system

Abstract

Provide a system. A method for generating a generative model includes: collecting data for training the generative model; means for pre-processing the collected data; means for training a generative model using the preprocessed data; means for generating a manzai script based on a theme or keywords provided by a user; a means for providing the generated comic script to a user; A system including:

Description

[Technical Field]

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

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

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

[0004] Generative AI technology is rapidly developing, but it is difficult for many average users to understand its performance differences and advantages. Furthermore, in the field of language generation in particular, it is necessary to appropriately handle the characteristics and humor of the Japanese language. Furthermore, there is a lack of concrete methods to effectively demonstrate the practicality and performance of generative AI. Therefore, a method is needed within Japan to widely and clearly demonstrate the capabilities of generative AI. [Means for solving the problem]

[0005] To solve this problem, the present invention provides the following means. A system is constructed that includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training a generative model using the preprocessed data, means for generating a manzai script based on themes and keywords provided by the user, and means for providing the generated manzai script to the user. This system enables the generative AI to provide users with concrete, practical deliverables while making the most of the characteristics of the Japanese language. Furthermore, the performance of the generative AI can be widely disseminated through the process and results.

[0006] A "generative model" is a machine learning model that generative AI is trained to generate new data or content for a specific task.

[0007] "Means of collecting data" refers to the methods and techniques used to obtain the data needed to train the generative model, including web scraping, using APIs, or extracting data from databases.

[0008] "Preprocessing means" refers to methods and techniques for converting collected data into a format that machine learning models can learn from effectively, including data cleansing, tokenization, and normalization.

[0009] "Theme and keywords" are elements that are specified by the user to indicate the content and style that the user expects from the manzai script to be generated.

[0010] A "manzai script" refers to the script or scenario of a comedy skit in the form of manzai, a dialogue-style piece of text intended to be performed by two or more comedians.

[0011] "Means of generation" refers to the technology or method by which the generation AI creates a new manzai script based on themes and keywords provided by the user.

[0012] "Means of delivery" refers to the technology or method for communicating the generated manzai script to users, including distribution via email, web application, or mobile app.

[0013] "Characteristics of the Japanese language" refers to the grammar, vocabulary, modes of expression, and cultural background that are unique to the Japanese language, and includes techniques and methods for dealing with these appropriately. [Brief explanation of the drawings]

[0014] [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

[0015] 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.

[0016] First, the terms used in the following description will be explained.

[0017] 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).

[0018] 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.

[0019] 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.

[0020] 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.

[0021] 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."

[0022] [First embodiment]

[0023] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0024] 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.

[0025] 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).

[0026] 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.

[0027] 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.

[0028] 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.

[0029] 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.

[0030] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0031] 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.

[0032] 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.

[0033] 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.

[0034] 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."

[0035] This invention provides a system that uses a generative AI to generate a manzai script and publicly demonstrates the AI's capabilities through the process and results. Specific embodiments are described below.

[0036] First, the server collects the data necessary to train the generative model. This can be done by web scraping, using APIs, or extracting data from existing databases. The collected data includes scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0037] The collected data is preprocessed by the server. Examples of preprocessing include text cleansing (removing noise data), tokenization (splitting into words or sentences), and normalization (conversion into a unified format), which prepares the data in a format suitable for training a generative model.

[0038] The server then trains a generative AI model based on the preprocessed data. Deep learning techniques are used for this purpose. The model is trained over multiple epochs to effectively capture the characteristics of the Japanese language and its humorous elements.

[0039] Users input the theme and keywords of the manzai using their device. Based on this input sent from the device to the server, the server generates a manzai script using the generated generative AI model. The generated script is a dialogue-style script that includes natural exchanges and humor in line with the specific theme.

[0040] Finally, the server sends the generated manzai script to the user's device. The user can practice the manzai based on the provided script and use it in actual performances. In addition, by sharing the process and results of practicing manzai, the user can demonstrate the performance of the generative AI to others.

[0041] For example, if a user inputs the keywords "pet topic" and "comedy element," the server will generate a manzai script based on this information, including funny pet-related and everyday anecdotes. The script is provided in an interactive format, including anecdotes such as "an incident that occurred while walking the dog" or "unexpected cat behavior." Users can use this script to practice manzai and aim to win the M-1 Grand Prix. Users can also share the capabilities of the generating AI by posting their practice process on social media or video platforms.

[0042] The processing flow will be explained below.

[0043] Step 1:

[0044] The server collects data to train the generative model, including scripts from past manzai competitions, literature on comedy, and Japanese dialogue data, through web scraping, API usage, and extraction from existing databases.

[0045] Step 2:

[0046] The server preprocesses the collected data, which includes text cleansing (removing noise data), tokenization (splitting into words and sentences), and normalization (converting into a unified format) to prepare the data in a format suitable for training a generative model.

[0047] Step 3:

[0048] The server uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0049] Step 4:

[0050] The user inputs the theme and keywords of the manzai using a terminal, which are then sent to the server.

[0051] Step 5:

[0052] The server uses the generated AI model to generate a manzai script based on themes and keywords provided by the user. This generation process creates natural dialogue that takes into account Japanese grammar and humor.

[0053] Step 6:

[0054] The generated manzai script is sent from the server to the user's device, where the user can view and check the received script.

[0055] Step 7:

[0056] The user practices a manzai based on the provided manzai script. If necessary, the user can modify the script content or input additional themes or keywords to generate the script again.

[0057] Step 8:

[0058] Ultimately, users can complete the manzai and use it in actual performances. In addition, by sharing the practice process and results on social media and video platforms, the performance and effectiveness of the generative AI can be widely promoted.

[0059] Example 1

[0060] 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."

[0061] Conventional text generation systems have had difficulty effectively incorporating the linguistic characteristics of Japanese and humorous elements when generating natural dialogue-style text based on specific themes or keywords. Furthermore, the generated text often fails to meet the user's intentions or expectations, resulting in scripts that lack entertainment value and practicality.

[0062] 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.

[0063] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a sentence generation script based on a theme or keywords provided by a user, and means for providing the generated sentence generation script to the user. This enables script generation in a natural, interactive format that takes into account the linguistic characteristics and humorous elements of the Japanese language.

[0064] A "generative model" refers to an algorithm that uses artificial intelligence techniques to generate output such as text or images based on specific input.

[0065] "Data collection" refers to the process of gathering the information needed to train a generative model from the internet or existing databases.

[0066] "Preprocessing" refers to the process of converting collected data into a format suitable for training a generative model by means of noise removal, tokenization, normalization, etc.

[0067] "User" refers to a person who uses the system to create a text generation script based on a specific theme or keyword.

[0068] "Themes and keywords" refer to important words and phrases that the user specifies to the system to determine the content and direction of the script to be generated.

[0069] A "text generation script" refers to a dialogue-style text created using a generative model based on a specific theme or keyword.

[0070] "Linguistic characteristics of Japanese" refers to characteristics that include the grammatical structure, expression methods, and cultural background that are unique to Japanese.

[0071] "Natural dialogue" refers to fluent, coherent conversation that is natural in human communication.

[0072] "Entertainment event performance data" refers to data on entertainment events that have been held in the past, and is the basic data used to generate high-quality scripts by training a generative model based on that data.

[0073] "Laughter literature" refers to research materials, books, papers, etc. on humor and comedy, and refers to the subjects used to train generative models.

[0074] MODE FOR CARRYING OUT THE INVENTION

[0075] This invention is a system that uses a generative AI model to create a sentence generation script based on a specific theme or keyword, and provides the results to the user.

[0076] Hardware and software used

[0077] The server collects data, preprocesses it, trains it, and generates scripts to train the generative model. Specifically, the following hardware and software are used:

[0078] Hardware: Computational servers with high-performance CPUs and GPUs, and large-capacity storage devices for storing data.

[0079] Software: The web scraping tool "BeautifulSoup" is used for data collection, the natural language processing library "MeCab" is used for data preprocessing, and the deep learning framework "TENSORFLOW (registered trademark)" or "PyTorch" is used for training the generative model.

[0080] Data collection

[0081] The server uses web scraping tools, APIs, and existing databases to collect the data needed to train the generative model. This data includes performance data from entertainment events and literature on comedy, such as scripts from past manzai competitions, books on comedy, and papers.

[0082] Data Preprocessing

[0083] The collected data is pre-processed by the server. Specifically, the following processes are performed:

[0084] Text cleansing: Remove noise data and unnecessary characters.

[0085] Tokenization: Use "MeCab" to divide the data into words and sentences.

[0086] Normalization: Transforming data into a uniform format and making it consistent.

[0087] Training generative AI models

[0088] Based on the preprocessed data, the server trains a generative AI model using the deep learning frameworks TensorFlow or PyTorch. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0089] User input

[0090] The user uses a device to input themes and keywords into the generative AI model. For example, keywords such as "pet topics" or "comedy elements" are input. This input information is then sent from the device to the server.

[0091] Creating and providing sentence generation scripts

[0092] The server uses a generative AI model to generate a text generation script based on themes and keywords submitted by the user. This script is a natural dialogue based on a specific theme, incorporating Japanese linguistic characteristics and humorous elements.

[0093] The generated script is provided to the user's device from the server. The user can practice generating sentences based on the provided script and use it in actual performances.

[0094] Prompt Sentence Examples

[0095] The following example is one of the prompts that a user can input to a generative AI model.

[0096] Create a text generation script about pets, with a comedic twist. For example, include a story about something that happened while walking a dog or an unexpected cat behavior.

[0097] The above is a specific embodiment for carrying out the present invention.

[0098] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0099] Step 1:

[0100] The server collects data to train the generative model. This data includes performance data on entertainment events that incorporate Japanese linguistic characteristics and humorous elements, as well as literature on laughter. Specifically, web scraping tools, APIs, and existing databases are used. Web scraping tools such as "BeautifulSoup" are used to collect comedy scripts from the Internet. Comedy-related literature is obtained through APIs, and the necessary data is extracted from existing databases. The collected data is saved in a format suitable for training the generative model.

[0101] Step 2:

[0102] The server pre-processes the collected data, which includes the following specific steps:

[0103] Cleansing: Remove noise data, specifically, remove unnecessary characters and advertisements.

[0104] Tokenization: We use the natural language processing library "MeCab" to divide the data into words and sentences.

[0105] Normalization: Transforming data into a uniform format and making it consistent.

[0106] The input is the collected raw data, and the output is pre-processed data suitable for training a generative model, which is then cleansed, tokenized, and stored in a normalized format.

[0107] Step 3:

[0108] The server uses the preprocessed data to train a generative AI model. Using the deep learning frameworks "TensorFlow" or "PyTorch," the model is trained over multiple epochs. The input is the preprocessed data, and the output is the trained generative AI model. Specifically, the server supplies the training data to the model in batches and updates the model's parameters as learning progresses. The end result is a generative AI model that effectively captures linguistic characteristics and humorous elements.

[0109] Step 4:

[0110] The user uses a device to input specific themes and keywords into the generative AI model. This input is sent to the server through an input form on the device. For example, the user might input "pet topics" or "comedy elements." The input consists of themes and keywords, and this information sent to the server becomes the input for processing.

[0111] Step 5:

[0112] The server uses a generative AI model to generate a text generation script based on the theme and keywords sent by the user. The input is the theme and keywords, and the output is the generated text generation script. During the generation process, the generative AI model creates dialogue-style text that is optimal for the theme and keywords. For example, it generates dialogue-style text that includes specific humorous elements, such as "happenings that happen while walking the dog" or "unexpected cat behavior."

[0113] Step 6:

[0114] The server provides the generated sentence generation script to the user's device. The user practices sentence generation by reading the provided script aloud. The input is the generated script, and the output is the script displayed on the user's device. The user can practice using this script and use it in actual performances. By publishing the practice process on social media or video platforms, users can spread the performance and potential of generative AI to many people.

[0115] (Application example 1)

[0116] 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."

[0117] Conventional AI systems have limitations in the quality of the content they generate and the way it is provided to users, making it difficult to achieve satisfactory results, especially in the entertainment field. They also lack convenience in terms of how users utilize and share the generated content. Furthermore, there is a need for a method that is easy for users to use and that also allows them to demonstrate the performance of the generating AI.

[0118] 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.

[0119] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, and means for training the generative model using the preprocessed data. This allows the server to generate a manzai script based on a theme or keywords provided by a user, and allows the user to practice and share the generated manzai script using a smartphone.

[0120] A "generative model" is an AI model that is trained to extract specific patterns and information from input data and perform problem solving or predictions.

[0121] "Means for collecting data" refers to the technical or methodological process for collecting the data required to train a generative model.

[0122] "Data pre-processing means" refers to the technical methods and processes used to process collected data in order to convert it into a form suitable for analysis or training.

[0123] "Generative model training" refers to the technical methods and processes that use collected and pre-processed data to train an AI model to perform a specific task.

[0124] "Means for generating manzai scripts based on themes and keywords" refers to a technical method or process for automatically generating dialogue-style text for manzai using user-provided topics and key words.

[0125] "Means for providing the generated manzai script to the user" refers to the technical methods and processes for transmitting the manzai script generated by AI to the user's device and making it viewable.

[0126] "Means that allow users to practice and share the generated manzai script using a smartphone" refers to the technical methods and processes that allow users to practice the generated manzai script using a smartphone app and share the content with other people.

[0127] This invention is a system that uses AI to generate manzai scripts and allows users to practice and share the scripts via smartphone. To realize this system, the following hardware and software are used.

[0128] Hardware

[0129] Server: A high-performance computer for back-end processing, which collects data, preprocesses the data, trains the generative model, and provides scripts to users.

[0130] Client device: A smartphone (iOS or ANDROID (registered trademark)). A device for providing the generated manzai script to the user.

[0131] software

[0132] Server side:

[0133] Programming language: Python

[0134] Deep Learning Framework: TensorFlow or PyTorch

[0135] Data processing libraries: NumPy, Pandas

[0136] Client side:

[0137] Smartphone app: Swift (iOS) or Kotlin (Android)

[0138] System Elements

[0139] 1. Data collection and preprocessing

[0140] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is preprocessed through text cleansing, tokenization, and normalization.

[0141] 2. Training the generative model

[0142] Based on the preprocessed data, a generative AI model is trained using deep learning techniques. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0143] 3. User input of themes and keywords

[0144] The user uses a smartphone app to input the theme or keywords of the manzai, such as prompt sentences like "school memories" or "sports day."

[0145] 4. Providing a generation script

[0146] The server uses a generative AI model to generate a manzai script based on themes and keywords provided by the user. The script is conversational and includes natural-looking interactions and humor that fit the specific theme.

[0147] 5. Script Distribution and Sharing

[0148] The server sends the generated script to the user's smartphone, and the user uses this script to practice their manzai. The script can also be shared on social media and video platforms via the smartphone app.

[0149] Examples of prompt statements

[0150] Theme: School Memories

[0151] Keywords: Sports day

[0152] This allows users to practice using the generated manzai script and share their performances with others, making it a useful way to widely promote the capabilities of the generative AI.

[0153] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0154] Step 1:

[0155] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is saved in text file format.

[0156] Step 2:

[0157] The server performs text cleansing (removing noise data), tokenization (dividing data into words and sentences), and normalization (converting data into a unified format) on the collected data. The preprocessed data is then formatted in a way that is suitable for training generative AI models.

[0158] Step 3:

[0159] The server uses the preprocessed data to train a generative AI model. Using deep learning techniques, the model effectively learns the linguistic characteristics and humorous elements of Japanese over multiple epochs. Once trained, the model becomes the basis for AI-based manzai script generation.

[0160] Step 4:

[0161] The user opens the smartphone app and inputs the theme or keywords of the manzai. For example, they input prompts such as "school memories" or "sports day," and the input data is sent from the app to the server. This input data becomes a request for the generative AI model.

