine

[1188] This software is located on a server and compares the target data with multiple generative AI models to determine the results of generative AI creations. GPT and BERT are used as generative AI models.

[1189] 4. Emotion Engine

[1190] This software is located on the server and analyzes target data to recognize user emotions. It uses text analysis and image analysis techniques.

[1191] 5. Receiving Device

[1192] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user. Smart glasses are used as this receiving device.

[1193] Hardware and software used

[1194] Hardware

[1195] Smart glasses (e.g., Google Glass, Vuzix Blade)

[1196] Server (e.g. AWS, Google Cloud)

[1197] software

[1198] Web framework: Flask

[1199] Generative AI decision engine: AIAnalyzer (library of hypotheses)

[1200] Emotion recognition engine: SentimentAnalyzer (hypothetical library)

[1201] Data processing and calculation methods

[1202] The analysis device (server) performs the following processes.

[1203] 1. Receiving and formatting data

[1204] The data sent from the input device is received by the server, where it is converted to JSON format or similar and formatted for analysis.

[1205] 2. Discrimination by Generative AI

[1206] Using a generative AI model (e.g., GPT, BERT, etc.), the probability that the target data is generated by generative AI is analyzed. The analysis result is expressed in percentage form.

[1207] 3. Emotion recognition

[1208] The emotion engine analyzes the user's emotions from the target data and classifies them as positive, negative, neutral, etc.

[1209] 4. Sending the results

[1210] The analysis results and emotion recognition results are formatted and transmitted to a receiving device.

[1211] Specific examples

[1212] Let's take the example of a customer in a brick-and-mortar clothing store who sends a message asking, "Is this clothing popular on social media?" This message is sent to the server as a prompt as follows:

[1213] Prompt Sentence Examples

[1214] Is this outfit creating a buzz on social media?

[1215] The analysis results are displayed as follows: "Probability of message generated by AI: 85%, Emotion: Neutral." Based on this information, customer service staff can quickly respond appropriately.

[1216] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1217] Step 1:

[1218] A user inputs target digital data such as text or images into an input device such as smart glasses. Specifically, the user inputs a message such as "Is this outfit trending on social media?" using voice or text input. This input data is sent to the smart glasses.

[1219] Step 2:

[1220] The smart glasses send the input digital data to the server, where it is converted into JSON format and passed to an analysis device. If the input sentence is something like "Is this clothing popular on social media?", it will be sent as JSON data.

[1221] Step 3:

[1222] The analysis device receives the data received by the server and formats it. The formatted data is converted into a format that can be analyzed by a generative AI model (GPT, BERT, etc.). For example, text data is formatted according to a standard format.

[1223] Step 4:

[1224] The server's analysis engine compares and analyzes the formatted data with the generative AI model. Specifically, input data is fed into a generative AI model such as GPT or BERT, and the degree of match and probability are calculated to determine whether the data was generated by the generative AI. For example, for the data "Is this clothing item trending on social media?", the probability that it was generated by the generative AI is determined to be 85%.

[1225] Step 5:

[1226] The server's emotion engine recognizes the user's emotion based on the analysis results. Text analysis technology is used to classify emotions from the target data (text). For example, the emotion engine will recognize the sentence "Is this clothing popular on social media?" as "neutral."

[1227] Step 6:

[1228] The server formats the generated AI's judgment results and emotion recognition results into JSON format and sends them to the smart glasses, which act as a receiving device. The formatted data includes the generated AI's probability (85%) and emotion classification (neutral).

[1229] Step 7:

[1230] The smart glasses then present the analysis results and emotion recognition results to the user. The display shows "Probability of message generated by AI: 85%, Emotion: Neutral," and the customer service staff can use this information to respond appropriately.

[1231] 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.

[1232] 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.

[1233] 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.

[1234] [Fourth embodiment]

[1235] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1236] 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.

[1237] 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).

[1238] 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.

[1239] 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.

[1240] 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).

[1241] 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.

[1242] 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.

[1243] 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.

[1244] 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.

[1245] 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.

[1246] 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.

[1247] 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."

[1248] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. Specific examples of the present invention are described in detail below.

[1249] System Configuration

[1250] This system consists of the following main components:

[1251] 1. Input Device (Terminal)

[1252] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[1253] 2. Server

[1254] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[1255] 3. Analysis Engine

[1256] It is software or an application located on a server that compares target data with multiple generative AI models and calculates the degree of match. The analytical engine uses natural language processing technology and machine learning algorithms.

[1257] 4. Receiving device (terminal)

[1258] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[1259] Program processing overview

[1260] 1. Enter and submit data

[1261] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[1262] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[1263] 2. Data Receipt and Analysis

[1264] The server receives data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[1265] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[1266] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[1267] 3. Sending and displaying analysis results

[1268] The server formats the analysis results in JSON format or similar and sends them to the receiving device.

[1269] The device visually presents the analysis results to the user, for example, displaying a message such as "There is an 85% chance that this sentence was created by a generative AI."

[1270] Specific examples

[1271] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[1272] 1. Enter and submit data

[1273] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[1274] The user clicks the "Start Analysis" button and the text is sent to the server.

[1275] 2. Data Receipt and Analysis

[1276] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[1277] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[1278] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[1279] 3. Sending and displaying analysis results

[1280] The server formats the "75% match" result in JSON format and sends it to the receiving device.

[1281] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI."

[1282] In this way, the invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing inappropriate use of generative AI.

