system

Abstract

Provide a system. [Solution] means for receiving a prompt from a user; means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt; means for transmitting the sub-prompts to respective artificial intelligence models; means for receiving results from said artificial intelligence model; a means for integrating the results received; and means for returning the consolidated results to the 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] Currently, it is difficult for users to effectively use multiple AI models in a coordinated manner. This requires users to go through complex processes to combine the strengths of different AI models to achieve consistent results. Furthermore, there is no efficient method for analyzing prompts and independently processing multiple tasks based on them. Therefore, there is a need for a system that allows users to easily use multiple AI models and obtain integrated results. [Means for solving the problem]

[0005] In order to solve the above-mentioned problems, the present invention provides the following means: a system including means for receiving a prompt from a user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, and means for returning the integrated results to the user. This allows a user to use multiple AI models in seamless cooperation through a single interface, and to obtain efficiently integrated results.

[0006] A "prompt" is input information such as an instruction or question given by a user to a system.

[0007] The "means for receiving a prompt" is a function that allows the system to obtain a prompt input by the user.

[0008] The "analysis means" is a function for analyzing the received prompt, understanding its contents, and dividing it into appropriate tasks.

[0009] A "subprompt" is a direction or question derived from the original prompt that has been split off by a parsing tool.

[0010] An "artificial intelligence model" is a machine learning model trained to perform a specific task.

[0011] The "means for sending to an artificial intelligence model" is a function for sending a sub-prompt to an appropriate artificial intelligence model.

[0012] "Means for receiving results" is a function for obtaining the results output from the artificial intelligence model.

[0013] "Means for integrating" is a function for combining results from multiple artificial intelligence models into a single unified result.

[0014] The "means for returning results" is a function for returning the integrated results to the user. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0016] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

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

[0018] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0019] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0020] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0021] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0023] [First embodiment]

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

[0025] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0028] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

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

[0032] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0033] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0034] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0035] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0036] The present invention provides a system that allows a user to easily order a primary prompt and coordinates the prompt across multiple artificial intelligence models, including means for receiving a prompt from a user, means for parsing the prompt and dividing it into multiple sub-prompts, means for sending the sub-prompts to respective artificial intelligence models, means for receiving results from the artificial intelligence models, means for aggregating the results, and means for returning the aggregated results to the user.

[0037] Program processing

[0038] server

[0039] 1. Accepting the prompt

[0040] The server receives the prompt sent from the user terminal.

[0041] Example: A user enters the prompt "Generate an image of a beautiful landscape and add a description."

[0042] 2. Parsing the prompt

[0043] The server analyzes the received prompts and splits them into different tasks (e.g., image generation, description generation).

[0044] 3. Generating a subprompt

[0045] The server generates sub-prompts for each task based on the analysis results.

[0046] Example: Generate a sub-prompt for "Generate an image of a beautiful landscape" and another for "Generate a description of the image."

[0047] 4. Sending a sub-prompt

[0048] The server sends the generated sub-prompt to the appropriate artificial intelligence model, for example, sending a sub-prompt such as "Generate an image of a beautiful landscape" to an image generation model, and a sub-prompt such as "Generate a description of the image" to a text generation model.

[0049] 5. Acceptance of Results

[0050] The server receives the output results from each artificial intelligence model.

[0051] Example: Receive a landscape image generated from an image generation model and a description generated from a text generation model.

[0052] 6. Synthesis of results

[0053] The server combines the received images and descriptions into a single consolidated result.

[0054] Example: The integrated results include a landscape image and its description.

[0055] 7. Return of results

[0056] The server transmits the integrated results to the user terminal.

[0057] User terminal

[0058] 1. Enter the prompt

[0059] The user enters a prompt into the interface.

[0060] Example: User types "Generate an image of a beautiful landscape and add a description."

[0061] 2. Sending a prompt

[0062] The user terminal sends the entered prompt to the server.

[0063] 3. Displaying the results

[0064] The results returned by the server are displayed to the user.

[0065] Example: A beautiful landscape image and its description are displayed.