[0162] Step 5:

[0163] The server receives themes and keywords submitted by users and uses a generative AI model to generate a manzai script, which includes natural-sounding interactions and humor that fit the specific theme.

[0164] Step 6:

[0165] The generated manzai script is sent from the server to the user's smartphone, and the user practices the manzai using the received script.

[0166] Step 7:

[0167] Users can share the manzai scripts they have practiced on social media and video platforms via a smartphone app, which also serves as a way to widely publicize the capabilities of the generative AI.

[0168] This allows a manzai script generation system using generative AI to embody a series of processes from data collection to providing and sharing the generated script.

[0169] 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.

[0170] The present invention provides a system that combines a generation AI and an emotion engine to generate a manzai script and provide it to a user. Specific embodiments are described below.

[0171] First, the server collects data to train the generative model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Methods of collection include web scraping, using APIs, and extracting data from existing databases.

[0172] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. At this stage, the data is formatted in a way that is suitable for training generative AI models.

[0173] The server then uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0174] The user inputs the theme and keywords of the manzai from the terminal, and the terminal sends this information to the server.

[0175] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh and episodes that they liked in the past, and reflects this in the generated script.

[0176] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During the generation process, natural dialogue is created, taking into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0177] The generated manzai script is sent from the server to the terminal, where the user can check the received script on the terminal and modify or regenerate it as necessary.

[0178] Finally, users practice using a manzai script provided by a generative AI using an emotion engine, which includes real-time updates and personalized elements based on past user emotional data.

[0179] As a specific example, if a user enters the theme "travel mishaps" and the keywords "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. Users can use this script to practice a stand-up comedy routine and post the process on social media or video platforms, thereby widely showcasing the capabilities of the generative AI and emotion engine.

[0180] As described above, the present invention combines a generation AI and an emotion engine to realize a system that generates manzai scripts that are more personalized for users and efficiently provides the results.

[0181] The processing flow will be explained below.

[0182] Step 1:

[0183] The server collects data to train the generative model. Specifically, it uses web scraping to obtain publicly available data on the Internet and APIs to extract data from existing databases. It also collects scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0184] Step 2:

[0185] The server preprocesses the collected data. At this stage, the data is first cleansed. Specifically, noise data and unnecessary information are removed. Next, the data is tokenized, and text is divided into words and sentences. The data is then normalized to standardize its format and prepare it in a format suitable for training a generative AI model.

[0186] Step 3:

[0187] The server uses the preprocessed data to train a generative AI model, typically using deep learning techniques. The model is trained over multiple epochs to learn the characteristics of the Japanese language and elements of humor.

[0188] Step 4:

[0189] The user inputs the theme and keywords of the manzai from the terminal. The user enters the content in the input field and clicks the "Send" button. This information is sent from the terminal to the server.

[0190] Step 5:

[0191] When the server receives themes and keywords provided by the user, the emotion engine is activated. The emotion engine collects the user's past emotional data and analyzes emotional changes and patterns. This allows it to understand the trends in manzai scripts that users prefer.

[0192] Step 6:

[0193] The server uses a generative AI model combined with an emotion engine to generate a manzai script based on themes and keywords provided by the user, and references the user's past emotional data to create an optimal script based on their emotions.

[0194] Step 7:

[0195] The server sends the generated manzai script to the terminal, which then sends the script to the user's terminal, where it is received and displayed.

[0196] Step 8:

[0197] The user can review the received manzai script on their device and modify or regenerate it as needed. The emotion engine recognizes the user's emotions in real time and updates the script accordingly.

[0198] Step 9:

[0199] Users practice manzai based on the provided manzai script. The emotion engine continuously monitors the user's reactions and adjusts the script content as necessary.

[0200] Step 10:

[0201] Finally, users can use the generated manzai script to perform an actual comedy routine, and by sharing the process and results on social media and video platforms, they can widely promote the capabilities of the generative AI and emotion engine.

[0202] For example, if a user inputs the theme "Work Failure Stories" and the keyword "Office," the server will refer to similar stories that the user liked in the past and use the emotion engine to generate a stand-up comedy script that the user can most easily relate to and that includes humorous elements. The user can then practice using this script and finally perform it, bringing laughter to the audience.

[0203] Example 2

[0204] 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."

[0205] Conventional manzai script generation systems have had difficulty generating personalized content that reflects individual users' emotions and preferences. They also lack the ability to generate natural dialogue that takes into account Japanese grammar and humor, resulting in a decline in the quality of the generated results. Furthermore, there was no system that reflected the user's emotions in real time during the generation process, making it difficult to increase user satisfaction.

[0206] 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 collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by a user, means for providing the generated manzai script to the user, and means for collecting user emotion data and adjusting the script content. This makes it possible to generate a high-quality manzai script personalized to the user's emotions and preferences.

[0207] "Data for training a generative model" refers to the information needed to train a generative AI model, including performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0208] "Means of collection" refers to the techniques and methods for effectively obtaining specific information, including web scraping, use of APIs, and extraction from existing databases.

[0209] "Preprocessing means" refers to a series of processes for converting collected data into an appropriate format, including data cleansing (removing noisy data), tokenization (dividing data), and normalization (standardizing data formats).

[0210] "Means for training a generative model" refers to techniques and methods for training a generative AI model using preprocessed data, including using deep learning techniques to train the model over a number of epochs.

[0211] "Means for generating a manzai script based on themes and keywords provided by the user" refers to a technology for automatically generating a manzai script based on user input, and the generated script includes natural dialogue that takes into account Japanese grammar and humor.

[0212] "Means for providing the generated manzai script to the user" refers to a technique or method for sending the generated manzai script to the user's terminal so that the user can check, modify, and regenerate it.

[0213] "Means for collecting user emotional data and adjusting script content" refers to technology that analyzes the user's past emotional data and adaptively changes the script content based on that data, using an emotion engine.

[0214] The present invention is a manzai script generation system that combines a generative AI model and an emotion engine, and provides users with personalized manzai scripts. Specific embodiments are described below.

[0215] First, the server collects data to train the generative AI model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Possible collection methods include web scraping, using APIs, and extracting data from existing databases. The hardware and software used include Cruise API and various web scraping tools.

[0216] Next, the server preprocesses the collected data. Preprocessing includes data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). At this stage, the data is formatted in a way that is suitable for training generative AI models. Specific software used includes the pandas library for data cleansing, SudachiPy for tokenization, and the nlp toolkit for normalization.

[0217] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques, possibly using frameworks like TensorFlow or PyTorch. The training process involves multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0218] When a user inputs a theme or keyword for a comedy routine, the device receives the information and sends it to the server. For example, if a user inputs the keywords "travel failure stories" and "family trip," the device sends this to the server.

[0219] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh in the past and their favorite episodes, and reflects this in the generated script. The emotion analysis technology used includes emotion recognition APIs and history learning technology.

[0220] The server then combines the generative AI model with an emotion engine to generate a manzai script based on themes and keywords entered by the user. The generation process creates natural dialogue that takes into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0221] The generated manzai script is sent from the server to the terminal. The user can check the received script on the terminal and modify or regenerate it as necessary. For example, if the user is not satisfied with part of the generated script, they can enter new keywords and regenerate it.

[0222] Finally, users practice using a manzai script provided by the generative AI using the emotion engine. The generated script includes real-time updates and personalized elements based on past user emotional data. By publishing the practiced script on social media and video platforms, users can widely demonstrate the capabilities of the generative AI and emotion engine.

[0223] As a specific example, if a user enters the theme "travel failure stories" and the keyword "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes incidents that occurred during the trip and funny interactions between family members. An example of a prompt for the generation AI model is, "Please generate a stand-up comedy script based on travel failure stories. The theme is travel failure stories, and the keyword is family trip."

[0224] As described above, the system of the present invention combines a generative AI model and an emotion engine to provide users with personalized manzai scripts and deliver the results efficiently.

[0225] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0226] Step 1: Data collection

[0227] The server collects data to train the generative AI model. The inputs include performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Specifically, it uses web scraping tools and APIs to obtain information from the internet and extracts the necessary data from existing databases. The output is the collected dataset.

[0228] Step 2: Data Preprocessing

[0229] The server preprocesses the collected data. The input is the collected dataset. Specifically, it first cleanses the data using the pandas library to remove noise. Then it splits the data into tokens using SudachiPy, and finally it normalizes the data format using the nlp toolkit. The output is the preprocessed dataset.

[0230] Step 3: Training the generative AI model

[0231] The server trains a generative AI model using the preprocessed data. The input is a preprocessed dataset. Specifically, it trains a deep learning model using TensorFlow or PyTorch. The model is trained over multiple epochs to learn the characteristics and humor of the Japanese language. The output is a trained generative AI model.

[0232] Step 4: Receiving User Input

[0233] The terminal receives input of the theme and keywords of the manzai from the user. The input includes the theme and keywords provided by the user. Specifically, when the user enters keywords such as "travel failure story" and "family trip" into the input field, the terminal sends this to the server. The output is the theme and keywords sent to the server.

[0234] Step 5: Collect and process emotion data

[0235] The server uses an emotion engine to collect and process the user's past emotional data. The input is the user's past emotional data. Specifically, it uses an emotion recognition API to collect scenes that made the user smile and favorite episodes from the past, and analyzes emotional patterns. The output is the analyzed emotional data.

[0236] Step 6: Generate the Manzai script

[0237] The server combines a generative AI model and an emotion engine to generate a manzai script based on the theme and keywords entered by the user. The inputs include the theme, keywords, and emotion data. Specifically, the generative AI model generates the script while taking into account Japanese language grammar and humor, and the emotion engine adjusts the script content in real time to reflect the user's emotion data. The output is the generated manzai script.

[0238] Step 7: Publish and check your Manzai script

[0239] The server sends the generated manzai script to the terminal. The input is the generated manzai script. Specifically, the server sends the script, the terminal receives it, and displays it to the user. The user checks the script and requests corrections or regeneration as necessary. The output is the manzai script checked on the terminal.

[0240] Step 8: Practice and share your scripts

[0241] The user practices based on the provided manzai script. The input is the confirmed manzai script. The specific operation is for the user to check the script, practice, record the process, and publish it on social media or a video platform. The output is the user's practice results and published content.

[0242] The above are the specific processing steps in the program of this system.

[0243] (Application example 2)

[0244] 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."

[0245] Conventional manzai script generation systems generate scripts uniformly without considering the user's emotions, resulting in a lack of personalization according to the preferences and emotions of individual users. Furthermore, they do not sufficiently consider the linguistic characteristics of the Japanese language, making it difficult to generate natural dialogue.

[0246] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by the user, means for providing the generated manzai script to the user, and means for recognizing the user's emotions and adjusting the content of the manzai script based on the emotions. This makes it possible to generate a personalized manzai script according to the user's emotional state.

[0247] "Means of collecting data for training a generative model" refers to means of collecting performance data from manzai competitions, literature on comedy, Japanese dialogue data, etc., using methods such as web scraping, using APIs, and extracting data from existing databases.

[0248] "Means for preprocessing collected data" refers to means for performing processes on collected data such as cleansing (removal of noise data), tokenization (dividing data), and normalization (standardization of data format).

[0249] "Means for training a generative model using preprocessed data" refers to means for training a generative AI model using deep learning techniques with preprocessed data as input.

[0250] "Means for generating a manzai script based on themes and keywords provided by the user" refers to means for generating a manzai script using a generative AI model based on themes and keywords entered by the user.

[0251] The "means for providing the generated manzai script to the user" is a means for transmitting the generated script to the terminal so that the user can check it.

[0252] "Means for recognizing the user's emotions and adjusting the content of the manzai script based on that" refers to means for analyzing the user's past emotional data and real-time emotional state, and using that information to adjust the content of the generated script as appropriate.

[0253] "Means for generating natural dialogue taking into account the linguistic characteristics of Japanese" refers to means for generating a manzai script as natural dialogue taking into account the grammar, humor, and linguistic characteristics unique to Japanese.

[0254] "Means including data on the performance of manzai competitions and literature on humor" refers to means of including data on the performance of manzai competitions and literature on comedy in the dataset required to train a generative model.

[0255] The present invention uses a system that combines a generative AI model and an emotion engine to generate and provide a personalized manzai script to a user. Specific embodiments of the present invention are described below.

[0256] First, the server collects data to train the generative model. The data collection methods include web scraping, API usage, and extraction from existing databases. The data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0257] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. This prepares the collected data in a format suitable for learning generative AI models.

[0258] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques and is trained over multiple epochs to learn Japanese grammar and humorous elements. The training dataset includes performance data from manzai competitions and literature on humor.

[0259] The user inputs the theme and keywords of the manzai through the terminal. The terminal then sends this information to the server. At this point, the emotion engine is introduced. The emotion engine references the user's past emotional data and adjusts the content of the manzai script based on that. For example, it can learn episodes that made the user laugh in the past and favorite keywords, and reflect these in the generated script.

[0260] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During this generation process, natural dialogue is created taking into account the characteristics of the Japanese language. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0261] The generated manzai script is sent from the server to the device, where the user can review it. The user can review the script and modify or regenerate it as needed. Finally, the user can practice using the generated manzai script and post the results on social media or video platforms.

[0262] As a concrete example, consider the case where a user inputs the topic "Travel Mishaps" and the keyword "Family Trip." In this case, the emotion engine will refer to past stories that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. The following is an example of the prompt that generates the script:

[0263] You are an AI that generates funny stand-up scripts. In the past, users found the "Travel Mishaps" and "Family Trip" episodes particularly funny.

[0264] Create the following script:

[0265] Theme: Travel

[0266] Keywords: family trip, travel incident

[0267] Based on the script generated in this way, users can practice their manzai and widely share the process and results on social media and video platforms.

[0268] This allows for the generation of personalized manzai scripts according to the user's emotional state, improving the entertainment experience.

[0269] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0270] Step 1:

[0271] The server collects data to train the generative model. Specifically, it collects performance data from manzai competitions, literature on comedy, and Japanese dialogue data by web scraping, using APIs, and extracting data from existing databases. The input is this raw data, and the output is the collected dataset.

[0272] Step 2:

[0273] The server preprocesses the collected data. Specifically, it performs data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). The input is the collected dataset, and the output is the preprocessed data.

[0274] Step 3:

[0275] The server uses the preprocessed data to train a generative AI model. Specifically, it uses deep learning techniques to train the model to learn Japanese grammar and humorous elements. The input is the preprocessed data, and the output is a trained generative AI model.

[0276] Step 4:

[0277] The user inputs the theme and keywords of the manzai from the terminal. Specifically, the theme and keywords are entered in an input form and sent to the server. The input is the theme and keywords specified by the user, and the output is the request data sent to the server.

[0278] Step 5:

[0279] The server uses an emotion engine to recognize the user's emotional data. Specifically, it adjusts prompts by referring to the user's past laughing episodes and favorite keywords. The input is the user's emotional data, themes, and keywords, and the output is an adjusted prompt sentence.

[0280] Step 6:

[0281] The server uses the adjusted prompt sentence to ask the generative AI model to generate a manzai script. Specifically, the adjusted prompt sentence is input into the generative AI model to generate a script. The input is the adjusted prompt sentence, and the output is the generated manzai script.

[0282] Step 7:

[0283] The server sends the generated manzai script to the terminal. Specifically, it sends the generated script to the terminal so that the user can check it. The input is the generated manzai script, and the output is the script displayed on the user's terminal.

[0284] Step 8:

[0285] The user checks the received script and modifies or regenerates it as necessary. Specifically, the user checks the script on the terminal and sends instructions to the server to edit or regenerate it. The input is the generated script, and the output is the modified or newly generated script.