[1283] The processing flow will be explained below.

[1284] Detailed processing steps of the program

[1285] Step 1:

[1286] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[1287] Step 2:

[1288] The terminal converts the input digital data into JSON format, which makes it easier for the server to parse.

[1289] Step 3:

[1290] The device sends the converted JSON data as an HTTP POST request to the server's API endpoint.

[1291] Step 4:

[1292] The server receives the POST request, converts the received data from JSON back to an internal parsing format, and temporarily stores it in a database.

[1293] Step 5:

[1294] The server adds new analysis jobs to the queue, which contain the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[1295] Step 6:

[1296] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[1297] Step 7:

[1298] The analytics engine compares the incoming data with multiple generative AI models, specifically analyzing the following:

[1299] Contextual consistency

[1300] Grammatical accuracy

[1301] Vocabulary frequency and patterns

[1302] Prevalence of similar text or images

[1303] Step 8:

[1304] The server obtains individual probability values ​​from each generative AI model and combines them to calculate the final match rate. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, they will be combined to calculate a match rate of 75%.

[1305] Step 9:

[1306] The server converts the analysis results, including the final degree of match, into JSON format and sends it to the terminal as an HTTP response.

[1307] Step 10:

[1308] The device then analyzes the analysis results and displays them in a format that is easy for the user to understand. For example, the device might display something like, "There is a 75% chance that this sentence was created by a generative AI."

[1309] Step 11:

[1310] The user checks the analysis results and takes the next action if necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[1311] Example 1

[1312] 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."

[1313] In recent years, digital data created by generative AI has become widely used in educational institutions and companies, raising concerns about its misuse and unauthorized use. The present invention aims to solve these problems and promote the proper use of generative AI by providing a system that can accurately determine whether digital data was created by generative AI.

[1314] 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.

[1315] In this invention, the server includes a means for transmitting target digital data to an analytical engine, a means for the analytical engine to compare and analyze the target data with multiple generative AI models, and a means for calculating a final degree of match by combining the degrees of match obtained from each generative AI model, thereby making it possible to determine with high accuracy whether digital data was created by a generative AI.

[1316] "Target digital data" refers to electronic text or image data acquired from an input device as the target for analysis.

[1317] An "input device" is a device such as a computer, smartphone, or tablet that allows a user to input digital data.

[1318] "Analysis Engine" means a program or software that compares and analyzes acquired digital data with a generated AI model.

[1319] A "generative AI model" is an artificial intelligence model trained to generate text and images using natural language processing and machine learning algorithms.

[1320] "Matchability" is a numerical value that indicates the likelihood that the target digital data was created by a generative AI model, and is calculated as an analysis result.

[1321] The "receiving device" is a device that receives the analysis results sent from the server and presents them to the user.

[1322] The "percentage display format" is a method of expressing the degree of agreement of the analysis results as a percentage.

[1323] The present invention relates to a system that analyzes target digital data and determines whether it was created by generative AI. This system can be used in educational institutions, companies, and other organizations to monitor the use of generative AI and take appropriate measures. A specific embodiment of this system is described in detail below.

[1324] System Configuration

[1325] This system consists of the following main components:

[1326] 1. Input Device (Terminal)

[1327] A device that allows a user to input digital data (text or images). This device includes personal computers, smartphones, tablets, etc.

[1328] 2. Server

[1329] This is a computer system that receives data sent from input devices and analyzes it using generative AI. The server has a built-in analysis engine and is responsible for comparing and analyzing the data with multiple generative AI models.

[1330] 3. Analysis Engine

[1331] It is a software or application located on a server that compares target data with multiple generative AI models and calculates the degree of agreement. The analysis engine uses natural language processing technology and machine learning algorithms, specifically models such as GPT, BERT, and T5.

[1332] 4. Receiving device (terminal)

[1333] A device that receives analysis results sent from the server and presents them to the user. It is often the same as the input device, but it can also be different.

[1334] Entering and Submitting Data

[1335] The user uses an input device to input the digital data to be analyzed (e.g., text or images) into the input form. Specifically, the user types text using a keyboard or selects an image from a file selection dialog.

[1336] The user clicks the "Start Analysis" button in the input form. This action sends the input data to the server. Specifically, an HTTP request is generated and a payload containing the data is sent to the server.

[1337] Receiving and analyzing data

[1338] The server receives the data sent by the user. Specifically, the server's web server captures the HTTP request and retrieves the body data.

[1339] The server converts the received data into a standard format such as JSON and formats it for analysis. Specifically, a data parser analyzes the raw data, extracts the necessary information, and formats it.

[1340] The server's analysis engine compares and analyzes the received data with multiple generative AI models (e.g., GPT, BERT, T5). Specifically, the server calls the analysis engine, inputs the data into the generative AI model, and analyzes the context, grammar, and vocabulary usage patterns.

[1341] The server then combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm to combine the match scores from different models and calculates the final score using methods such as a weighted average.

[1342] Sending and displaying analysis results

[1343] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results and serializes them into a JSON-formatted string.

[1344] The server then sends the formatted analysis results to the receiving device. Specifically, it generates an HTTP response and sends it to the device.

[1345] The device visually presents the analysis results to the user. Specifically, the device's web application parses the JSON data and displays a message saying, "There is an 85% chance that this sentence was created by generative AI."