[0066] Specific examples

[0067] Example 1: Image and description generation process

[0068] The user enters the prompt "Create an image of a beautiful landscape and add a description."

[0069] The user terminal sends a prompt to the server.

[0070] The server receives the prompt and splits it into an image generation task and a description generation task.

[0071] The server sends a sub-prompt to the image generation model saying, "Generate an image of a beautiful landscape," and a sub-prompt to the text generation model saying, "Generate a description of the image."

[0072] The server receives the landscape image generated from the image generation model and the explanatory text generated from the text generation model.

[0073] The server aggregates these results and sends them to the user terminal.

[0074] The user terminal displays the received landscape image and explanatory text on the interface.

[0075] The user checks the generated image and description.

[0076] This invention provides a system that allows users to seamlessly utilize multiple artificial intelligence models and efficiently process complex tasks, enabling users to easily obtain high-quality results.

[0077] The processing flow will be explained below.

[0078] Step 1:

[0079] The user enters a prompt into the interface.

[0080] For example, type "Generate an image of a beautiful landscape and add a description."

[0081] Step 2:

[0082] The terminal sends the prompt entered by the user to the server.

[0083] For example: A prompt is sent to the server.

[0084] Step 3:

[0085] The server receives the prompt.

[0086] Example: Accept the prompt "Generate an image of a beautiful landscape and add a description."

[0087] Step 4:

[0088] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[0089] Example: Split into tasks "image generation" and "description generation".

[0090] Step 5:

[0091] The server generates sub-prompts corresponding to each task.

[0092] Example: Generate the sub-prompts "Generate an image of a beautiful landscape" and "Generate a description of the generated image."

[0093] Step 6:

[0094] The server sends the generated sub-prompts to the selected artificial intelligence model.

[0095] Example: Send an image generation model the sub-prompt "Generate an image of a beautiful landscape," and a text generation model the sub-prompt "Generate a description of the generated image."

[0096] Step 7:

[0097] The server receives the output results from each artificial intelligence model.

[0098] Example: Accepts beautiful landscape images generated from an image generation model and explanatory text generated from a text generation model.

[0099] Step 8:

[0100] The server consolidates the results it receives.

[0101] Example: A landscape image and its description are packaged together as a single data package.

[0102] Step 9:

[0103] The server transmits the integrated results to the user terminal.

[0104] Example: Sending a synthesized landscape image and its description to a user device.

[0105] Step 10:

[0106] The terminal displays the results received from the server to the user.

[0107] Example: A user interface displays an image of a beautiful landscape and a description of it.

[0108] Step 11:

[0109] The user checks the displayed results.

[0110] Example: View and evaluate the generated landscape images and their descriptions.

[0111] Example 1

[0112] 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."

[0113] In today's world, users are required to efficiently process complex tasks using multiple AI models, but existing systems lack a simple way to achieve this. A method is needed that allows users to seamlessly execute different tasks, such as image generation and text generation, within a single system, and then integrate and efficiently use the results.

[0114] 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.

[0115] In this invention, the server includes means for receiving a prompt sentence from a user terminal, means for analyzing the received prompt sentence and dividing it into multiple sub-prompt sentences based on the prompt sentence, means for sending the sub-prompt sentences to each generative AI model, means for receiving results from the generative AI models, means for integrating the received results, and means for returning the integrated results to the user terminal, thereby enabling a user to efficiently process complex tasks using multiple AI models and easily obtain the integrated results.

[0116] A "prompt" is a textual instruction for a user to enter a specific task or request.

[0117] A "user terminal" is a device (e.g., a computer, smartphone, tablet, etc.) that a user uses to input prompt sentences and communicate with a server.

[0118] The "server" is a central computer system that receives prompts, analyzes them, distributes tasks to generative AI models, and synthesizes and returns the results to the user device.

[0119] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task (e.g., image generation or text generation).

[0120] An "image generation model" is an artificial intelligence model that is trained to generate images based on input prompts.

[0121] A "text generation model" is an artificial intelligence model that is trained to generate natural language text based on an input prompt.