[0286] Step 9:

[0287] Finally, the user practices a manzai based on the generated script and publishes the results on social media or video platforms. Specifically, the user performs a manzai based on the script, records the process, and publishes it as a video. The input is the generated script, and the output is the published manzai video.

[0288] 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.

[0289] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0290] 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.

[0291] [Second embodiment]

[0292] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0293] 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.

[0294] 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).

[0295] 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.

[0296] 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.

[0297] 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).

[0298] 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.

[0299] 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.

[0300] 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.

[0301] 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.

[0302] 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.

[0303] 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."

[0304] This invention provides a system that uses a generative AI to generate a manzai script and publicly demonstrates the AI's capabilities through the process and results. Specific embodiments are described below.

[0305] First, the server collects the data necessary to train the generative model. This can be done by web scraping, using APIs, or extracting data from existing databases. The collected data includes scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0306] The collected data is preprocessed by the server. Examples of preprocessing include text cleansing (removing noise data), tokenization (splitting into words or sentences), and normalization (conversion into a unified format), which prepares the data in a format suitable for training a generative model.

[0307] The server then trains a generative AI model based on the preprocessed data. Deep learning techniques are used for this purpose. The model is trained over multiple epochs to effectively capture the characteristics of the Japanese language and its humorous elements.

[0308] Users input the theme and keywords of the manzai using their device. Based on this input sent from the device to the server, the server generates a manzai script using the generated generative AI model. The generated script is a dialogue-style script that includes natural exchanges and humor in line with the specific theme.

[0309] Finally, the server sends the generated manzai script to the user's device. The user can practice the manzai based on the provided script and use it in actual performances. In addition, by sharing the process and results of practicing manzai, the user can demonstrate the performance of the generative AI to others.

[0310] For example, if a user inputs the keywords "pet topic" and "comedy element," the server will generate a manzai script based on this information, including funny pet-related and everyday anecdotes. The script is provided in an interactive format, including anecdotes such as "an incident that occurred while walking the dog" or "unexpected cat behavior." Users can use this script to practice manzai and aim to win the M-1 Grand Prix. Users can also share the capabilities of the generating AI by posting their practice process on social media or video platforms.

[0311] The processing flow will be explained below.

[0312] Step 1:

[0313] The server collects data to train the generative model, including scripts from past manzai competitions, literature on comedy, and Japanese dialogue data, through web scraping, API usage, and extraction from existing databases.

[0314] Step 2:

[0315] The server preprocesses the collected data, which includes text cleansing (removing noise data), tokenization (splitting into words and sentences), and normalization (converting into a unified format) to prepare the data in a format suitable for training a generative model.

[0316] Step 3:

[0317] The server uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0318] Step 4:

[0319] The user inputs the theme and keywords of the manzai using a terminal, which are then sent to the server.

[0320] Step 5:

[0321] The server uses the generated AI model to generate a manzai script based on themes and keywords provided by the user. This generation process creates natural dialogue that takes into account Japanese grammar and humor.

[0322] Step 6:

[0323] The generated manzai script is sent from the server to the user's device, where the user can view and check the received script.

[0324] Step 7:

[0325] The user practices a manzai based on the provided manzai script. If necessary, the user can modify the script content or input additional themes or keywords to generate the script again.

[0326] Step 8:

[0327] Ultimately, users can complete the manzai and use it in actual performances. In addition, by sharing the practice process and results on social media and video platforms, the performance and effectiveness of the generative AI can be widely promoted.

[0328] Example 1

[0329] 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."

[0330] Conventional text generation systems have had difficulty effectively incorporating the linguistic characteristics of Japanese and humorous elements when generating natural dialogue-style text based on specific themes or keywords. Furthermore, the generated text often fails to meet the user's intentions or expectations, resulting in scripts that lack entertainment value and practicality.

[0331] 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.

[0332] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a sentence generation script based on a theme or keywords provided by a user, and means for providing the generated sentence generation script to the user. This enables script generation in a natural, interactive format that takes into account the linguistic characteristics and humorous elements of the Japanese language.

[0333] A "generative model" refers to an algorithm that uses artificial intelligence techniques to generate output such as text or images based on specific input.

[0334] "Data collection" refers to the process of gathering the information needed to train a generative model from the internet or existing databases.

[0335] "Preprocessing" refers to the process of converting collected data into a format suitable for training a generative model by means of noise removal, tokenization, normalization, etc.

[0336] "User" refers to a person who uses the system to create a text generation script based on a specific theme or keyword.

[0337] "Themes and keywords" refer to important words and phrases that the user specifies to the system to determine the content and direction of the script to be generated.

[0338] A "text generation script" refers to a dialogue-style text created using a generative model based on a specific theme or keyword.

[0339] "Linguistic characteristics of Japanese" refers to characteristics that include the grammatical structure, expression methods, and cultural background that are unique to Japanese.

[0340] "Natural dialogue" refers to fluent, coherent conversation that is natural in human communication.

[0341] "Entertainment event performance data" refers to data on entertainment events that have been held in the past, and is the basic data used to generate high-quality scripts by training a generative model based on that data.

[0342] "Laughter literature" refers to research materials, books, papers, etc. on humor and comedy, and refers to the subjects used to train generative models.

[0343] MODE FOR CARRYING OUT THE INVENTION

[0344] This invention is a system that uses a generative AI model to create a sentence generation script based on a specific theme or keyword, and provides the results to the user.

[0345] Hardware and software used

[0346] The server collects data, preprocesses it, trains it, and generates scripts to train the generative model. Specifically, the following hardware and software are used:

[0347] Hardware: Computational servers with high-performance CPUs and GPUs, and large-capacity storage devices for storing data.

[0348] Software: The web scraping tool "BeautifulSoup" is used for data collection, the natural language processing library "MeCab" is used for data preprocessing, and the deep learning frameworks "TensorFlow" or "PyTorch" are used for training the generative model.

[0349] Data collection

[0350] The server uses web scraping tools, APIs, and existing databases to collect the data needed to train the generative model. This data includes performance data from entertainment events and literature on comedy, such as scripts from past manzai competitions, books on comedy, and papers.

[0351] Data Preprocessing

[0352] The collected data is pre-processed by the server. Specifically, the following processes are performed:

[0353] Text cleansing: Remove noise data and unnecessary characters.

[0354] Tokenization: Use "MeCab" to divide the data into words and sentences.

[0355] Normalization: Transforming data into a uniform format and making it consistent.

[0356] Training generative AI models

[0357] Based on the preprocessed data, the server trains a generative AI model using the deep learning frameworks TensorFlow or PyTorch. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0358] User input

[0359] The user uses a device to input themes and keywords into the generative AI model. For example, keywords such as "pet topics" or "comedy elements" are input. This input information is then sent from the device to the server.

[0360] Creating and providing sentence generation scripts

[0361] The server uses a generative AI model to generate a text generation script based on themes and keywords submitted by the user. This script is a natural dialogue based on a specific theme, incorporating Japanese linguistic characteristics and humorous elements.

[0362] The generated script is provided to the user's device from the server. The user can practice generating sentences based on the provided script and use it in actual performances.

[0363] Prompt Sentence Examples

[0364] The following example is one of the prompts that a user can input to a generative AI model.

[0365] Create a text generation script about pets, with a comedic twist. For example, include a story about something that happened while walking a dog or an unexpected cat behavior.

[0366] The above is a specific embodiment for carrying out the present invention.

[0367] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0368] Step 1:

[0369] The server collects data to train the generative model. This data includes performance data on entertainment events that incorporate Japanese linguistic characteristics and humorous elements, as well as literature on laughter. Specifically, web scraping tools, APIs, and existing databases are used. Web scraping tools such as "BeautifulSoup" are used to collect comedy scripts from the Internet. Comedy-related literature is obtained through APIs, and the necessary data is extracted from existing databases. The collected data is saved in a format suitable for training the generative model.

[0370] Step 2:

[0371] The server pre-processes the collected data, which includes the following specific steps:

[0372] Cleansing: Remove noise data, specifically, remove unnecessary characters and advertisements.

[0373] Tokenization: We use the natural language processing library "MeCab" to divide the data into words and sentences.

[0374] Normalization: Transforming data into a uniform format and making it consistent.

[0375] The input is the collected raw data, and the output is pre-processed data suitable for training a generative model, which is then cleansed, tokenized, and stored in a normalized format.

[0376] Step 3:

[0377] The server uses the preprocessed data to train a generative AI model. Using the deep learning frameworks "TensorFlow" or "PyTorch," the model is trained over multiple epochs. The input is the preprocessed data, and the output is the trained generative AI model. Specifically, the server supplies the training data to the model in batches and updates the model's parameters as learning progresses. The end result is a generative AI model that effectively captures linguistic characteristics and humorous elements.

[0378] Step 4:

[0379] The user uses a device to input specific themes and keywords into the generative AI model. This input is sent to the server through an input form on the device. For example, the user might input "pet topics" or "comedy elements." The input consists of themes and keywords, and this information sent to the server becomes the input for processing.

[0380] Step 5:

[0381] The server uses a generative AI model to generate a text generation script based on the theme and keywords sent by the user. The input is the theme and keywords, and the output is the generated text generation script. During the generation process, the generative AI model creates dialogue-style text that is optimal for the theme and keywords. For example, it generates dialogue-style text that includes specific humorous elements, such as "happenings that happen while walking the dog" or "unexpected cat behavior."

[0382] Step 6:

[0383] The server provides the generated sentence generation script to the user's device. The user practices sentence generation by reading the provided script aloud. The input is the generated script, and the output is the script displayed on the user's device. The user can practice using this script and use it in actual performances. By publishing the practice process on social media or video platforms, users can spread the performance and potential of generative AI to many people.

[0384] (Application example 1)

[0385] 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."

[0386] Conventional AI systems have limitations in the quality of the content they generate and the way it is provided to users, making it difficult to achieve satisfactory results, especially in the entertainment field. They also lack convenience in terms of how users utilize and share the generated content. Furthermore, there is a need for a method that is easy for users to use and that also allows them to demonstrate the performance of the generating AI.

[0387] 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.

[0388] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, and means for training the generative model using the preprocessed data. This allows the server to generate a manzai script based on a theme or keywords provided by a user, and allows the user to practice and share the generated manzai script using a smartphone.

[0389] A "generative model" is an AI model that is trained to extract specific patterns and information from input data and perform problem solving or predictions.

[0390] "Means for collecting data" refers to the technical or methodological process for collecting the data required to train a generative model.

[0391] "Data pre-processing means" refers to the technical methods and processes used to process collected data in order to convert it into a form suitable for analysis or training.

[0392] "Generative model training" refers to the technical methods and processes that use collected and pre-processed data to train an AI model to perform a specific task.

[0393] "Means for generating manzai scripts based on themes and keywords" refers to a technical method or process for automatically generating dialogue-style text for manzai using user-provided topics and key words.

[0394] "Means for providing the generated manzai script to the user" refers to the technical methods and processes for transmitting the manzai script generated by AI to the user's device and making it viewable.

[0395] "Means that allow users to practice and share the generated manzai script using a smartphone" refers to the technical methods and processes that allow users to practice the generated manzai script using a smartphone app and share the content with other people.

[0396] This invention is a system that uses AI to generate manzai scripts and allows users to practice and share the scripts via smartphone. To realize this system, the following hardware and software are used.

[0397] Hardware

[0398] Server: A high-performance computer for back-end processing, which collects data, preprocesses the data, trains the generative model, and provides scripts to users.

[0399] Client device: A smartphone (iOS or Android). This device is used to provide the generated manzai script to the user.

[0400] software

[0401] Server side:

[0402] Programming language: Python

[0403] Deep Learning Framework: TensorFlow or PyTorch

[0404] Data processing libraries: NumPy, Pandas

[0405] Client side:

[0406] Smartphone app: Swift (iOS) or Kotlin (Android)

[0407] System Elements

[0408] 1. Data collection and preprocessing

[0409] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is preprocessed through text cleansing, tokenization, and normalization.

[0410] 2. Training the generative model

[0411] Based on the preprocessed data, a generative AI model is trained using deep learning techniques. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0412] 3. User input of themes and keywords

[0413] The user uses a smartphone app to input the theme or keywords of the manzai, such as prompt sentences like "school memories" or "sports day."

[0414] 4. Providing a generation script

[0415] The server uses a generative AI model to generate a manzai script based on themes and keywords provided by the user. The script is conversational and includes natural-looking interactions and humor that fit the specific theme.

[0416] 5. Script Distribution and Sharing

[0417] The server sends the generated script to the user's smartphone, and the user uses this script to practice their manzai. The script can also be shared on social media and video platforms via the smartphone app.

[0418] Examples of prompt statements

[0419] Theme: School Memories

[0420] Keywords: Sports day

[0421] This allows users to practice using the generated manzai script and share their performances with others, making it a useful way to widely promote the capabilities of the generative AI.

[0422] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0423] Step 1:

[0424] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is saved in text file format.

[0425] Step 2:

[0426] The server performs text cleansing (removing noise data), tokenization (dividing data into words and sentences), and normalization (converting data into a unified format) on the collected data. The preprocessed data is then formatted in a way that is suitable for training generative AI models.

[0427] Step 3:

[0428] The server uses the preprocessed data to train a generative AI model. Using deep learning techniques, the model effectively learns the linguistic characteristics and humorous elements of Japanese over multiple epochs. Once trained, the model becomes the basis for AI-based manzai script generation.

[0429] Step 4:

[0430] The user opens the smartphone app and inputs the theme or keywords of the manzai. For example, they input prompts such as "school memories" or "sports day," and the input data is sent from the app to the server. This input data becomes a request for the generative AI model.

[0431] Step 5:

[0432] The server receives themes and keywords submitted by users and uses a generative AI model to generate a manzai script, which includes natural-sounding interactions and humor that fit the specific theme.

[0433] Step 6:

[0434] The generated manzai script is sent from the server to the user's smartphone, and the user practices the manzai using the received script.

[0435] Step 7:

[0436] Users can share the manzai scripts they have practiced on social media and video platforms via a smartphone app, which also serves as a way to widely publicize the capabilities of the generative AI.

[0437] This allows a manzai script generation system using generative AI to embody a series of processes from data collection to providing and sharing the generated script.

[0438] 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.

[0439] The present invention provides a system that combines a generation AI and an emotion engine to generate a manzai script and provide it to a user. Specific embodiments are described below.

[0440] First, the server collects data to train the generative model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Methods of collection include web scraping, using APIs, and extracting data from existing databases.

[0441] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. At this stage, the data is formatted in a way that is suitable for training generative AI models.

[0442] The server then uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0443] The user inputs the theme and keywords of the manzai from the terminal, and the terminal sends this information to the server.

[0444] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh and episodes that they liked in the past, and reflects this in the generated script.

[0445] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During the generation process, natural dialogue is created, taking into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0446] The generated manzai script is sent from the server to the terminal, where the user can check the received script on the terminal and modify or regenerate it as necessary.

[0447] Finally, users practice using a manzai script provided by a generative AI using an emotion engine, which includes real-time updates and personalized elements based on past user emotional data.

[0448] As a specific example, if a user enters the theme "travel mishaps" and the keywords "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. Users can use this script to practice a stand-up comedy routine and post the process on social media or video platforms, thereby widely showcasing the capabilities of the generative AI and emotion engine.

[0449] As described above, the present invention combines a generation AI and an emotion engine to realize a system that generates manzai scripts that are more personalized for users and efficiently provides the results.

[0450] The processing flow will be explained below.