[1346] Specific examples

[1347] For example, consider a case where a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[1348] Prompt Sentence Examples

[1349] Below are some example prompts to input to the generative AI model:

[1350] "Please help me write the report that's due today."

[1351] "Write a data analysis explanation for a school project."

[1352] "Please translate the following text into an English fan fiction story."

[1353] This prompt statement can be used to check the system's operation and to demonstrate it.

[1354] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1355] Step 1:

[1356] A user inputs digital data using an input device. The user enters text or images to be analyzed into an input form. In this process, the user types text content using a keyboard or uploads an image file using a file selection dialog. The input data is temporarily stored in local storage on the input device. Input = digital data to be analyzed, Output = raw data stored on the input device.

[1357] Step 2:

[1358] The user clicks the "Start Analysis" button, which sends the input data to the server. Specifically, the input device generates an HTTP POST request, which includes the input data as the payload. Input = raw data stored on the input device, Output = data payload sent to the server.

[1359] Step 3:

[1360] The server receives data sent from the input device. The server's web server captures the HTTP request and obtains the payload data. Input = data payload sent from the input device, Output = formatted data to pass to the analysis engine.

[1361] Step 4:

[1362] The server converts the data it receives into a standard format such as JSON and formats it for analysis. Specifically, the server's data parser analyzes the raw data, extracts the necessary information, and serializes it. Input = data payload received by the server, output = formatted data in JSON format.

[1363] Step 5:

[1364] The server's analysis engine compares and analyzes the formatted data with multiple generative AI models. Specifically, the server calls the analysis engine, inputs the formatted data into generative AI models such as GPT, BERT, and T5, and analyzes the context, grammar, and vocabulary usage patterns. Input = formatted data in JSON format, output = match score for each generative AI model.

[1365] Step 6:

[1366] The server combines the match scores from each generative AI model to calculate the final match score. Specifically, the server uses an algorithm (e.g., weighted average) to combine the match scores from different models to calculate the overall score. Input = match score for each generative AI model, output = final match score.

[1367] Step 7:

[1368] The server formats the calculated analysis results in JSON format or similar and sends them to the receiving device. Specifically, the server creates a dictionary structure containing the analysis results, serializes it into a JSON-formatted string, and sends it as an HTTP response. Input = final match score, output = JSON-formatted analysis results sent to the receiving device.

[1369] Step 8:

[1370] The device visually presents the analysis results received from the server to the user. Specifically, the device's web application parses the JSON data and displays the message, "There is an 85% chance that this sentence was created by generative AI." Input = JSON-formatted analysis results received from the server, Output = visual analysis results presented to the user.

[1371] Specific examples

[1372] Consider a case in which a user is a student at an educational institution and is submitting an English composition assignment. The user enters the sentence, "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device and clicks the "Start Analysis" button. The sentence is then sent to the server. The server formats the text data received from the input device for analysis and passes it to the analysis engine. The server's analysis engine compares the sentence with generative AI models such as GPT, BERT, and T5, analyzing the context and grammar. The server then combines the match scores obtained from each generative AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) to calculate a final match score (e.g., 75%). The server formats the "75% match" result in JSON format and sends it to the receiving device. The device then displays a message to the user saying, "There is a 75% chance that this sentence was created by a generative AI." In this way, the present invention reduces the burden on educational institutions and provides a powerful tool for monitoring and preventing the inappropriate use of generative AI.

[1373] Prompt Sentence Examples

[1374] Below are some example prompts to input to the generative AI model:

[1375] "Please help me write the report that's due today."

[1376] "Write a data analysis explanation for a school project."

[1377] "Please translate the following text into an English fan fiction story."

[1378] This prompt statement can be used to check the system's operation and to demonstrate it.

[1379] (Application example 1)

[1380] 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."

[1381] In recent years, the generation of digital data using generative AI has rapidly spread, resulting in a need to determine whether it was created by generative AI. As this technology becomes more widespread, the risk of unauthorized use of text and images created by generative AI also increases. It is particularly important for educational institutions and companies to properly monitor the use of generative AI and take necessary measures. However, current systems face the challenge of making this determination efficiently and accurately.

[1382] 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.

[1383] In this invention, the server includes means for transmitting digital data input from a user terminal to an analysis server, means for comparing and analyzing the target data using multiple generative AI models, and means for returning the analysis results in percentage format to the terminal, thereby enabling accurate real-time determination of whether the digital data was created by a generative AI.

[1384] A "terminal" is a device through which a user inputs digital data, and specifically refers to a personal computer, smartphone, tablet, etc.

[1385] An "analysis server" is a computer system that receives digital data sent from a terminal and analyzes the data using a generative AI model.

[1386] A "generative AI model" is an algorithm or model that uses artificial intelligence technology to generate digital data, and examples include GPT, BERT, and T5.

[1387] "Comparison and analysis" is the process in which the analysis server uses a generative AI model to compare input data with known data sets and analyze the data's features and patterns.

[1388] The "percentage display format" is a display format in which the analysis results are quantified and the degree of agreement or reliability is expressed as a percentage.

[1389] "Visual presentation means" refers to an interface or method for displaying analysis results in a way that is easy for users to understand, such as using graphs, text, charts, etc.

[1390] "Synthesis" refers to the process of integrating the results obtained from multiple generative AI models to derive the final analysis results.

[1391] In order to implement the present invention, the following system and means are utilized: Each component of the system is designed based on the claims.