[0122] The "means for parsing a prompt sentence" is a software component that has the function of parsing a prompt sentence and dividing it into multiple sub-prompt sentences.

[0123] A "sub-prompt sentence" is a detailed instruction for a specific task that is extracted as a result of parsing a prompt sentence.

[0124] A "means for aggregating results" is a software component that has the ability to combine individual results from a generative AI model into a single aggregated result.

[0125] The present invention provides a system that allows users to easily input prompt sentences and process complex tasks using multiple generative AI models. This system consists of a server and a user terminal.

[0126] server

[0127] The server has the functions of receiving prompt sentences from the user device, analyzing the received prompt sentences and dividing them into multiple sub-prompts, sending the sub-prompts to the appropriate generative AI model, and receiving the results from the generative AI model, integrating them, and sending them back to the user device.

[0128] User terminal

[0129] The user terminal provides an interface for the user to input a prompt sentence, which is then sent to the server, which processes the sentence and displays the result.

[0130] Specific implementation example

[0131] Entering a prompt statement

[0132] The user inputs a prompt sentence into the interface of the user terminal, such as "Create an image of a beautiful landscape and add a description of it." This prompt sentence is sent to the server as an HTTP request.

[0133] Parsing the prompt statement

[0134] The server analyzes the received prompt sentence. It uses natural language processing (NLP) technology to divide the prompt sentence into an image generation task and a text generation task. Specifically, it generates the following sub-prompt sentences:

[0135] Image generation subprompt: "Generate an image of a beautiful landscape."

[0136] Text generation subprompt: "Generate a description for that image."

[0137] Sending a subprompt statement

[0138] The server sends the image generation subprompt to the image generation model and the text generation subprompt to the text generation model. These models are generative AI models trained specifically for each task. The server sends the subprompt sentence as an API request and waits for a response from the model.

[0139] Acceptance and integration of results

[0140] The server receives the results from the generative AI models. For example, an image generation model might return an image of a beautiful landscape, and a text generation model might return a description of that image. It then combines these results and stores them in a single data format (e.g., JSON).

[0141] Returning the results

[0142] The server sends the integrated results to the user's device, which then displays the results on its interface. The user can then view the beautiful landscape images and their descriptions.

[0143] This system allows users to efficiently process complex tasks using multiple generative AI models and easily obtain integrated results. By clarifying what prompts will produce what results, the user experience is greatly improved.

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

[0145] Step 1:

[0146] Entering a prompt statement

[0147] The user inputs a prompt sentence through the interface of the user terminal.

[0148] Input: The prompt text entered by the user (e.g., "Generate an image of a beautiful landscape and add a description.")

[0149] Output: The user terminal generates an HTTP request to send the prompt text to the server.

[0150] Step 2:

[0151] Receiving a prompt

[0152] The server receives the prompt sentence sent from the user terminal.

[0153] Input: Prompt sent from the user's terminal to the server

[0154] Output: Stores the prompt text in an internal data structure.

[0155] Step 3:

[0156] Parsing the prompt statement

[0157] The server analyzes the received prompt sentence using natural language processing technology and divides it into multiple sub-prompt sentences.

[0158] Input: The received prompt (e.g., "Generate an image of a beautiful landscape and add a description.")

[0159] Output: Split sub-prompt sentences (e.g. "Generate an image of a beautiful landscape", "Generate a description of the image")

[0160] Step 4:

[0161] Sending a subprompt statement

[0162] The server sends the split subprompt sentences to the appropriate generative AI model. For example, an image generation subprompt is sent to an image generation model, and an explanatory sentence generation subprompt is sent to a text generation model.

[0163] Input: Separated subprompt sentence

[0164] Output: API request to the generative AI model

[0165] Step 5:

[0166] Acceptance of results

[0167] The server receives the results returned by the generative AI models: generated images from the image generation model, and generated explanatory text from the text generation model.

[0168] Input: Responses from each generative AI model (e.g., generated landscape images, generated descriptions)

[0169] Output: Stores the resulting data in an internal data structure.