[0451] Step 1:

[0452] The server collects data to train the generative model. Specifically, it uses web scraping to obtain publicly available data on the Internet and APIs to extract data from existing databases. It also collects scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0453] Step 2:

[0454] The server preprocesses the collected data. At this stage, the data is first cleansed. Specifically, noise data and unnecessary information are removed. Next, the data is tokenized, and text is divided into words and sentences. The data is then normalized to standardize its format and prepare it in a format suitable for training a generative AI model.

[0455] Step 3:

[0456] The server uses the preprocessed data to train a generative AI model, typically using deep learning techniques. The model is trained over multiple epochs to learn the characteristics of the Japanese language and elements of humor.

[0457] Step 4:

[0458] The user inputs the theme and keywords of the manzai from the terminal. The user enters the content in the input field and clicks the "Send" button. This information is sent from the terminal to the server.

[0459] Step 5:

[0460] When the server receives themes and keywords provided by the user, the emotion engine is activated. The emotion engine collects the user's past emotional data and analyzes emotional changes and patterns. This allows it to understand the trends in manzai scripts that users prefer.

[0461] Step 6:

[0462] The server uses a generative AI model combined with an emotion engine to generate a manzai script based on themes and keywords provided by the user, and references the user's past emotional data to create an optimal script based on their emotions.

[0463] Step 7:

[0464] The server sends the generated manzai script to the terminal, which then sends the script to the user's terminal, where it is received and displayed.

[0465] Step 8:

[0466] The user can review the received manzai script on their device and modify or regenerate it as needed. The emotion engine recognizes the user's emotions in real time and updates the script accordingly.

[0467] Step 9:

[0468] Users practice manzai based on the provided manzai script. The emotion engine continuously monitors the user's reactions and adjusts the script content as necessary.

[0469] Step 10:

[0470] Finally, users can use the generated manzai script to perform an actual comedy routine, and by sharing the process and results on social media and video platforms, they can widely promote the capabilities of the generative AI and emotion engine.

[0471] For example, if a user inputs the theme "Work Failure Stories" and the keyword "Office," the server will refer to similar stories that the user liked in the past and use the emotion engine to generate a stand-up comedy script that the user can most easily relate to and that includes humorous elements. The user can then practice using this script and finally perform it, bringing laughter to the audience.

[0472] Example 2

[0473] 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."

[0474] Conventional manzai script generation systems have had difficulty generating personalized content that reflects individual users' emotions and preferences. They also lack the ability to generate natural dialogue that takes into account Japanese grammar and humor, resulting in a decline in the quality of the generated results. Furthermore, there was no system that reflected the user's emotions in real time during the generation process, making it difficult to increase user satisfaction.

[0475] 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 collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by a user, means for providing the generated manzai script to the user, and means for collecting user emotion data and adjusting the script content. This makes it possible to generate a high-quality manzai script personalized to the user's emotions and preferences.

[0476] "Data for training a generative model" refers to the information needed to train a generative AI model, including performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0477] "Means of collection" refers to the techniques and methods for effectively obtaining specific information, including web scraping, use of APIs, and extraction from existing databases.

[0478] "Preprocessing means" refers to a series of processes for converting collected data into an appropriate format, including data cleansing (removing noisy data), tokenization (dividing data), and normalization (standardizing data formats).

[0479] "Means for training a generative model" refers to techniques and methods for training a generative AI model using preprocessed data, including using deep learning techniques to train the model over a number of epochs.

[0480] "Means for generating a manzai script based on themes and keywords provided by the user" refers to a technology for automatically generating a manzai script based on user input, and the generated script includes natural dialogue that takes into account Japanese grammar and humor.

[0481] "Means for providing the generated manzai script to the user" refers to a technique or method for sending the generated manzai script to the user's terminal so that the user can check, modify, and regenerate it.

[0482] "Means for collecting user emotional data and adjusting script content" refers to technology that analyzes the user's past emotional data and adaptively changes the script content based on that data, using an emotion engine.

[0483] The present invention is a manzai script generation system that combines a generative AI model and an emotion engine, and provides users with personalized manzai scripts. Specific embodiments are described below.

[0484] First, the server collects data to train the generative AI model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Possible collection methods include web scraping, using APIs, and extracting data from existing databases. The hardware and software used include Cruise API and various web scraping tools.

[0485] Next, the server preprocesses the collected data. Preprocessing includes data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). At this stage, the data is formatted in a way that is suitable for training generative AI models. Specific software used includes the pandas library for data cleansing, SudachiPy for tokenization, and the nlp toolkit for normalization.

[0486] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques, possibly using frameworks like TensorFlow or PyTorch. The training process involves multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0487] When a user inputs a theme or keyword for a comedy routine, the device receives the information and sends it to the server. For example, if a user inputs the keywords "travel failure stories" and "family trip," the device sends this to the server.

[0488] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh in the past and their favorite episodes, and reflects this in the generated script. The emotion analysis technology used includes emotion recognition APIs and history learning technology.

[0489] The server then combines the generative AI model with an emotion engine to generate a manzai script based on themes and keywords entered by the user. The generation process creates natural dialogue that takes into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0490] The generated manzai script is sent from the server to the terminal. The user can check the received script on the terminal and modify or regenerate it as necessary. For example, if the user is not satisfied with part of the generated script, they can enter new keywords and regenerate it.

[0491] Finally, users practice using a manzai script provided by the generative AI using the emotion engine. The generated script includes real-time updates and personalized elements based on past user emotional data. By publishing the practiced script on social media and video platforms, users can widely demonstrate the capabilities of the generative AI and emotion engine.

[0492] As a specific example, if a user enters the theme "travel failure stories" and the keyword "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes incidents that occurred during the trip and funny interactions between family members. An example of a prompt for the generation AI model is, "Please generate a stand-up comedy script based on travel failure stories. The theme is travel failure stories, and the keyword is family trip."

[0493] As described above, the system of the present invention combines a generative AI model and an emotion engine to provide users with personalized manzai scripts and deliver the results efficiently.

[0494] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0495] Step 1: Data collection

[0496] The server collects data to train the generative AI model. The inputs include performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Specifically, it uses web scraping tools and APIs to obtain information from the internet and extracts the necessary data from existing databases. The output is the collected dataset.

[0497] Step 2: Data Preprocessing

[0498] The server preprocesses the collected data. The input is the collected dataset. Specifically, it first cleanses the data using the pandas library to remove noise. Then it splits the data into tokens using SudachiPy, and finally it normalizes the data format using the nlp toolkit. The output is the preprocessed dataset.

[0499] Step 3: Training the generative AI model

[0500] The server trains a generative AI model using the preprocessed data. The input is a preprocessed dataset. Specifically, it trains a deep learning model using TensorFlow or PyTorch. The model is trained over multiple epochs to learn the characteristics and humor of the Japanese language. The output is a trained generative AI model.

[0501] Step 4: Receiving User Input

[0502] The terminal receives input of the theme and keywords of the manzai from the user. The input includes the theme and keywords provided by the user. Specifically, when the user enters keywords such as "travel failure story" and "family trip" into the input field, the terminal sends this to the server. The output is the theme and keywords sent to the server.

[0503] Step 5: Collect and process emotion data

[0504] The server uses an emotion engine to collect and process the user's past emotional data. The input is the user's past emotional data. Specifically, it uses an emotion recognition API to collect scenes that made the user smile and favorite episodes from the past, and analyzes emotional patterns. The output is the analyzed emotional data.

[0505] Step 6: Generate the Manzai script

[0506] The server combines a generative AI model and an emotion engine to generate a manzai script based on the theme and keywords entered by the user. The inputs include the theme, keywords, and emotion data. Specifically, the generative AI model generates the script while taking into account Japanese language grammar and humor, and the emotion engine adjusts the script content in real time to reflect the user's emotion data. The output is the generated manzai script.

[0507] Step 7: Publish and check your Manzai script

[0508] The server sends the generated manzai script to the terminal. The input is the generated manzai script. Specifically, the server sends the script, the terminal receives it, and displays it to the user. The user checks the script and requests corrections or regeneration as necessary. The output is the manzai script checked on the terminal.

[0509] Step 8: Practice and share your scripts

[0510] The user practices based on the provided manzai script. The input is the confirmed manzai script. The specific operation is for the user to check the script, practice, record the process, and publish it on social media or a video platform. The output is the user's practice results and published content.

[0511] The above are the specific processing steps in the program of this system.

[0512] (Application example 2)

[0513] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0514] Conventional manzai script generation systems generate scripts uniformly without considering the user's emotions, resulting in a lack of personalization according to the preferences and emotions of individual users. Furthermore, they do not sufficiently consider the linguistic characteristics of the Japanese language, making it difficult to generate natural dialogue.

[0515] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by the user, means for providing the generated manzai script to the user, and means for recognizing the user's emotions and adjusting the content of the manzai script based on the emotions. This makes it possible to generate a personalized manzai script according to the user's emotional state.

[0516] "Means of collecting data for training a generative model" refers to means of collecting performance data from manzai competitions, literature on comedy, Japanese dialogue data, etc., using methods such as web scraping, using APIs, and extracting data from existing databases.

[0517] "Means for preprocessing collected data" refers to means for performing processes on collected data such as cleansing (removal of noise data), tokenization (dividing data), and normalization (standardization of data format).

[0518] "Means for training a generative model using preprocessed data" refers to means for training a generative AI model using deep learning techniques with preprocessed data as input.

[0519] "Means for generating a manzai script based on themes and keywords provided by the user" refers to means for generating a manzai script using a generative AI model based on themes and keywords entered by the user.

[0520] The "means for providing the generated manzai script to the user" is a means for transmitting the generated script to the terminal so that the user can check it.

[0521] "Means for recognizing the user's emotions and adjusting the content of the manzai script based on that" refers to means for analyzing the user's past emotional data and real-time emotional state, and using that information to adjust the content of the generated script as appropriate.

[0522] "Means for generating natural dialogue taking into account the linguistic characteristics of Japanese" refers to means for generating a manzai script as natural dialogue taking into account the grammar, humor, and linguistic characteristics unique to Japanese.

[0523] "Means including data on the performance of manzai competitions and literature on humor" refers to means of including data on the performance of manzai competitions and literature on comedy in the dataset required to train a generative model.

[0524] The present invention uses a system that combines a generative AI model and an emotion engine to generate and provide a personalized manzai script to a user. Specific embodiments of the present invention are described below.

[0525] First, the server collects data to train the generative model. The data collection methods include web scraping, API usage, and extraction from existing databases. The data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0526] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. This prepares the collected data in a format suitable for learning generative AI models.

[0527] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques and is trained over multiple epochs to learn Japanese grammar and humorous elements. The training dataset includes performance data from manzai competitions and literature on humor.

[0528] The user inputs the theme and keywords of the manzai through the terminal. The terminal then sends this information to the server. At this point, the emotion engine is introduced. The emotion engine references the user's past emotional data and adjusts the content of the manzai script based on that. For example, it can learn episodes that made the user laugh in the past and favorite keywords, and reflect these in the generated script.

[0529] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During this generation process, natural dialogue is created taking into account the characteristics of the Japanese language. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0530] The generated manzai script is sent from the server to the device, where the user can review it. The user can review the script and modify or regenerate it as needed. Finally, the user can practice using the generated manzai script and post the results on social media or video platforms.

[0531] As a concrete example, consider the case where a user inputs the topic "Travel Mishaps" and the keyword "Family Trip." In this case, the emotion engine will refer to past stories that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. The following is an example of the prompt that generates the script:

[0532] You are an AI that generates funny stand-up scripts. In the past, users found the "Travel Mishaps" and "Family Trip" episodes particularly funny.

[0533] Create the following script:

[0534] Theme: Travel

[0535] Keywords: family trip, travel incident

[0536] Based on the script generated in this way, users can practice their manzai and widely share the process and results on social media and video platforms.

[0537] This allows for the generation of personalized manzai scripts according to the user's emotional state, improving the entertainment experience.

[0538] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0539] Step 1:

[0540] The server collects data to train the generative model. Specifically, it collects performance data from manzai competitions, literature on comedy, and Japanese dialogue data by web scraping, using APIs, and extracting data from existing databases. The input is this raw data, and the output is the collected dataset.

[0541] Step 2:

[0542] The server preprocesses the collected data. Specifically, it performs data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). The input is the collected dataset, and the output is the preprocessed data.

[0543] Step 3:

[0544] The server uses the preprocessed data to train a generative AI model. Specifically, it uses deep learning techniques to train the model to learn Japanese grammar and humorous elements. The input is the preprocessed data, and the output is a trained generative AI model.

[0545] Step 4:

[0546] The user inputs the theme and keywords of the manzai from the terminal. Specifically, the theme and keywords are entered in an input form and sent to the server. The input is the theme and keywords specified by the user, and the output is the request data sent to the server.

[0547] Step 5:

[0548] The server uses an emotion engine to recognize the user's emotional data. Specifically, it adjusts prompts by referring to the user's past laughing episodes and favorite keywords. The input is the user's emotional data, themes, and keywords, and the output is an adjusted prompt sentence.

[0549] Step 6:

[0550] The server uses the adjusted prompt sentence to ask the generative AI model to generate a manzai script. Specifically, the adjusted prompt sentence is input into the generative AI model to generate a script. The input is the adjusted prompt sentence, and the output is the generated manzai script.

[0551] Step 7:

[0552] The server sends the generated manzai script to the terminal. Specifically, it sends the generated script to the terminal so that the user can check it. The input is the generated manzai script, and the output is the script displayed on the user's terminal.

[0553] Step 8:

[0554] The user checks the received script and modifies or regenerates it as necessary. Specifically, the user checks the script on the terminal and sends instructions to the server to edit or regenerate it. The input is the generated script, and the output is the modified or newly generated script.

[0555] Step 9:

[0556] Finally, the user practices a manzai based on the generated script and publishes the results on social media or video platforms. Specifically, the user performs a manzai based on the script, records the process, and publishes it as a video. The input is the generated script, and the output is the published manzai video.

[0557] 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.

[0558] 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.

[0559] 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.

[0560] [Third embodiment]

[0561] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0562] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0563] 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).

[0564] 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.

[0565] 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.

[0566] 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).

[0567] 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.

[0568] 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.

[0569] 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.

[0570] 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.

[0571] 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.

[0572] 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."

[0573] This invention provides a system that uses a generative AI to generate a manzai script and publicly demonstrates the AI's capabilities through the process and results. Specific embodiments are described below.

[0574] First, the server collects the data necessary to train the generative model. This can be done by web scraping, using APIs, or extracting data from existing databases. The collected data includes scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0575] The collected data is preprocessed by the server. Examples of preprocessing include text cleansing (removing noise data), tokenization (splitting into words or sentences), and normalization (conversion into a unified format), which prepares the data in a format suitable for training a generative model.

[0576] The server then trains a generative AI model based on the preprocessed data. Deep learning techniques are used for this purpose. The model is trained over multiple epochs to effectively capture the characteristics of the Japanese language and its humorous elements.

[0577] Users input the theme and keywords of the manzai using their device. Based on this input sent from the device to the server, the server generates a manzai script using the generated generative AI model. The generated script is a dialogue-style script that includes natural exchanges and humor in line with the specific theme.

[0578] Finally, the server sends the generated manzai script to the user's device. The user can practice the manzai based on the provided script and use it in actual performances. In addition, by sharing the process and results of practicing manzai, the user can demonstrate the performance of the generative AI to others.

[0579] For example, if a user inputs the keywords "pet topic" and "comedy element," the server will generate a manzai script based on this information, including funny pet-related and everyday anecdotes. The script is provided in an interactive format, including anecdotes such as "an incident that occurred while walking the dog" or "unexpected cat behavior." Users can use this script to practice manzai and aim to win the M-1 Grand Prix. Users can also share the capabilities of the generating AI by posting their practice process on social media or video platforms.

[0580] The processing flow will be explained below.