[1392] System Configuration

[1393] This system consists of the following main components:

[1394] 1. Device:

[1395] A terminal is a device through which a user inputs digital data. Specifically, a personal computer, smartphone, tablet, etc. is used. The user inputs digital data into the terminal and sends it to an analysis server.

[1396] 2. Analysis server:

[1397] The analysis server is a computer system that receives digital data sent from the terminal and analyzes the data using a generative AI model. The server has generative AI models such as GPT, BERT, and T5 installed, and processes the data for analysis.

[1398] 3. Generative AI Model:

[1399] A generative AI model is an algorithm or model that uses generative AI to generate digital data. Examples include GPT, BERT, and T5. This model analyzes input digital data by comparing it with existing datasets.

[1400] 4. Data Receipt and Analysis:

[1401] The server receives the digital data sent from the device and converts it into a standard format such as JSON. The converted data is formatted for analysis and compared and analyzed using a generative AI model. The analysis results are calculated in the form of a confidence level (%).

[1402] 5. Presentation of results:

[1403] The analysis results are returned to the terminal in the form of a percentage. The terminal provides an interface to visually present the analysis results to the user, allowing the user to easily understand whether the digital data was created by generative AI.

[1404] Process flow:

[1405] As a concrete example, consider the analysis of the sentence "He finished writing his thesis."

[1406] 1. Data entry and transmission via terminal:

[1407] The user enters the sentence "He finished writing his thesis" into the input form on the terminal. By pressing the send button, this data is sent to the analysis server.

[1408] 2. Data reception and analysis by the server:

[1409] The analysis server formats the received digital data for analysis and performs the analysis using a generative AI model, such as GPT, BERT, or T5, to analyze the grammar and structure of the sentence.

[1410] 3. Sending and displaying analysis results:

[1411] The server calculates the analysis results and returns them to the device in the form of a percentage, which then displays a message to the user, such as "There is an 85% chance that this sentence was created by a generative AI."

[1412] The system uses Python hoformat to send and receive data between the server and the device. Specific analysis is performed based on a generative AI model, and the reliability of the analysis results is calculated.

[1413] The above process and components used enable the present invention to accurately determine in real time whether digital data entered by a user was created by generative AI.

[1414] Example prompt sentence:

[1415] "He finished writing his thesis."

[1416] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1417] Step 1:

[1418] A user uses a terminal to input digital data. Specifically, the user enters text data such as "He has finished writing his thesis" into the terminal's input form and clicks the "Submit" button. This input data is later formatted into JSON format to be sent to an analysis server. At this stage, the input is the text data entered by the user, and the output is JSON format data.

[1419] Step 2:

[1420] The terminal sends the digital data entered by the user to the analysis server. The terminal uses an HTTP POST request to send JSON-formatted data to the server. This request allows the data sent from the terminal to reach the analysis server, allowing the server to begin processing the received data. The input is JSON-formatted text data entered by the user, and the output is an HTTP request to the server.

[1421] Step 3:

[1422] The server receives the digital data sent from the device and formats it for analysis. Specifically, it converts the received JSON data into an appropriate data format that can be analyzed by the generative AI model. This conversion process includes a processing step that converts JSON into an internal data structure that can be processed by the analysis engine. The input is JSON-formatted data, and the output is internal data for the analysis engine.

[1423] Step 4:

[1424] The server's analysis engine analyzes the data using generative AI models. This analysis process uses multiple generative AI models, such as GPT, BERT, and T5, to compare the input digital data with existing datasets. Specifically, it performs grammar checks, contextual analysis, and the detection of specific lexical patterns. The input for this step is the internal data for the analysis engine, and the output is a confidence score analyzed by each generative AI model.

[1425] Step 5:

[1426] The server aggregates the analysis results from each generative AI model and calculates the final analysis result. Specifically, it combines the confidence scores obtained from each model to calculate the final match and confidence. In this step, the analysis results are consolidated using average and maximum values ​​based on context and patterns. The input is the confidence scores from each generative AI model, and the output is the overall analysis result.

[1427] Step 6:

[1428] The server formats the final analysis results in JSON format and sends them to the terminal. Specifically, the analysis results generated internally on the server are converted into JSON data format and sent back to the terminal as an HTTP response. The input is the comprehensive analysis results, and the output is JSON format data.

[1429] Step 7:

[1430] The device visually presents the analysis results received from the server to the user. For example, a message such as "There is an 85% chance that this sentence was created by a generative AI" is displayed on the screen. The input is the analysis results in JSON format received from the server, and the output is a visual display for the user.

[1431] These steps allow the system to determine in real time whether digital data was created by generative AI.

[1432] 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.

[1433] The present invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed. Specific examples of the present invention are described in detail below.

[1434] System Configuration

[1435] This system consists of the following main components:

[1436] 1. Input Device (Terminal)

[1437] A device that allows users to input digital data (text or images). Examples of such devices include PCs, smartphones, and tablets.

[1438] 2. Server

[1439] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[1440] 3. Analysis Engine

[1441] This software or application resides on a server and compares target data with multiple generative AI models to determine the results of generative AI creations.

[1442] 4. Emotion Engine

[1443] Software or application located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[1444] 5. Receiving Device (Terminal)

[1445] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user.

[1446] Program processing overview

[1447] 1. Enter and submit data

[1448] The user enters text and images into a form on an input device, a process that can be carried out through a typical text editor or digital image upload interface.