[0581] Step 1:

[0582] The server collects data to train the generative model, including scripts from past manzai competitions, literature on comedy, and Japanese dialogue data, through web scraping, API usage, and extraction from existing databases.

[0583] Step 2:

[0584] The server preprocesses the collected data, which includes text cleansing (removing noise data), tokenization (splitting into words and sentences), and normalization (converting into a unified format) to prepare the data in a format suitable for training a generative model.

[0585] Step 3:

[0586] The server uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0587] Step 4:

[0588] The user inputs the theme and keywords of the manzai using a terminal, which are then sent to the server.

[0589] Step 5:

[0590] The server uses the generated AI model to generate a manzai script based on themes and keywords provided by the user. This generation process creates natural dialogue that takes into account Japanese grammar and humor.

[0591] Step 6:

[0592] The generated manzai script is sent from the server to the user's device, where the user can view and check the received script.

[0593] Step 7:

[0594] The user practices a manzai based on the provided manzai script. If necessary, the user can modify the script content or input additional themes or keywords to generate the script again.

[0595] Step 8:

[0596] Ultimately, users can complete the manzai and use it in actual performances. In addition, by sharing the practice process and results on social media and video platforms, the performance and effectiveness of the generative AI can be widely promoted.

[0597] Example 1

[0598] 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."

[0599] Conventional text generation systems have had difficulty effectively incorporating the linguistic characteristics of Japanese and humorous elements when generating natural dialogue-style text based on specific themes or keywords. Furthermore, the generated text often fails to meet the user's intentions or expectations, resulting in scripts that lack entertainment value and practicality.

[0600] 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.

[0601] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a sentence generation script based on a theme or keywords provided by a user, and means for providing the generated sentence generation script to the user. This enables script generation in a natural, interactive format that takes into account the linguistic characteristics and humorous elements of the Japanese language.

[0602] A "generative model" refers to an algorithm that uses artificial intelligence techniques to generate output such as text or images based on specific input.

[0603] "Data collection" refers to the process of gathering the information needed to train a generative model from the internet or existing databases.

[0604] "Preprocessing" refers to the process of converting collected data into a format suitable for training a generative model by means of noise removal, tokenization, normalization, etc.

[0605] "User" refers to a person who uses the system to create a text generation script based on a specific theme or keyword.

[0606] "Themes and keywords" refer to important words and phrases that the user specifies to the system to determine the content and direction of the script to be generated.

[0607] A "text generation script" refers to a dialogue-style text created using a generative model based on a specific theme or keyword.

[0608] "Linguistic characteristics of Japanese" refers to characteristics that include the grammatical structure, expression methods, and cultural background that are unique to Japanese.

[0609] "Natural dialogue" refers to fluent, coherent conversation that is natural in human communication.

[0610] "Entertainment event performance data" refers to data on entertainment events that have been held in the past, and is the basic data used to generate high-quality scripts by training a generative model based on that data.

[0611] "Laughter literature" refers to research materials, books, papers, etc. on humor and comedy, and refers to the subjects used to train generative models.

[0612] MODE FOR CARRYING OUT THE INVENTION

[0613] This invention is a system that uses a generative AI model to create a sentence generation script based on a specific theme or keyword, and provides the results to the user.

[0614] Hardware and software used

[0615] The server collects data, preprocesses it, trains it, and generates scripts to train the generative model. Specifically, the following hardware and software are used:

[0616] Hardware: Computational servers with high-performance CPUs and GPUs, and large-capacity storage devices for storing data.

[0617] Software: The web scraping tool "BeautifulSoup" is used for data collection, the natural language processing library "MeCab" is used for data preprocessing, and the deep learning frameworks "TensorFlow" or "PyTorch" are used for training the generative model.

[0618] Data collection

[0619] The server uses web scraping tools, APIs, and existing databases to collect the data needed to train the generative model. This data includes performance data from entertainment events and literature on comedy, such as scripts from past manzai competitions, books on comedy, and papers.

[0620] Data Preprocessing

[0621] The collected data is pre-processed by the server. Specifically, the following processes are performed:

[0622] Text cleansing: Remove noise data and unnecessary characters.

[0623] Tokenization: Use "MeCab" to divide the data into words and sentences.

[0624] Normalization: Transforming data into a uniform format and making it consistent.

[0625] Training generative AI models

[0626] Based on the preprocessed data, the server trains a generative AI model using the deep learning frameworks TensorFlow or PyTorch. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0627] User input

[0628] The user uses a device to input themes and keywords into the generative AI model. For example, keywords such as "pet topics" or "comedy elements" are input. This input information is then sent from the device to the server.

[0629] Creating and providing sentence generation scripts

[0630] The server uses a generative AI model to generate a text generation script based on themes and keywords submitted by the user. This script is a natural dialogue based on a specific theme, incorporating Japanese linguistic characteristics and humorous elements.

[0631] The generated script is provided to the user's device from the server. The user can practice generating sentences based on the provided script and use it in actual performances.

[0632] Prompt Sentence Examples

[0633] The following example is one of the prompts that a user can input to a generative AI model.

[0634] Create a text generation script about pets, with a comedic twist. For example, include a story about something that happened while walking a dog or an unexpected cat behavior.

[0635] The above is a specific embodiment for carrying out the present invention.

[0636] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0637] Step 1:

[0638] The server collects data to train the generative model. This data includes performance data on entertainment events that incorporate Japanese linguistic characteristics and humorous elements, as well as literature on laughter. Specifically, web scraping tools, APIs, and existing databases are used. Web scraping tools such as "BeautifulSoup" are used to collect comedy scripts from the Internet. Comedy-related literature is obtained through APIs, and the necessary data is extracted from existing databases. The collected data is saved in a format suitable for training the generative model.

[0639] Step 2:

[0640] The server pre-processes the collected data, which includes the following specific steps:

[0641] Cleansing: Remove noise data, specifically, remove unnecessary characters and advertisements.

[0642] Tokenization: We use the natural language processing library "MeCab" to divide the data into words and sentences.

[0643] Normalization: Transforming data into a uniform format and making it consistent.

[0644] The input is the collected raw data, and the output is pre-processed data suitable for training a generative model, which is then cleansed, tokenized, and stored in a normalized format.

[0645] Step 3:

[0646] The server uses the preprocessed data to train a generative AI model. Using the deep learning frameworks "TensorFlow" or "PyTorch," the model is trained over multiple epochs. The input is the preprocessed data, and the output is the trained generative AI model. Specifically, the server supplies the training data to the model in batches and updates the model's parameters as learning progresses. The end result is a generative AI model that effectively captures linguistic characteristics and humorous elements.

[0647] Step 4:

[0648] The user uses a device to input specific themes and keywords into the generative AI model. This input is sent to the server through an input form on the device. For example, the user might input "pet topics" or "comedy elements." The input consists of themes and keywords, and this information sent to the server becomes the input for processing.

[0649] Step 5:

[0650] The server uses a generative AI model to generate a text generation script based on the theme and keywords sent by the user. The input is the theme and keywords, and the output is the generated text generation script. During the generation process, the generative AI model creates dialogue-style text that is optimal for the theme and keywords. For example, it generates dialogue-style text that includes specific humorous elements, such as "happenings that happen while walking the dog" or "unexpected cat behavior."

[0651] Step 6:

[0652] The server provides the generated sentence generation script to the user's device. The user practices sentence generation by reading the provided script aloud. The input is the generated script, and the output is the script displayed on the user's device. The user can practice using this script and use it in actual performances. By publishing the practice process on social media or video platforms, users can spread the performance and potential of generative AI to many people.

[0653] (Application example 1)

[0654] 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."

[0655] Conventional AI systems have limitations in the quality of the content they generate and the way it is provided to users, making it difficult to achieve satisfactory results, especially in the entertainment field. They also lack convenience in terms of how users utilize and share the generated content. Furthermore, there is a need for a method that is easy for users to use and that also allows them to demonstrate the performance of the generating AI.

[0656] 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.

[0657] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, and means for training the generative model using the preprocessed data. This allows the server to generate a manzai script based on a theme or keywords provided by a user, and allows the user to practice and share the generated manzai script using a smartphone.

[0658] A "generative model" is an AI model that is trained to extract specific patterns and information from input data and perform problem solving or predictions.

[0659] "Means for collecting data" refers to the technical or methodological process for collecting the data required to train a generative model.

[0660] "Data pre-processing means" refers to the technical methods and processes used to process collected data in order to convert it into a form suitable for analysis or training.

[0661] "Generative model training" refers to the technical methods and processes that use collected and pre-processed data to train an AI model to perform a specific task.

[0662] "Means for generating manzai scripts based on themes and keywords" refers to a technical method or process for automatically generating dialogue-style text for manzai using user-provided topics and key words.

[0663] "Means for providing the generated manzai script to the user" refers to the technical methods and processes for transmitting the manzai script generated by AI to the user's device and making it viewable.

[0664] "Means that allow users to practice and share the generated manzai script using a smartphone" refers to the technical methods and processes that allow users to practice the generated manzai script using a smartphone app and share the content with other people.

[0665] This invention is a system that uses AI to generate manzai scripts and allows users to practice and share the scripts via smartphone. To realize this system, the following hardware and software are used.

[0666] Hardware

[0667] Server: A high-performance computer for back-end processing, which collects data, preprocesses the data, trains the generative model, and provides scripts to users.

[0668] Client device: A smartphone (iOS or Android). This device is used to provide the generated manzai script to the user.

[0669] software

[0670] Server side:

[0671] Programming language: Python

[0672] Deep Learning Framework: TensorFlow or PyTorch

[0673] Data processing libraries: NumPy, Pandas

[0674] Client side:

[0675] Smartphone app: Swift (iOS) or Kotlin (Android)

[0676] System Elements

[0677] 1. Data collection and preprocessing

[0678] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is preprocessed through text cleansing, tokenization, and normalization.

[0679] 2. Training the generative model

[0680] Based on the preprocessed data, a generative AI model is trained using deep learning techniques. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0681] 3. User input of themes and keywords

[0682] The user uses a smartphone app to input the theme or keywords of the manzai, such as prompt sentences like "school memories" or "sports day."

[0683] 4. Providing a generation script

[0684] The server uses a generative AI model to generate a manzai script based on themes and keywords provided by the user. The script is conversational and includes natural-looking interactions and humor that fit the specific theme.

[0685] 5. Script Distribution and Sharing

[0686] The server sends the generated script to the user's smartphone, and the user uses this script to practice their manzai. The script can also be shared on social media and video platforms via the smartphone app.

[0687] Examples of prompt statements

[0688] Theme: School Memories

[0689] Keywords: Sports day

[0690] This allows users to practice using the generated manzai script and share their performances with others, making it a useful way to widely promote the capabilities of the generative AI.

[0691] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0692] Step 1:

[0693] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is saved in text file format.

[0694] Step 2:

[0695] The server performs text cleansing (removing noise data), tokenization (dividing data into words and sentences), and normalization (converting data into a unified format) on the collected data. The preprocessed data is then formatted in a way that is suitable for training generative AI models.

[0696] Step 3:

[0697] The server uses the preprocessed data to train a generative AI model. Using deep learning techniques, the model effectively learns the linguistic characteristics and humorous elements of Japanese over multiple epochs. Once trained, the model becomes the basis for AI-based manzai script generation.

[0698] Step 4:

[0699] The user opens the smartphone app and inputs the theme or keywords of the manzai. For example, they input prompts such as "school memories" or "sports day," and the input data is sent from the app to the server. This input data becomes a request for the generative AI model.

[0700] Step 5:

[0701] The server receives themes and keywords submitted by users and uses a generative AI model to generate a manzai script, which includes natural-sounding interactions and humor that fit the specific theme.

[0702] Step 6:

[0703] The generated manzai script is sent from the server to the user's smartphone, and the user practices the manzai using the received script.

[0704] Step 7:

[0705] Users can share the manzai scripts they have practiced on social media and video platforms via a smartphone app, which also serves as a way to widely publicize the capabilities of the generative AI.

[0706] This allows a manzai script generation system using generative AI to embody a series of processes from data collection to providing and sharing the generated script.

[0707] 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.

[0708] The present invention provides a system that combines a generation AI and an emotion engine to generate a manzai script and provide it to a user. Specific embodiments are described below.

[0709] First, the server collects data to train the generative model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Methods of collection include web scraping, using APIs, and extracting data from existing databases.

[0710] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. At this stage, the data is formatted in a way that is suitable for training generative AI models.

[0711] The server then uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0712] The user inputs the theme and keywords of the manzai from the terminal, and the terminal sends this information to the server.

[0713] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh and episodes that they liked in the past, and reflects this in the generated script.

[0714] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During the generation process, natural dialogue is created, taking into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0715] The generated manzai script is sent from the server to the terminal, where the user can check the received script on the terminal and modify or regenerate it as necessary.

[0716] Finally, users practice using a manzai script provided by a generative AI using an emotion engine, which includes real-time updates and personalized elements based on past user emotional data.

[0717] As a specific example, if a user enters the theme "travel mishaps" and the keywords "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. Users can use this script to practice a stand-up comedy routine and post the process on social media or video platforms, thereby widely showcasing the capabilities of the generative AI and emotion engine.

[0718] As described above, the present invention combines a generation AI and an emotion engine to realize a system that generates manzai scripts that are more personalized for users and efficiently provides the results.

[0719] The processing flow will be explained below.

[0720] Step 1:

[0721] The server collects data to train the generative model. Specifically, it uses web scraping to obtain publicly available data on the Internet and APIs to extract data from existing databases. It also collects scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0722] Step 2:

[0723] The server preprocesses the collected data. At this stage, the data is first cleansed. Specifically, noise data and unnecessary information are removed. Next, the data is tokenized, and text is divided into words and sentences. The data is then normalized to standardize its format and prepare it in a format suitable for training a generative AI model.

[0724] Step 3:

[0725] The server uses the preprocessed data to train a generative AI model, typically using deep learning techniques. The model is trained over multiple epochs to learn the characteristics of the Japanese language and elements of humor.

[0726] Step 4:

[0727] The user inputs the theme and keywords of the manzai from the terminal. The user enters the content in the input field and clicks the "Send" button. This information is sent from the terminal to the server.

[0728] Step 5:

[0729] When the server receives themes and keywords provided by the user, the emotion engine is activated. The emotion engine collects the user's past emotional data and analyzes emotional changes and patterns. This allows it to understand the trends in manzai scripts that users prefer.

[0730] Step 6:

[0731] The server uses a generative AI model combined with an emotion engine to generate a manzai script based on themes and keywords provided by the user, and references the user's past emotional data to create an optimal script based on their emotions.

[0732] Step 7:

[0733] The server sends the generated manzai script to the terminal, which then sends the script to the user's terminal, where it is received and displayed.

[0734] Step 8:

[0735] The user can review the received manzai script on their device and modify or regenerate it as needed. The emotion engine recognizes the user's emotions in real time and updates the script accordingly.

[0736] Step 9:

[0737] Users practice manzai based on the provided manzai script. The emotion engine continuously monitors the user's reactions and adjusts the script content as necessary.

[0738] Step 10:

[0739] Finally, users can use the generated manzai script to perform an actual comedy routine, and by sharing the process and results on social media and video platforms, they can widely promote the capabilities of the generative AI and emotion engine.

[0740] For example, if a user inputs the theme "Work Failure Stories" and the keyword "Office," the server will refer to similar stories that the user liked in the past and use the emotion engine to generate a stand-up comedy script that the user can most easily relate to and that includes humorous elements. The user can then practice using this script and finally perform it, bringing laughter to the audience.

[0741] Example 2

[0742] 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."

[0743] Conventional manzai script generation systems have had difficulty generating personalized content that reflects individual users' emotions and preferences. They also lack the ability to generate natural dialogue that takes into account Japanese grammar and humor, resulting in a decline in the quality of the generated results. Furthermore, there was no system that reflected the user's emotions in real time during the generation process, making it difficult to increase user satisfaction.

[0744] 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 collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by a user, means for providing the generated manzai script to the user, and means for collecting user emotion data and adjusting the script content. This makes it possible to generate a high-quality manzai script personalized to the user's emotions and preferences.

[0745] "Data for training a generative model" refers to the information needed to train a generative AI model, including performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0746] "Means of collection" refers to the techniques and methods for effectively obtaining specific information, including web scraping, use of APIs, and extraction from existing databases.

[0747] "Preprocessing means" refers to a series of processes for converting collected data into an appropriate format, including data cleansing (removing noisy data), tokenization (dividing data), and normalization (standardizing data formats).

[0748] "Means for training a generative model" refers to techniques and methods for training a generative AI model using preprocessed data, including using deep learning techniques to train the model over a number of epochs.

[0749] "Means for generating a manzai script based on themes and keywords provided by the user" refers to a technology for automatically generating a manzai script based on user input, and the generated script includes natural dialogue that takes into account Japanese grammar and humor.

[0750] "Means for providing the generated manzai script to the user" refers to a technique or method for sending the generated manzai script to the user's terminal so that the user can check, modify, and regenerate it.

[0751] "Means for collecting user emotional data and adjusting script content" refers to technology that analyzes the user's past emotional data and adaptively changes the script content based on that data, using an emotion engine.

[0752] The present invention is a manzai script generation system that combines a generative AI model and an emotion engine, and provides users with personalized manzai scripts. Specific embodiments are described below.

[0753] First, the server collects data to train the generative AI model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Possible collection methods include web scraping, using APIs, and extracting data from existing databases. The hardware and software used include Cruise API and various web scraping tools.

[0754] Next, the server preprocesses the collected data. Preprocessing includes data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). At this stage, the data is formatted in a way that is suitable for training generative AI models. Specific software used includes the pandas library for data cleansing, SudachiPy for tokenization, and the nlp toolkit for normalization.

[0755] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques, possibly using frameworks like TensorFlow or PyTorch. The training process involves multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0756] When a user inputs a theme or keyword for a comedy routine, the device receives the information and sends it to the server. For example, if a user inputs the keywords "travel failure stories" and "family trip," the device sends this to the server.

[0757] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh in the past and their favorite episodes, and reflects this in the generated script. The emotion analysis technology used includes emotion recognition APIs and history learning technology.

[0758] The server then combines the generative AI model with an emotion engine to generate a manzai script based on themes and keywords entered by the user. The generation process creates natural dialogue that takes into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0759] The generated manzai script is sent from the server to the terminal. The user can check the received script on the terminal and modify or regenerate it as necessary. For example, if the user is not satisfied with part of the generated script, they can enter new keywords and regenerate it.

[0760] Finally, users practice using a manzai script provided by the generative AI using the emotion engine. The generated script includes real-time updates and personalized elements based on past user emotional data. By publishing the practiced script on social media and video platforms, users can widely demonstrate the capabilities of the generative AI and emotion engine.

[0761] As a specific example, if a user enters the theme "travel failure stories" and the keyword "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes incidents that occurred during the trip and funny interactions between family members. An example of a prompt for the generation AI model is, "Please generate a stand-up comedy script based on travel failure stories. The theme is travel failure stories, and the keyword is family trip."

[0762] As described above, the system of the present invention combines a generative AI model and an emotion engine to provide users with personalized manzai scripts and deliver the results efficiently.

[0763] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0764] Step 1: Data collection

[0765] The server collects data to train the generative AI model. The inputs include performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Specifically, it uses web scraping tools and APIs to obtain information from the internet and extracts the necessary data from existing databases. The output is the collected dataset.

[0766] Step 2: Data Preprocessing

[0767] The server preprocesses the collected data. The input is the collected dataset. Specifically, it first cleanses the data using the pandas library to remove noise. Then it splits the data into tokens using SudachiPy, and finally it normalizes the data format using the nlp toolkit. The output is the preprocessed dataset.

[0768] Step 3: Training the generative AI model

[0769] The server trains a generative AI model using the preprocessed data. The input is a preprocessed dataset. Specifically, it trains a deep learning model using TensorFlow or PyTorch. The model is trained over multiple epochs to learn the characteristics and humor of the Japanese language. The output is a trained generative AI model.

[0770] Step 4: Receiving User Input

[0771] The terminal receives input of the theme and keywords of the manzai from the user. The input includes the theme and keywords provided by the user. Specifically, when the user enters keywords such as "travel failure story" and "family trip" into the input field, the terminal sends this to the server. The output is the theme and keywords sent to the server.

[0772] Step 5: Collect and process emotion data

[0773] The server uses an emotion engine to collect and process the user's past emotional data. The input is the user's past emotional data. Specifically, it uses an emotion recognition API to collect scenes that made the user smile and favorite episodes from the past, and analyzes emotional patterns. The output is the analyzed emotional data.

[0774] Step 6: Generate the Manzai script

[0775] The server combines a generative AI model and an emotion engine to generate a manzai script based on the theme and keywords entered by the user. The inputs include the theme, keywords, and emotion data. Specifically, the generative AI model generates the script while taking into account Japanese language grammar and humor, and the emotion engine adjusts the script content in real time to reflect the user's emotion data. The output is the generated manzai script.

[0776] Step 7: Publish and check your Manzai script

[0777] The server sends the generated manzai script to the terminal. The input is the generated manzai script. Specifically, the server sends the script, the terminal receives it, and displays it to the user. The user checks the script and requests corrections or regeneration as necessary. The output is the manzai script checked on the terminal.

[0778] Step 8: Practice and share your scripts

[0779] The user practices based on the provided manzai script. The input is the confirmed manzai script. The specific operation is for the user to check the script, practice, record the process, and publish it on social media or a video platform. The output is the user's practice results and published content.

[0780] The above are the specific processing steps in the program of this system.

[0781] (Application example 2)

[0782] 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."

[0783] Conventional manzai script generation systems generate scripts uniformly without considering the user's emotions, resulting in a lack of personalization according to the preferences and emotions of individual users. Furthermore, they do not sufficiently consider the linguistic characteristics of the Japanese language, making it difficult to generate natural dialogue.

[0784] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by the user, means for providing the generated manzai script to the user, and means for recognizing the user's emotions and adjusting the content of the manzai script based on the emotions. This makes it possible to generate a personalized manzai script according to the user's emotional state.

[0785] "Means of collecting data for training a generative model" refers to means of collecting performance data from manzai competitions, literature on comedy, Japanese dialogue data, etc., using methods such as web scraping, using APIs, and extracting data from existing databases.

[0786] "Means for preprocessing collected data" refers to means for performing processes on collected data such as cleansing (removal of noise data), tokenization (dividing data), and normalization (standardization of data format).

[0787] "Means for training a generative model using preprocessed data" refers to means for training a generative AI model using deep learning techniques with preprocessed data as input.

[0788] "Means for generating a manzai script based on themes and keywords provided by the user" refers to means for generating a manzai script using a generative AI model based on themes and keywords entered by the user.

[0789] The "means for providing the generated manzai script to the user" is a means for transmitting the generated script to the terminal so that the user can check it.

[0790] "Means for recognizing the user's emotions and adjusting the content of the manzai script based on that" refers to means for analyzing the user's past emotional data and real-time emotional state, and using that information to adjust the content of the generated script as appropriate.

[0791] "Means for generating natural dialogue taking into account the linguistic characteristics of Japanese" refers to means for generating a manzai script as natural dialogue taking into account the grammar, humor, and linguistic characteristics unique to Japanese.

[0792] "Means including data on the performance of manzai competitions and literature on humor" refers to means of including data on the performance of manzai competitions and literature on comedy in the dataset required to train a generative model.

[0793] The present invention uses a system that combines a generative AI model and an emotion engine to generate and provide a personalized manzai script to a user. Specific embodiments of the present invention are described below.

[0794] First, the server collects data to train the generative model. The data collection methods include web scraping, API usage, and extraction from existing databases. The data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[0795] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. This prepares the collected data in a format suitable for learning generative AI models.

[0796] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques and is trained over multiple epochs to learn Japanese grammar and humorous elements. The training dataset includes performance data from manzai competitions and literature on humor.

[0797] The user inputs the theme and keywords of the manzai through the terminal. The terminal then sends this information to the server. At this point, the emotion engine is introduced. The emotion engine references the user's past emotional data and adjusts the content of the manzai script based on that. For example, it can learn episodes that made the user laugh in the past and favorite keywords, and reflect these in the generated script.

[0798] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During this generation process, natural dialogue is created taking into account the characteristics of the Japanese language. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0799] The generated manzai script is sent from the server to the device, where the user can review it. The user can review the script and modify or regenerate it as needed. Finally, the user can practice using the generated manzai script and post the results on social media or video platforms.

[0800] As a concrete example, consider the case where a user inputs the topic "Travel Mishaps" and the keyword "Family Trip." In this case, the emotion engine will refer to past stories that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. The following is an example of the prompt that generates the script:

[0801] You are an AI that generates funny stand-up scripts. In the past, users found the "Travel Mishaps" and "Family Trip" episodes particularly funny.

[0802] Create the following script:

[0803] Theme: Travel

[0804] Keywords: family trip, travel incident

[0805] Based on the script generated in this way, users can practice their manzai and widely share the process and results on social media and video platforms.

[0806] This allows for the generation of personalized manzai scripts according to the user's emotional state, improving the entertainment experience.

[0807] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0808] Step 1:

[0809] The server collects data to train the generative model. Specifically, it collects performance data from manzai competitions, literature on comedy, and Japanese dialogue data by web scraping, using APIs, and extracting data from existing databases. The input is this raw data, and the output is the collected dataset.

[0810] Step 2:

[0811] The server preprocesses the collected data. Specifically, it performs data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). The input is the collected dataset, and the output is the preprocessed data.

[0812] Step 3:

[0813] The server uses the preprocessed data to train a generative AI model. Specifically, it uses deep learning techniques to train the model to learn Japanese grammar and humorous elements. The input is the preprocessed data, and the output is a trained generative AI model.

[0814] Step 4:

[0815] The user inputs the theme and keywords of the manzai from the terminal. Specifically, the theme and keywords are entered in an input form and sent to the server. The input is the theme and keywords specified by the user, and the output is the request data sent to the server.

[0816] Step 5:

[0817] The server uses an emotion engine to recognize the user's emotional data. Specifically, it adjusts prompts by referring to the user's past laughing episodes and favorite keywords. The input is the user's emotional data, themes, and keywords, and the output is an adjusted prompt sentence.

[0818] Step 6:

[0819] The server uses the adjusted prompt sentence to ask the generative AI model to generate a manzai script. Specifically, the adjusted prompt sentence is input into the generative AI model to generate a script. The input is the adjusted prompt sentence, and the output is the generated manzai script.

[0820] Step 7:

[0821] The server sends the generated manzai script to the terminal. Specifically, it sends the generated script to the terminal so that the user can check it. The input is the generated manzai script, and the output is the script displayed on the user's terminal.

[0822] Step 8:

[0823] The user checks the received script and modifies or regenerates it as necessary. Specifically, the user checks the script on the terminal and sends instructions to the server to edit or regenerate it. The input is the generated script, and the output is the modified or newly generated script.

[0824] Step 9:

[0825] Finally, the user practices a manzai based on the generated script and publishes the results on social media or video platforms. Specifically, the user performs a manzai based on the script, records the process, and publishes it as a video. The input is the generated script, and the output is the published manzai video.

[0826] 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.

[0827] 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.

[0828] 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.

[0829] [Fourth embodiment]

[0830] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[0831] 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.

[0832] 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).

[0833] 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.

[0834] 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.

[0835] 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).

[0836] 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.

[0837] 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.

[0838] 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.

[0839] 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.

[0840] 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.

[0841] 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.

[0842] 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."

[0843] This invention provides a system that uses a generative AI to generate a manzai script and publicly demonstrates the AI's capabilities through the process and results. Specific embodiments are described below.

[0844] First, the server collects the data necessary to train the generative model. This can be done by web scraping, using APIs, or extracting data from existing databases. The collected data includes scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0845] The collected data is preprocessed by the server. Examples of preprocessing include text cleansing (removing noise data), tokenization (splitting into words or sentences), and normalization (conversion into a unified format), which prepares the data in a format suitable for training a generative model.

[0846] The server then trains a generative AI model based on the preprocessed data. Deep learning techniques are used for this purpose. The model is trained over multiple epochs to effectively capture the characteristics of the Japanese language and its humorous elements.

[0847] Users input the theme and keywords of the manzai using their device. Based on this input sent from the device to the server, the server generates a manzai script using the generated generative AI model. The generated script is a dialogue-style script that includes natural exchanges and humor in line with the specific theme.

[0848] Finally, the server sends the generated manzai script to the user's device. The user can practice the manzai based on the provided script and use it in actual performances. In addition, by sharing the process and results of practicing manzai, the user can demonstrate the performance of the generative AI to others.

[0849] For example, if a user inputs the keywords "pet topic" and "comedy element," the server will generate a manzai script based on this information, including funny pet-related and everyday anecdotes. The script is provided in an interactive format, including anecdotes such as "an incident that occurred while walking the dog" or "unexpected cat behavior." Users can use this script to practice manzai and aim to win the M-1 Grand Prix. Users can also share the capabilities of the generating AI by posting their practice process on social media or video platforms.

[0850] The processing flow will be explained below.

[0851] Step 1:

[0852] The server collects data to train the generative model, including scripts from past manzai competitions, literature on comedy, and Japanese dialogue data, through web scraping, API usage, and extraction from existing databases.

[0853] Step 2:

[0854] The server preprocesses the collected data, which includes text cleansing (removing noise data), tokenization (splitting into words and sentences), and normalization (converting into a unified format) to prepare the data in a format suitable for training a generative model.

[0855] Step 3:

[0856] The server uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0857] Step 4:

[0858] The user inputs the theme and keywords of the manzai using a terminal, which are then sent to the server.

[0859] Step 5:

[0860] The server uses the generated AI model to generate a manzai script based on themes and keywords provided by the user. This generation process creates natural dialogue that takes into account Japanese grammar and humor.

[0861] Step 6:

[0862] The generated manzai script is sent from the server to the user's device, where the user can view and check the received script.

[0863] Step 7:

[0864] The user practices a manzai based on the provided manzai script. If necessary, the user can modify the script content or input additional themes or keywords to generate the script again.

[0865] Step 8:

[0866] Ultimately, users can complete the manzai and use it in actual performances. In addition, by sharing the practice process and results on social media and video platforms, the performance and effectiveness of the generative AI can be widely promoted.

[0867] Example 1

[0868] 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."

[0869] Conventional text generation systems have had difficulty effectively incorporating the linguistic characteristics of Japanese and humorous elements when generating natural dialogue-style text based on specific themes or keywords. Furthermore, the generated text often fails to meet the user's intentions or expectations, resulting in scripts that lack entertainment value and practicality.

[0870] 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.

[0871] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a sentence generation script based on a theme or keywords provided by a user, and means for providing the generated sentence generation script to the user. This enables script generation in a natural, interactive format that takes into account the linguistic characteristics and humorous elements of the Japanese language.

[0872] A "generative model" refers to an algorithm that uses artificial intelligence techniques to generate output such as text or images based on specific input.

[0873] "Data collection" refers to the process of gathering the information needed to train a generative model from the internet or existing databases.

[0874] "Preprocessing" refers to the process of converting collected data into a format suitable for training a generative model by means of noise removal, tokenization, normalization, etc.