[1449] The user clicks the "Start Analysis" button on the data input device, and the data is sent to the server.

[1450] 2. Data Receipt and Analysis

[1451] The server receives the data sent from the input device, converts it into a standard format such as JSON, and formats it for analysis.

[1452] The server-based analytics engine compares the incoming data with generative AI models, such as GPT, BERT, and T5.

[1453] The server then combines the scores obtained from each generative AI model to calculate the final score, which is calculated by analyzing context, grammar, and vocabulary usage patterns.

[1454] 3. Emotion Recognition Processing

[1455] The emotion engine in the server recognizes the user's emotions based on the analysis results. For text, it classifies emotions as negative, positive, neutral, etc. For images, it uses facial expression recognition technology to analyze emotions.

[1456] The server adds the emotion recognition results to the analysis results to create comprehensive feedback.

[1457] 4. Transmission and display of analysis results and emotion recognition results

[1458] The server formats the analysis results and emotion recognition results in JSON format or similar and sends them to the receiving device.

[1459] The device displays the received results in a user-friendly format, such as a message like, "There is an 85% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[1460] Specific examples

[1461] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[1462] 1. Enter and submit data

[1463] The user enters the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." into a form on an input device.

[1464] The user clicks the "Start Analysis" button and the text is sent to the server.

[1465] 2. Data Receipt and Analysis

[1466] The server formats the text data received from the input device for analysis and passes it to the analysis engine.

[1467] The analysis engine on the server compares sentences with generative AI models such as GPT, BERT, and T5 to analyze context and grammar.

[1468] The server combines the match scores obtained from each generated AI (e.g., 70% for GPT, 75% for BERT, and 80% for T5) and calculates the final match score (e.g., 75%).

[1469] 3. Emotion Recognition Processing

[1470] The emotion engine in the server recognizes the user's emotion as positive based on the content of the text.

[1471] 4. Transmission and display of analysis results and emotion recognition results

[1472] The server formats the result of "75% match" and the recognition result of "emotion is positive" in JSON format and sends them to the receiving device.

[1473] The device will display to the user, "There is a 75% chance that this sentence was created by a generative AI. Also, the user's sentiment is positive."

[1474] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[1475] The processing flow will be explained below.

[1476] Specific processing steps of the invention that combines the emotion engine

[1477] Step 1:

[1478] The user inputs text or images into an input device, i.e., digital data into an input form on a PC, smartphone, or tablet, and clicks the "Start Analysis" button.

[1479] Step 2:

[1480] The terminal converts the input digital data into JSON format, which is then sent to the server.

[1481] Step 3:

[1482] The server receives the JSON data, which is then formatted for parsing and temporarily stored in a database.

[1483] Step 4:

[1484] The server adds a new analysis job to the queue, which contains the digital data to be analyzed along with the necessary information to pass to the analysis engine.

[1485] Step 5:

[1486] The server sequentially takes analysis jobs from the queue and passes them to the analysis engine, which loads multiple generative AI models (e.g., GPT, BERT, T5, etc.).

[1487] Step 6:

[1488] The analytics engine compares incoming data with multiple generative AI models, analyzing:

[1489] Contextual consistency

[1490] Grammatical accuracy

[1491] Vocabulary frequency and patterns

[1492] Prevalence of similar text or images

[1493] Step 7:

[1494] The server obtains the individual probability values ​​from each generative AI model. For example, if the GPT model gives a match rate of 60%, the BERT model gives a match rate of 70%, and the T5 model gives a match rate of 80%, then the server combines these to calculate a match rate of 75%.

[1495] Step 8:

[1496] The server passes the results of the match analysis to the emotion engine, which then recognizes the user's emotion from the content of the received data. For text, it classifies emotions as negative, positive, neutral, etc., and for images, it uses facial expression recognition technology to analyze the emotion.

[1497] Step 9:

[1498] The server adds the emotion recognition results to the analysis results and generates comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[1499] Step 10:

[1500] The server converts the analysis results and emotion recognition results into JSON format and sends them to the receiving device as an HTTP response.

[1501] Step 11:

[1502] The device analyzes the received analysis results and emotion recognition results and displays them in a format that is easy for the user to understand. For example, a message such as "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive" may be displayed on the screen.

[1503] Step 12:

[1504] The user checks the analysis results and emotion recognition results and takes the following actions as necessary, such as correcting the sentence, re-entering it, or checking with the teacher.

[1505] Example 2

[1506] 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."

[1507] In recent years, technological advances in generative AI have made it easier to automatically generate text and images, increasing the risk of this being misused by educational institutions and businesses. Furthermore, there is a demand for feedback that takes into account the user's emotions, which has been difficult with conventional systems. Therefore, a system is needed that can simultaneously identify digital data automatically generated by generative AI, recognize the user's emotions, and provide feedback that reflects those emotions.

[1508] The identification process by the identification processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means. In this invention, the server includes means for acquiring target digital data from an information processing device, means for transmitting the acquired data to an analysis device, means for the analysis device to compare and analyze the target data with multiple generative AI models, means for recognizing the user's emotions based on the analysis results, and means for returning the analysis results and emotion recognition results in a percentage format to the receiving device. This enables the generation AI to identify automatically generated data and provide feedback that reflects the user's emotions.

[1509] An "information processing device" is a device that allows a user to input digital data, and is a wide range of terminals including PCs, smartphones, tablets, etc.