[0875] "User" refers to a person who uses the system to create a text generation script based on a specific theme or keyword.

[0876] "Themes and keywords" refer to important words and phrases that the user specifies to the system to determine the content and direction of the script to be generated.

[0877] A "text generation script" refers to a dialogue-style text created using a generative model based on a specific theme or keyword.

[0878] "Linguistic characteristics of Japanese" refers to characteristics that include the grammatical structure, expression methods, and cultural background that are unique to Japanese.

[0879] "Natural dialogue" refers to fluent, coherent conversation that is natural in human communication.

[0880] "Entertainment event performance data" refers to data on entertainment events that have been held in the past, and is the basic data used to generate high-quality scripts by training a generative model based on that data.

[0881] "Laughter literature" refers to research materials, books, papers, etc. on humor and comedy, and refers to the subjects used to train generative models.

[0882] MODE FOR CARRYING OUT THE INVENTION

[0883] This invention is a system that uses a generative AI model to create a sentence generation script based on a specific theme or keyword, and provides the results to the user.

[0884] Hardware and software used

[0885] The server collects data, preprocesses it, trains it, and generates scripts to train the generative model. Specifically, the following hardware and software are used:

[0886] Hardware: Computational servers with high-performance CPUs and GPUs, and large-capacity storage devices for storing data.

[0887] Software: The web scraping tool "BeautifulSoup" is used for data collection, the natural language processing library "MeCab" is used for data preprocessing, and the deep learning frameworks "TensorFlow" or "PyTorch" are used for training the generative model.

[0888] Data collection

[0889] The server uses web scraping tools, APIs, and existing databases to collect the data needed to train the generative model. This data includes performance data from entertainment events and literature on comedy, such as scripts from past manzai competitions, books on comedy, and papers.

[0890] Data Preprocessing

[0891] The collected data is pre-processed by the server. Specifically, the following processes are performed:

[0892] Text cleansing: Remove noise data and unnecessary characters.

[0893] Tokenization: Use "MeCab" to divide the data into words and sentences.

[0894] Normalization: Transforming data into a uniform format and making it consistent.

[0895] Training generative AI models

[0896] Based on the preprocessed data, the server trains a generative AI model using the deep learning frameworks TensorFlow or PyTorch. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0897] User input

[0898] The user uses a device to input themes and keywords into the generative AI model. For example, keywords such as "pet topics" or "comedy elements" are input. This input information is then sent from the device to the server.

[0899] Creating and providing sentence generation scripts

[0900] The server uses a generative AI model to generate a text generation script based on themes and keywords submitted by the user. This script is a natural dialogue based on a specific theme, incorporating Japanese linguistic characteristics and humorous elements.

[0901] The generated script is provided to the user's device from the server. The user can practice generating sentences based on the provided script and use it in actual performances.

[0902] Prompt Sentence Examples

[0903] The following example is one of the prompts that a user can input to a generative AI model.

[0904] Create a text generation script about pets, with a comedic twist. For example, include a story about something that happened while walking a dog or an unexpected cat behavior.

[0905] The above is a specific embodiment for carrying out the present invention.

[0906] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0907] Step 1:

[0908] The server collects data to train the generative model. This data includes performance data on entertainment events that incorporate Japanese linguistic characteristics and humorous elements, as well as literature on laughter. Specifically, web scraping tools, APIs, and existing databases are used. Web scraping tools such as "BeautifulSoup" are used to collect comedy scripts from the Internet. Comedy-related literature is obtained through APIs, and the necessary data is extracted from existing databases. The collected data is saved in a format suitable for training the generative model.

[0909] Step 2:

[0910] The server pre-processes the collected data, which includes the following specific steps:

[0911] Cleansing: Remove noise data, specifically, remove unnecessary characters and advertisements.

[0912] Tokenization: We use the natural language processing library "MeCab" to divide the data into words and sentences.

[0913] Normalization: Transforming data into a uniform format and making it consistent.

[0914] The input is the collected raw data, and the output is pre-processed data suitable for training a generative model, which is then cleansed, tokenized, and stored in a normalized format.

[0915] Step 3:

[0916] The server uses the preprocessed data to train a generative AI model. Using the deep learning frameworks "TensorFlow" or "PyTorch," the model is trained over multiple epochs. The input is the preprocessed data, and the output is the trained generative AI model. Specifically, the server supplies the training data to the model in batches and updates the model's parameters as learning progresses. The end result is a generative AI model that effectively captures linguistic characteristics and humorous elements.

[0917] Step 4:

[0918] The user uses a device to input specific themes and keywords into the generative AI model. This input is sent to the server through an input form on the device. For example, the user might input "pet topics" or "comedy elements." The input consists of themes and keywords, and this information sent to the server becomes the input for processing.

[0919] Step 5:

[0920] The server uses a generative AI model to generate a text generation script based on the theme and keywords sent by the user. The input is the theme and keywords, and the output is the generated text generation script. During the generation process, the generative AI model creates dialogue-style text that is optimal for the theme and keywords. For example, it generates dialogue-style text that includes specific humorous elements, such as "happenings that happen while walking the dog" or "unexpected cat behavior."

[0921] Step 6:

[0922] The server provides the generated sentence generation script to the user's device. The user practices sentence generation by reading the provided script aloud. The input is the generated script, and the output is the script displayed on the user's device. The user can practice using this script and use it in actual performances. By publishing the practice process on social media or video platforms, users can spread the performance and potential of generative AI to many people.

[0923] (Application example 1)

[0924] 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."

[0925] Conventional AI systems have limitations in the quality of the content they generate and the way it is provided to users, making it difficult to achieve satisfactory results, especially in the entertainment field. They also lack convenience in terms of how users utilize and share the generated content. Furthermore, there is a need for a method that is easy for users to use and that also allows them to demonstrate the performance of the generating AI.

[0926] 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.

[0927] In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, and means for training the generative model using the preprocessed data. This allows the server to generate a manzai script based on a theme or keywords provided by a user, and allows the user to practice and share the generated manzai script using a smartphone.

[0928] A "generative model" is an AI model that is trained to extract specific patterns and information from input data and perform problem solving or predictions.

[0929] "Means for collecting data" refers to the technical or methodological process for collecting the data required to train a generative model.

[0930] "Data pre-processing means" refers to the technical methods and processes used to process collected data in order to convert it into a form suitable for analysis or training.

[0931] "Generative model training" refers to the technical methods and processes that use collected and pre-processed data to train an AI model to perform a specific task.

[0932] "Means for generating manzai scripts based on themes and keywords" refers to a technical method or process for automatically generating dialogue-style text for manzai using user-provided topics and key words.

[0933] "Means for providing the generated manzai script to the user" refers to the technical methods and processes for transmitting the manzai script generated by AI to the user's device and making it viewable.

[0934] "Means that allow users to practice and share the generated manzai script using a smartphone" refers to the technical methods and processes that allow users to practice the generated manzai script using a smartphone app and share the content with other people.

[0935] This invention is a system that uses AI to generate manzai scripts and allows users to practice and share the scripts via smartphone. To realize this system, the following hardware and software are used.

[0936] Hardware

[0937] Server: A high-performance computer for back-end processing, which collects data, preprocesses the data, trains the generative model, and provides scripts to users.

[0938] Client device: A smartphone (iOS or Android). This device is used to provide the generated manzai script to the user.

[0939] software

[0940] Server side:

[0941] Programming language: Python

[0942] Deep Learning Framework: TensorFlow or PyTorch

[0943] Data processing libraries: NumPy, Pandas

[0944] Client side:

[0945] Smartphone app: Swift (iOS) or Kotlin (Android)

[0946] System Elements

[0947] 1. Data collection and preprocessing

[0948] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is preprocessed through text cleansing, tokenization, and normalization.

[0949] 2. Training the generative model

[0950] Based on the preprocessed data, a generative AI model is trained using deep learning techniques. The model is trained over multiple epochs to effectively learn the linguistic characteristics and humorous elements of Japanese.

[0951] 3. User input of themes and keywords

[0952] The user uses a smartphone app to input the theme or keywords of the manzai, such as prompt sentences like "school memories" or "sports day."

[0953] 4. Providing a generation script

[0954] The server uses a generative AI model to generate a manzai script based on themes and keywords provided by the user. The script is conversational and includes natural-looking interactions and humor that fit the specific theme.

[0955] 5. Script Distribution and Sharing

[0956] The server sends the generated script to the user's smartphone, and the user uses this script to practice their manzai. The script can also be shared on social media and video platforms via the smartphone app.

[0957] Examples of prompt statements

[0958] Theme: School Memories

[0959] Keywords: Sports day

[0960] This allows users to practice using the generated manzai script and share their performances with others, making it a useful way to widely promote the capabilities of the generative AI.

[0961] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0962] Step 1:

[0963] The server uses web scraping and APIs to collect data from the internet, such as manzai scripts, dialogue data, and literature on comedy. The collected data is saved in text file format.

[0964] Step 2:

[0965] The server performs text cleansing (removing noise data), tokenization (dividing data into words and sentences), and normalization (converting data into a unified format) on the collected data. The preprocessed data is then formatted in a way that is suitable for training generative AI models.

[0966] Step 3:

[0967] The server uses the preprocessed data to train a generative AI model. Using deep learning techniques, the model effectively learns the linguistic characteristics and humorous elements of Japanese over multiple epochs. Once trained, the model becomes the basis for AI-based manzai script generation.

[0968] Step 4:

[0969] The user opens the smartphone app and inputs the theme or keywords of the manzai. For example, they input prompts such as "school memories" or "sports day," and the input data is sent from the app to the server. This input data becomes a request for the generative AI model.

[0970] Step 5:

[0971] The server receives themes and keywords submitted by users and uses a generative AI model to generate a manzai script, which includes natural-sounding interactions and humor that fit the specific theme.

[0972] Step 6:

[0973] The generated manzai script is sent from the server to the user's smartphone, and the user practices the manzai using the received script.

[0974] Step 7:

[0975] Users can share the manzai scripts they have practiced on social media and video platforms via a smartphone app, which also serves as a way to widely publicize the capabilities of the generative AI.

[0976] This allows a manzai script generation system using generative AI to embody a series of processes from data collection to providing and sharing the generated script.

[0977] 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.

[0978] The present invention provides a system that combines a generation AI and an emotion engine to generate a manzai script and provide it to a user. Specific embodiments are described below.

[0979] First, the server collects data to train the generative model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Methods of collection include web scraping, using APIs, and extracting data from existing databases.

[0980] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. At this stage, the data is formatted in a way that is suitable for training generative AI models.

[0981] The server then uses the preprocessed data to train a generative AI model, which uses deep learning techniques and is trained over multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[0982] The user inputs the theme and keywords of the manzai from the terminal, and the terminal sends this information to the server.

[0983] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh and episodes that they liked in the past, and reflects this in the generated script.

[0984] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During the generation process, natural dialogue is created, taking into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[0985] The generated manzai script is sent from the server to the terminal, where the user can check the received script on the terminal and modify or regenerate it as necessary.

[0986] Finally, users practice using a manzai script provided by a generative AI using an emotion engine, which includes real-time updates and personalized elements based on past user emotional data.

[0987] As a specific example, if a user enters the theme "travel mishaps" and the keywords "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. Users can use this script to practice a stand-up comedy routine and post the process on social media or video platforms, thereby widely showcasing the capabilities of the generative AI and emotion engine.

[0988] As described above, the present invention combines a generation AI and an emotion engine to realize a system that generates manzai scripts that are more personalized for users and efficiently provides the results.

[0989] The processing flow will be explained below.

[0990] Step 1:

[0991] The server collects data to train the generative model. Specifically, it uses web scraping to obtain publicly available data on the Internet and APIs to extract data from existing databases. It also collects scripts from past manzai competitions, literature on comedy, and Japanese dialogue data.

[0992] Step 2:

[0993] The server preprocesses the collected data. At this stage, the data is first cleansed. Specifically, noise data and unnecessary information are removed. Next, the data is tokenized, and text is divided into words and sentences. The data is then normalized to standardize its format and prepare it in a format suitable for training a generative AI model.

[0994] Step 3:

[0995] The server uses the preprocessed data to train a generative AI model, typically using deep learning techniques. The model is trained over multiple epochs to learn the characteristics of the Japanese language and elements of humor.

[0996] Step 4:

[0997] The user inputs the theme and keywords of the manzai from the terminal. The user enters the content in the input field and clicks the "Send" button. This information is sent from the terminal to the server.

[0998] Step 5:

[0999] When the server receives themes and keywords provided by the user, the emotion engine is activated. The emotion engine collects the user's past emotional data and analyzes emotional changes and patterns. This allows it to understand the trends in manzai scripts that users prefer.

[1000] Step 6:

[1001] The server uses a generative AI model combined with an emotion engine to generate a manzai script based on themes and keywords provided by the user, and references the user's past emotional data to create an optimal script based on their emotions.

[1002] Step 7:

[1003] The server sends the generated manzai script to the terminal, which then sends the script to the user's terminal, where it is received and displayed.

[1004] Step 8:

[1005] The user can review the received manzai script on their device and modify or regenerate it as needed. The emotion engine recognizes the user's emotions in real time and updates the script accordingly.

[1006] Step 9:

[1007] Users practice manzai based on the provided manzai script. The emotion engine continuously monitors the user's reactions and adjusts the script content as necessary.

[1008] Step 10:

[1009] Finally, users can use the generated manzai script to perform an actual comedy routine, and by sharing the process and results on social media and video platforms, they can widely promote the capabilities of the generative AI and emotion engine.

[1010] For example, if a user inputs the theme "Work Failure Stories" and the keyword "Office," the server will refer to similar stories that the user liked in the past and use the emotion engine to generate a stand-up comedy script that the user can most easily relate to and that includes humorous elements. The user can then practice using this script and finally perform it, bringing laughter to the audience.

[1011] Example 2

[1012] 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."

[1013] Conventional manzai script generation systems have had difficulty generating personalized content that reflects individual users' emotions and preferences. They also lack the ability to generate natural dialogue that takes into account Japanese grammar and humor, resulting in a decline in the quality of the generated results. Furthermore, there was no system that reflected the user's emotions in real time during the generation process, making it difficult to increase user satisfaction.

[1014] 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 collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by a user, means for providing the generated manzai script to the user, and means for collecting user emotion data and adjusting the script content. This makes it possible to generate a high-quality manzai script personalized to the user's emotions and preferences.

[1015] "Data for training a generative model" refers to the information needed to train a generative AI model, including performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[1016] "Means of collection" refers to the techniques and methods for effectively obtaining specific information, including web scraping, use of APIs, and extraction from existing databases.

[1017] "Preprocessing means" refers to a series of processes for converting collected data into an appropriate format, including data cleansing (removing noisy data), tokenization (dividing data), and normalization (standardizing data formats).

[1018] "Means for training a generative model" refers to techniques and methods for training a generative AI model using preprocessed data, including using deep learning techniques to train the model over a number of epochs.

[1019] "Means for generating a manzai script based on themes and keywords provided by the user" refers to a technology for automatically generating a manzai script based on user input, and the generated script includes natural dialogue that takes into account Japanese grammar and humor.

[1020] "Means for providing the generated manzai script to the user" refers to a technique or method for sending the generated manzai script to the user's terminal so that the user can check, modify, and regenerate it.

[1021] "Means for collecting user emotional data and adjusting script content" refers to technology that analyzes the user's past emotional data and adaptively changes the script content based on that data, using an emotion engine.

[1022] The present invention is a manzai script generation system that combines a generative AI model and an emotion engine, and provides users with personalized manzai scripts. Specific embodiments are described below.