[1510] The "analysis device" is a device for analyzing acquired digital data, and is a module within the server that uses multiple generative AI models to compare and analyze data.

[1511] A "generative AI model" is an algorithm that automatically generates digital data using generative AI technology, and specifically includes GPT, BERT, T5, etc.

[1512] "Digital data" refers to data provided by a user via an input device, and includes text data, image data, and the like.

[1513] "Means for recognizing emotions" refers to software or algorithms for classifying and recognizing a user's emotions based on the analysis results, including text analysis and facial expression recognition technologies.

[1514] A "receiving device" is a device that receives the analysis results and emotion recognition results and presents them to the user, and includes smartphones, PCs, tablets, etc.

[1515] The "means for transmitting data" refers to a protocol or interface for transmitting digital data from the information processing device to the analysis device.

[1516] This invention relates to a system that combines the ability to determine whether target digital data was created by generative AI with the ability to recognize user emotions. This system can be used in educational settings and businesses to monitor the use of generative AI and take appropriate measures. Furthermore, by providing the results of user emotion recognition, more advanced feedback and analysis can be performed.

[1517] This system consists of the following main components:

[1518] 1. Input devices (information processing devices)

[1519] A device that allows users to input digital data (text and images). Examples of this device include PCs, smartphones, and tablets.

[1520] 2. Server (analysis device)

[1521] This is a computer system that receives and analyzes data sent from input devices. The server has a built-in analysis engine and emotion engine, and is responsible for analyzing the data and recognizing emotions.

[1522] 3. Analysis Engine

[1523] It is software or a program located on a server that compares target data with multiple generative AI models to determine the results of generative AI creations. Specifically, models such as GPT, BERT, and T5 are used.

[1524] 4. Emotion Engine

[1525] Software or a program located in a server that analyzes target data or additional user data (e.g., text or images) to recognize user emotions. Emotion engines use text analysis and image analysis techniques.

[1526] 5. Receiving Device

[1527] This is a device that receives analysis results and emotion recognition results sent from the server and presents them to the user.

[1528] Specific processing of the program

[1529] Digital data entered by a user through an information processing device is first sent to a server. The server's analysis engine compares the received data with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. Next, an emotion engine recognizes the user's emotions. The analysis results and emotion recognition results are integrated and sent to the receiving device, where they are finally presented to the user.

[1530] Specific examples

[1531] For example, consider a case where a user is a student at an educational institution and is submitting an English writing assignment.

[1532] The user enters the following sentence into the form on their PC: "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." When the user clicks the "Start Analysis" button, the sentence is sent to the server.

[1533] The server formats the text data and passes it to the analysis engine. The analysis engine compares the text with generative AI models such as GPT, BERT, and T5 to calculate the degree of match. For example, if GPT reports a match of 70%, BERT reports a match of 75%, and T5 reports a match of 80%, these will be combined to arrive at a final match of 75%.

[1534] Next, the emotion engine on the server recognizes the user's emotion as positive based on the content of the sentence. This integrates the analysis results with the emotion recognition results, and displays on the receiving device the message, "There is a 75% chance that this sentence was created by generative AI. The user's emotion is positive."

[1535] This invention makes it possible to efficiently monitor and prevent the use of generative AI, and by recognizing the user's emotions, it is possible to provide more detailed feedback, which is expected to be used in educational settings and businesses.

[1536] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1537] Step 1:

[1538] Users input text or images into a form on an input device such as a PC, smartphone, or tablet. Specifically, they enter text into a text area or upload an image file. This input data (e.g., the text "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." or an image file) is converted to JSON format on the device.

[1539] Input: Text or image data entered by a user into an input device

[1540] Output: Digital data formatted in JSON format

[1541] Step 2:

[1542] The user clicks the "Start Analysis" button on the input device. This operation sends the digital data to the server. When the send button is clicked, the device starts communication to send the data to the server.

[1543] Input: Digital data formatted in JSON format

[1544] Output: Server receives data

[1545] Step 3:

[1546] The server receives the digital data sent from the input device, where it is reconfirmed in JSON format and prepared to be passed to the analysis engine.

[1547] Input: JSON formatted data sent from the input device

[1548] Output: Formatted data to be passed to the analysis engine

[1549] Step 4:

[1550] The server's analytics engine analyzes the incoming digital data. This engine then invokes generative AI models such as GPT, BERT, and T5 to match the data. A match is calculated for each model, comparing context, grammar, and vocabulary usage patterns.

[1551] Input: Formatted digital data

[1552] Output: Match score by each generative AI model (e.g., 70% for GPT, 75% for BERT, 80% for T5)

[1553] Step 5:

[1554] The server then combines the match scores from each generative AI model to calculate the overall match score. For example, the match scores (70% for GPT, 75% for BERT, and 80% for T5) are averaged to calculate the final match score (e.g., 75%).

[1555] Input: Match scores from each generative AI model

[1556] Output: Overall match (e.g. 75%)

[1557] Step 6:

[1558] The emotion engine in the server recognizes the user's emotion based on the analysis results. For text, it classifies it as negative, positive, or neutral. For example, the sentence "My summer vacation was wonderful. We visited many places and enjoyed a lot of local foods." would be classified as positive. For images, emotion is analyzed using facial expression recognition technology.

[1559] Input: Formatted data and analysis results

[1560] Output: User's emotion category (e.g., positive)

[1561] Step 7:

[1562] The server combines the emotion recognition results with the analysis results to create comprehensive feedback, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's emotion is positive."

[1563] Input: Overall agreement and emotion category

[1564] Output: General feedback message

[1565] Step 8:

[1566] The server formats the aggregate feedback message in JSON format and sends it to the receiving device.

[1567] Input: General feedback message

[1568] Output: Feedback data in JSON format

[1569] Step 9:

[1570] The device displays the feedback data received from the server to the user, such as a message like, "There is a 75% chance that this sentence was created by a generative AI. The user's sentiment is positive."

[1571] Input: Feedback data in JSON format

[1572] Output: Feedback displayed in a format that the user can understand

[1573] (Application example 2)

[1574] 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."

[1575] Currently, it is difficult for brick-and-mortar stores to determine the authenticity and sentiment of customers' text messages in real time. This makes it difficult for customer service staff to respond appropriately, potentially hindering the improvement of the customer experience. It is also challenging to respond appropriately to false information and messages created by generative AI. These issues need to be resolved.

[1576] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1577] In this invention, the server includes a means for acquiring target digital data from an input device, a means for transmitting the acquired data to an analysis device, a means for the analysis device to compare and analyze the target data with multiple generative AI models, and a means for analyzing emotions and displaying the recognition results on a receiving device. This allows customer service staff to confirm the authenticity of customer messages and simultaneously recognize customer emotions in real time. This enables more appropriate and prompt customer service, and is expected to improve the customer experience.

[1578] "Target digital data" refers to any form of digital data that is the subject of analysis, including text, images, etc.

[1579] An "input device" is a device that allows a user to input digital data, and includes PCs, smartphones, tablets, smart glasses, etc.

[1580] An "analysis device" is a computer system that acquires target digital data and analyzes it using an analysis engine and an emotion recognition engine.

[1581] A "generative AI model" is a model generated using artificial intelligence, such as a natural language processing model such as GPT or BERT.

[1582] The "receiving device" is a device that receives the analysis results and emotion recognition results sent from the analyzing device and presents them to the user.

[1583] The "means for analyzing emotions and displaying the recognition results on the receiving device" is a means having the function of analyzing the user's emotions from the target digital data and displaying the results on the receiving device in real time.

[1584] This invention relates to a system that determines whether target digital data was created by generative AI and simultaneously recognizes the user's emotions. This system is intended to be used primarily in customer service situations in brick-and-mortar stores.

[1585] System Configuration

[1586] The system includes the following major components:

[1587] 1. Input Devices

[1588] A device that allows customers to input digital data (text and images). This includes wearable devices such as smart glasses.

[1589] 2. Analysis device

[1590] This is a computer system that receives and analyzes data sent from an input device. This device has a built-in analysis engine and emotion engine.

[1591] 3. Analysis Engine

[1592] This software is located on a server and compares the target data with multiple generative AI models to determine the results of generative AI creations. GPT and BERT are used as generative AI models.

[1593] 4. Emotion Engine

[1594] This software is located on the server and analyzes target data to recognize user emotions. It uses text analysis and image analysis techniques.

[1595] 5. Receiving Device

[1596] This device receives the analysis results and emotion recognition results sent from the server and presents them to the user. Smart glasses are used as this receiving device.

[1597] Hardware and software used

[1598] Hardware

[1599] Smart glasses (e.g., Google Glass, Vuzix Blade)

[1600] Server (e.g. AWS, Google Cloud)

[1601] software

[1602] Web framework: Flask

[1603] Generative AI decision engine: AIAnalyzer (library of hypotheses)

[1604] Emotion recognition engine: SentimentAnalyzer (hypothetical library)

[1605] Data processing and calculation methods

[1606] The analysis device (server) performs the following processes.

[1607] 1. Receiving and formatting data

[1608] The data sent from the input device is received by the server, where it is converted to JSON format or similar and formatted for analysis.

[1609] 2. Discrimination by Generative AI

[1610] Using a generative AI model (e.g., GPT, BERT, etc.), the probability that the target data is generated by generative AI is analyzed. The analysis result is expressed in percentage form.

[1611] 3. Emotion recognition

[1612] The emotion engine analyzes the user's emotions from the target data and classifies them as positive, negative, neutral, etc.

[1613] 4. Sending the results

[1614] The analysis results and emotion recognition results are formatted and transmitted to a receiving device.

[1615] Specific examples

[1616] Let's take the example of a customer in a brick-and-mortar clothing store who sends a message asking, "Is this clothing popular on social media?" This message is sent to the server as a prompt as follows:

[1617] Prompt Sentence Examples

[1618] Is this outfit creating a buzz on social media?

[1619] The analysis results are displayed as follows: "Probability of message generated by AI: 85%, Emotion: Neutral." Based on this information, customer service staff can quickly respond appropriately.

[1620] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1621] Step 1:

[1622] A user inputs target digital data such as text or images into an input device such as smart glasses. Specifically, the user inputs a message such as "Is this outfit trending on social media?" using voice or text input. This input data is sent to the smart glasses.

[1623] Step 2:

[1624] The smart glasses send the input digital data to the server, where it is converted into JSON format and passed to an analysis device. If the input sentence is something like "Is this clothing popular on social media?", it will be sent as JSON data.

[1625] Step 3:

[1626] The analysis device receives the data received by the server and formats it. The formatted data is converted into a format that can be analyzed by a generative AI model (GPT, BERT, etc.). For example, text data is formatted according to a standard format.