[1023] First, the server collects data to train the generative AI model. The collected data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Possible collection methods include web scraping, using APIs, and extracting data from existing databases. The hardware and software used include Cruise API and various web scraping tools.

[1024] Next, the server preprocesses the collected data. Preprocessing includes data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). At this stage, the data is formatted in a way that is suitable for training generative AI models. Specific software used includes the pandas library for data cleansing, SudachiPy for tokenization, and the nlp toolkit for normalization.

[1025] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques, possibly using frameworks like TensorFlow or PyTorch. The training process involves multiple epochs to learn the characteristics and humorous elements of the Japanese language.

[1026] When a user inputs a theme or keyword for a comedy routine, the device receives the information and sends it to the server. For example, if a user inputs the keywords "travel failure stories" and "family trip," the device sends this to the server.

[1027] This is where the emotion engine comes in. The emotion engine collects user emotional data and adjusts the content of the manzai script based on that data. For example, it learns scenes that made the user laugh in the past and their favorite episodes, and reflects this in the generated script. The emotion analysis technology used includes emotion recognition APIs and history learning technology.

[1028] The server then combines the generative AI model with an emotion engine to generate a manzai script based on themes and keywords entered by the user. The generation process creates natural dialogue that takes into account Japanese grammar and humor. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[1029] The generated manzai script is sent from the server to the terminal. The user can check the received script on the terminal and modify or regenerate it as necessary. For example, if the user is not satisfied with part of the generated script, they can enter new keywords and regenerate it.

[1030] Finally, users practice using a manzai script provided by the generative AI using the emotion engine. The generated script includes real-time updates and personalized elements based on past user emotional data. By publishing the practiced script on social media and video platforms, users can widely demonstrate the capabilities of the generative AI and emotion engine.

[1031] As a specific example, if a user enters the theme "travel failure stories" and the keyword "family trip," the emotion engine will refer to past episodes that made the user laugh and generate a script that includes incidents that occurred during the trip and funny interactions between family members. An example of a prompt for the generation AI model is, "Please generate a stand-up comedy script based on travel failure stories. The theme is travel failure stories, and the keyword is family trip."

[1032] As described above, the system of the present invention combines a generative AI model and an emotion engine to provide users with personalized manzai scripts and deliver the results efficiently.

[1033] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1034] Step 1: Data collection

[1035] The server collects data to train the generative AI model. The inputs include performance data from manzai competitions, literature on comedy, and Japanese dialogue data. Specifically, it uses web scraping tools and APIs to obtain information from the internet and extracts the necessary data from existing databases. The output is the collected dataset.

[1036] Step 2: Data Preprocessing

[1037] The server preprocesses the collected data. The input is the collected dataset. Specifically, it first cleanses the data using the pandas library to remove noise. Then it splits the data into tokens using SudachiPy, and finally it normalizes the data format using the nlp toolkit. The output is the preprocessed dataset.

[1038] Step 3: Training the generative AI model

[1039] The server trains a generative AI model using the preprocessed data. The input is a preprocessed dataset. Specifically, it trains a deep learning model using TensorFlow or PyTorch. The model is trained over multiple epochs to learn the characteristics and humor of the Japanese language. The output is a trained generative AI model.

[1040] Step 4: Receiving User Input

[1041] The terminal receives input of the theme and keywords of the manzai from the user. The input includes the theme and keywords provided by the user. Specifically, when the user enters keywords such as "travel failure story" and "family trip" into the input field, the terminal sends this to the server. The output is the theme and keywords sent to the server.

[1042] Step 5: Collect and process emotion data

[1043] The server uses an emotion engine to collect and process the user's past emotional data. The input is the user's past emotional data. Specifically, it uses an emotion recognition API to collect scenes that made the user smile and favorite episodes from the past, and analyzes emotional patterns. The output is the analyzed emotional data.

[1044] Step 6: Generate the Manzai script

[1045] The server combines a generative AI model and an emotion engine to generate a manzai script based on the theme and keywords entered by the user. The inputs include the theme, keywords, and emotion data. Specifically, the generative AI model generates the script while taking into account Japanese language grammar and humor, and the emotion engine adjusts the script content in real time to reflect the user's emotion data. The output is the generated manzai script.

[1046] Step 7: Publish and check your Manzai script

[1047] The server sends the generated manzai script to the terminal. The input is the generated manzai script. Specifically, the server sends the script, the terminal receives it, and displays it to the user. The user checks the script and requests corrections or regeneration as necessary. The output is the manzai script checked on the terminal.

[1048] Step 8: Practice and share your scripts

[1049] The user practices based on the provided manzai script. The input is the confirmed manzai script. The specific operation is for the user to check the script, practice, record the process, and publish it on social media or a video platform. The output is the user's practice results and published content.

[1050] The above are the specific processing steps in the program of this system.

[1051] (Application example 2)

[1052] 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."

[1053] Conventional manzai script generation systems generate scripts uniformly without considering the user's emotions, resulting in a lack of personalization according to the preferences and emotions of individual users. Furthermore, they do not sufficiently consider the linguistic characteristics of the Japanese language, making it difficult to generate natural dialogue.

[1054] The specification processing by the specification processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for collecting data for training a generative model, means for preprocessing the collected data, means for training the generative model using the preprocessed data, means for generating a manzai script based on a theme and keywords provided by the user, means for providing the generated manzai script to the user, and means for recognizing the user's emotions and adjusting the content of the manzai script based on the emotions. This makes it possible to generate a personalized manzai script according to the user's emotional state.

[1055] "Means of collecting data for training a generative model" refers to means of collecting performance data from manzai competitions, literature on comedy, Japanese dialogue data, etc., using methods such as web scraping, using APIs, and extracting data from existing databases.

[1056] "Means for preprocessing collected data" refers to means for performing processes on collected data such as cleansing (removal of noise data), tokenization (dividing data), and normalization (standardization of data format).

[1057] "Means for training a generative model using preprocessed data" refers to means for training a generative AI model using deep learning techniques with preprocessed data as input.

[1058] "Means for generating a manzai script based on themes and keywords provided by the user" refers to means for generating a manzai script using a generative AI model based on themes and keywords entered by the user.

[1059] The "means for providing the generated manzai script to the user" is a means for transmitting the generated script to the terminal so that the user can check it.

[1060] "Means for recognizing the user's emotions and adjusting the content of the manzai script based on that" refers to means for analyzing the user's past emotional data and real-time emotional state, and using that information to adjust the content of the generated script as appropriate.

[1061] "Means for generating natural dialogue taking into account the linguistic characteristics of Japanese" refers to means for generating a manzai script as natural dialogue taking into account the grammar, humor, and linguistic characteristics unique to Japanese.

[1062] "Means including data on the performance of manzai competitions and literature on humor" refers to means of including data on the performance of manzai competitions and literature on comedy in the dataset required to train a generative model.

[1063] The present invention uses a system that combines a generative AI model and an emotion engine to generate and provide a personalized manzai script to a user. Specific embodiments of the present invention are described below.

[1064] First, the server collects data to train the generative model. The data collection methods include web scraping, API usage, and extraction from existing databases. The data includes performance data from manzai competitions, literature on comedy, and Japanese dialogue data.

[1065] The collected data is preprocessed on the server, which includes data cleansing (removing noise data), tokenization (splitting data), normalization (standardizing data format), etc. This prepares the collected data in a format suitable for learning generative AI models.

[1066] The server then uses the preprocessed data to train a generative AI model. The generative model uses deep learning techniques and is trained over multiple epochs to learn Japanese grammar and humorous elements. The training dataset includes performance data from manzai competitions and literature on humor.

[1067] The user inputs the theme and keywords of the manzai through the terminal. The terminal then sends this information to the server. At this point, the emotion engine is introduced. The emotion engine references the user's past emotional data and adjusts the content of the manzai script based on that. For example, it can learn episodes that made the user laugh in the past and favorite keywords, and reflect these in the generated script.

[1068] The server combines a generative AI model and an emotion engine to generate a manzai script based on themes and keywords entered by the user. During this generation process, natural dialogue is created taking into account the characteristics of the Japanese language. The emotion engine also recognizes the user's emotions in real time and updates the script accordingly.

[1069] The generated manzai script is sent from the server to the device, where the user can review it. The user can review the script and modify or regenerate it as needed. Finally, the user can practice using the generated manzai script and post the results on social media or video platforms.

[1070] As a concrete example, consider the case where a user inputs the topic "Travel Mishaps" and the keyword "Family Trip." In this case, the emotion engine will refer to past stories that made the user laugh and generate a script that includes funny family interactions and mishaps that occurred during the trip. The following is an example of the prompt that generates the script:

[1071] You are an AI that generates funny stand-up scripts. In the past, users found the "Travel Mishaps" and "Family Trip" episodes particularly funny.

[1072] Create the following script:

[1073] Theme: Travel

[1074] Keywords: family trip, travel incident

[1075] Based on the script generated in this way, users can practice their manzai and widely share the process and results on social media and video platforms.

[1076] This allows for the generation of personalized manzai scripts according to the user's emotional state, improving the entertainment experience.

[1077] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1078] Step 1:

[1079] The server collects data to train the generative model. Specifically, it collects performance data from manzai competitions, literature on comedy, and Japanese dialogue data by web scraping, using APIs, and extracting data from existing databases. The input is this raw data, and the output is the collected dataset.

[1080] Step 2:

[1081] The server preprocesses the collected data. Specifically, it performs data cleansing (removing noise data), tokenization (splitting data), and normalization (standardizing data format). The input is the collected dataset, and the output is the preprocessed data.

[1082] Step 3:

[1083] The server uses the preprocessed data to train a generative AI model. Specifically, it uses deep learning techniques to train the model to learn Japanese grammar and humorous elements. The input is the preprocessed data, and the output is a trained generative AI model.

[1084] Step 4:

[1085] The user inputs the theme and keywords of the manzai from the terminal. Specifically, the theme and keywords are entered in an input form and sent to the server. The input is the theme and keywords specified by the user, and the output is the request data sent to the server.

[1086] Step 5:

[1087] The server uses an emotion engine to recognize the user's emotional data. Specifically, it adjusts prompts by referring to the user's past laughing episodes and favorite keywords. The input is the user's emotional data, themes, and keywords, and the output is an adjusted prompt sentence.

[1088] Step 6:

[1089] The server uses the adjusted prompt sentence to ask the generative AI model to generate a manzai script. Specifically, the adjusted prompt sentence is input into the generative AI model to generate a script. The input is the adjusted prompt sentence, and the output is the generated manzai script.

[1090] Step 7:

[1091] The server sends the generated manzai script to the terminal. Specifically, it sends the generated script to the terminal so that the user can check it. The input is the generated manzai script, and the output is the script displayed on the user's terminal.

[1092] Step 8:

[1093] The user checks the received script and modifies or regenerates it as necessary. Specifically, the user checks the script on the terminal and sends instructions to the server to edit or regenerate it. The input is the generated script, and the output is the modified or newly generated script.

[1094] Step 9:

[1095] Finally, the user practices a manzai based on the generated script and publishes the results on social media or video platforms. Specifically, the user performs a manzai based on the script, records the process, and publishes it as a video. The input is the generated script, and the output is the published manzai video.

[1096] 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.

[1097] 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.

[1098] 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.

[1099] 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.

[1100] 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.

[1101] 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.

[1102] 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).

[1103] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[1104] 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."

[1105] 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.

[1106] 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).

[1107] 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.

[1108] 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.

[1109] 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.

[1110] 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.

[1111] 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.

[1112] 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.

[1113] 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.

[1114] 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.

[1115] 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.

[1116] 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.

[1117] The following is further disclosed regarding the above embodiment.

[1118] (Claim 1)

[1119] a means for collecting data for training a generative model;

[1120] means for pre-processing the collected data;

[1121] means for training a generative model using the preprocessed data;

[1122] means for generating a manzai script based on a theme or keywords provided by a user;

[1123] a means for providing the generated comic script to a user;

[1124] A system including:

[1125] (Claim 2)

[1126] 2. The system according to claim 1, further comprising means for generating natural dialogue by taking into consideration the linguistic characteristics of the Japanese language when generating a manzai script.

[1127] (Claim 3)

[1128] The system of claim 1, further comprising means for including performance data of manzai competitions and literature on humor in the dataset from which the generative model is trained.

[1129] "Example 1"

[1130] (Claim 1)

[1131] a means for collecting data for training a generative model;

[1132] means for pre-processing the collected data;

[1133] means for training a generative model using the preprocessed data;

[1134] means for generating a text generation script based on themes and keywords provided by a user;

[1135] a means for providing the generated sentence generation script to a user;

[1136] A system including:

[1137] (Claim 2)

[1138] 2. The system according to claim 1, further comprising means for generating natural dialogue taking into account the linguistic characteristics of Japanese.

[1139] (Claim 3)

[1140] 10. The system of claim 1, further comprising means for including performance data of entertainment events and literature on laughter in the dataset from which the generative model is trained.

[1141] "Application Example 1"

[1142] (Claim 1)

[1143] a means for collecting data for training a generative model;

[1144] means for pre-processing the collected data;

[1145] means for training a generative model using the preprocessed data;

[1146] means for generating a manzai script based on a theme or keywords provided by a user;

[1147] a means for providing the generated comic script to a user;

[1148] A means for allowing a user to practice and share the generated manzai script using a smartphone;

[1149] A system including:

[1150] (Claim 2)

[1151] 2. The system according to claim 1, further comprising means for generating natural dialogue by taking into consideration the linguistic characteristics of the Japanese language when generating a manzai script.

[1152] (Claim 3)

[1153] The system of claim 1, further comprising means for including performance data of manzai competitions and literature on humor in the dataset from which the generative model is trained.

[1154] "Example 2: Combining Emotion Engines"

[1155] (Claim 1)

[1156] a means for collecting data for training a generative model;

[1157] means for pre-processing the collected data;

[1158] means for training a generative model using the preprocessed data;

[1159] means for generating a manzai script based on a theme or keywords provided by a user;

[1160] a means for providing the generated comic script to a user;

[1161] A means for collecting user emotion data and adjusting script content;

[1162] A system including:

[1163] (Claim 2)

[1164] 2. The system according to claim 1, further comprising means for generating natural dialogue by taking into consideration the linguistic characteristics of the Japanese language when generating a manzai script.

[1165] (Claim 3)

[1166] The system of claim 1, further comprising means for including performance data of manzai competitions and literature on humor in the dataset from which the generative model is trained.

[1167] "Application example 2 when combining emotion engines"

[1168] (Claim 1)

[1169] a means for collecting data for training a generative model;

[1170] means for pre-processing the collected data;

[1171] means for training a generative model using the preprocessed data;

[1172] means for generating a manzai script based on a theme or keywords provided by a user;

[1173] a means for providing the generated comic script to a user;

[1174] means for recognizing a user's emotions and adjusting the content of the comic script based thereon;

[1175] A system including:

[1176] (Claim 2)

[1177] 2. The system according to claim 1, further comprising means for generating natural dialogue by taking into consideration the linguistic characteristics of the Japanese language when generating a manzai script.

[1178] (Claim 3)

[1179] The system of claim 1, further comprising means for including performance data of manzai competitions and literature on humor in the dataset from which the generative model is trained. [Explanation of symbols]

[1180] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. a means for collecting data for training a generative model; means for pre-processing the collected data; means for training a generative model using the preprocessed data; means for generating a manzai script based on a theme or keywords provided by a user; a means for providing the generated comic script to a user; A system including:

2. 2. The system according to claim 1, further comprising means for generating natural dialogue by taking into consideration the linguistic characteristics of the Japanese language when generating a manzai script.

3. The system of claim 1 , further comprising means for including performance data of stand-up comedy competitions and literature on humor in the dataset from which the generative model is trained.

Images