[1627] Step 4:

[1628] The server's analysis engine compares and analyzes the formatted data with the generative AI model. Specifically, input data is fed into a generative AI model such as GPT or BERT, and the degree of match and probability are calculated to determine whether the data was generated by the generative AI. For example, for the data "Is this clothing item trending on social media?", the probability that it was generated by the generative AI is determined to be 85%.

[1629] Step 5:

[1630] The server's emotion engine recognizes the user's emotion based on the analysis results. Text analysis technology is used to classify emotions from the target data (text). For example, the emotion engine will recognize the sentence "Is this clothing popular on social media?" as "neutral."

[1631] Step 6:

[1632] The server formats the generated AI's judgment results and emotion recognition results into JSON format and sends them to the smart glasses, which act as a receiving device. The formatted data includes the generated AI's probability (85%) and emotion classification (neutral).

[1633] Step 7:

[1634] The smart glasses then present the analysis results and emotion recognition results to the user. The display shows "Probability of message generated by AI: 85%, Emotion: Neutral," and the customer service staff can use this information to respond appropriately.

[1635] 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.

[1636] 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.

[1637] 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.

[1638] 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.

[1639] 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.

[1640] 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.

[1641] 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).

[1642] 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, motorcycles, and other devices, 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.

[1643] 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."

[1644] 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.

[1645] 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).

[1646] 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.

[1647] 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.

[1648] 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.

[1649] 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.

[1650] 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.

[1651] 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.

[1652] 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.

[1653] 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.

[1654] 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.

[1655] 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.

[1656] The following is further disclosed regarding the above embodiment.

[1657] (Claim 1)

[1658] means for acquiring digital data of interest from an input device;

[1659] means for transmitting the acquired data to an analysis engine;

[1660] A means for the analysis engine to compare and analyze multiple generative AI models and target data;

[1661] means for transmitting the analysis results back to the receiving device in a percentage format;

[1662] means for presenting the analysis results on the receiving device;

[1663] A system including:

[1664] (Claim 2)

[1665] The system of claim 1, wherein the target digital data is a text.

[1666] (Claim 3)

[1667] 2. The system of claim 1, wherein the digital data of the object is an image.

[1668] "Example 1"

[1669] (Claim 1)

[1670] means for acquiring digital data of interest from an input device;

[1671] means for transmitting the acquired data to an analysis engine;

[1672] A means for the analysis engine to compare and analyze multiple generative AI models and target data;

[1673] A means for calculating a final matching score by combining the matching scores obtained from each generative AI model;

[1674] means for transmitting the analysis results back to the receiving device in a percentage format;

[1675] means for presenting the analysis results on the receiving device;

[1676] A system including:

[1677] (Claim 2)

[1678] The system of claim 1, wherein the target digital data is a text.

[1679] (Claim 3)

[1680] 2. The system of claim 1, wherein the digital data of the object is an image.

[1681] "Application Example 1"

[1682] (Claim 1)

[1683] a terminal for inputting target digital data;

[1684] means for transmitting the input data to an analysis server;

[1685] A means for the analysis server to compare and analyze the target data using multiple generative AI models;

[1686] A means for returning the analysis results to the terminal in a percentage format;

[1687] a means for visually presenting the analysis results to the user on the device;

[1688] a means of sending input data to the generative AI model and synthesizing results from the model; and

[1689] A system including:

[1690] (Claim 2)

[1691] 2. The system of claim 1, wherein the target digital data is text data.

[1692] (Claim 3)

[1693] The system of claim 1, wherein the target digital data is visual data.

[1694] "Example 2: Combining Emotion Engines"

[1695] (Claim 1)

[1696] means for acquiring target digital data from an information processing device;

[1697] means for transmitting the acquired data to an analysis device;

[1698] A means for the analysis device to compare and analyze multiple generative AI models and target data;

[1699] means for recognizing a user's emotion based on the analysis result;

[1700] means for returning the analysis result and the emotion recognition result to the receiving device in a percentage format;

[1701] means for presenting the analysis result and the emotion recognition result in the receiving device;

[1702] A system including:

[1703] (Claim 2)

[1704] 2. The system of claim 1, wherein the target digital data is text data.

[1705] (Claim 3)

[1706] 2. The system of claim 1, wherein the digital data of interest is image data.

[1707] "Application example 2 when combining emotion engines"

[1708] (Claim 1)

[1709] means for acquiring target digital data from an input device;

[1710] means for transmitting the acquired data to an analysis device;

[1711] A means for the analysis device to compare and analyze multiple generative AI models and target data;

[1712] means for returning the analysis result to the receiving device in a percentage format;

[1713] means for presenting the analysis results on the receiving device;

[1714] means for analyzing emotions and displaying the recognition results on a receiving device;

[1715] A system including:

[1716] (Claim 2)

[1717] The system of claim 1, wherein the target digital data is a text.

[1718] (Claim 3)

[1719] 2. The system of claim 1, wherein the digital data of the object is an image. [Explanation of symbols]

[1720] 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. means for acquiring digital data of interest from an input device; means for transmitting the acquired data to an analysis engine; The analytical engine has a means to compare and analyze multiple generative AI models with the target data, means for transmitting the analysis results back to the receiving device in a percentage format; means for presenting the analysis results on the receiving device; A system including:

2. 2. The system of claim 1, wherein the target digital data is a text.

3. The system of claim 1 , wherein the digital data of the object is an image.

Images