[0170] Step 6:

[0171] Integration of results

[0172] The server then combines the results it receives, for example combining images and descriptions into a single data format (e.g., a JSON object).

[0173] Input: Accepted result data (e.g., generated landscape image, generated description)

[0174] Output: Consolidated results (e.g., a JSON object containing image URLs and descriptions)

[0175] Step 7:

[0176] Returning the results

[0177] The server returns the consolidated results to the user terminal.

[0178] Input: Consolidated results (e.g., JSON format data)

[0179] Output: HTTP response to the user's device

[0180] Step 8:

[0181] Displaying the results

[0182] The user terminal receives the results returned from the server and displays them to the user.

[0183] Input: Integration results sent from the server (e.g., JSON format data)

[0184] Output: The result shown to the user (e.g., a landscape image and its description)

[0185] Through the above process, users can efficiently obtain the results of complex tasks that go through multiple generative AI models.

[0186] (Application example 1)

[0187] 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."

[0188] Conventional customer support systems for virtual stores have been limited to using individual models or simple FAQ databases to answer user questions, making it difficult to provide integrated, advanced answers. As a result, there has been a problem in that it takes a lot of time and effort for users to obtain the information they are looking for.

[0189] 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.

[0190] In this invention, the server includes means for receiving a prompt from a user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, means for returning the integrated results to the user, and means for analyzing questions about products in a virtual store and providing related information, thereby enabling the user to obtain detailed and comprehensive information about products by inputting a prompt once.

[0191] "Means for receiving prompts from a user" refers to a device or software function that allows the system to receive questions or requests entered by a user.

[0192] The "means for analyzing a prompt and dividing it into a plurality of sub-prompts based on the prompt" refers to the function of a device or software that understands a received prompt and breaks it down into a plurality of specific tasks.

[0193] The "means for sending sub-prompts to respective artificial intelligence models" refers to a device or software function that sends the divided sub-prompts to the appropriate artificial intelligence models so that they can be processed.

[0194] "Means for receiving results from an artificial intelligence model" refers to a device or software function for receiving results generated by an artificial intelligence model.

[0195] "Means for integrating received results" refers to the function of a device or software that combines the results received from each artificial intelligence model into a single, consistent result.

[0196] A "means for returning the consolidated results to the user" is a device or software function that displays or presents the consolidated results to the user.

[0197] "Means for analyzing product-related questions within a virtual store and providing related information" refers to a device or software function for analyzing product-related questions entered by users in a virtual store and providing related information in response to those questions.

[0198] An "image generation model" is a type of artificial intelligence that generates images based on input text prompts.

[0199] A "sentence generation model" is a type of artificial intelligence that generates sentences based on input text prompts.

[0200] A "review analysis model" is a type of artificial intelligence that analyzes product reviews and summarizes or evaluates their content.

[0201] The present invention is a system that allows users to easily order key prompts and coordinates prompts across multiple artificial intelligence models. This system aims to enable multiple artificial intelligence models to cooperate and generate answers to questions about products asked by users, particularly in virtual stores.

[0202] The main components of the system are as follows:

[0203] 1. Accepting the prompt

[0204] The user inputs a question about a product in the virtual store. For example, the user inputs a prompt such as "How is the camera performance of this smartphone?" This prompt is sent from the user terminal to the server.

[0205] 2. Parsing the prompt

[0206] The server parses the received prompt and divides it into multiple sub-prompts based on different tasks, for example, "analyzing image samples from a camera," "presenting technical specifications of a camera," and "summarizing user reviews."

[0207] 3. Generate and send a subprompt

[0208] Based on the analysis results, the server generates sub-prompts appropriate for each task and sends them to each AI model, such as "Generate an image sample from the camera," "Explain the camera's technical specifications," or "Summarize user reviews about the camera's performance."

[0209] 4. Acceptance of Results

[0210] The server receives the output results from each artificial intelligence model (image generation model, sentence generation model, review analysis model).

[0211] 5. Synthesis of results

[0212] The server then combines the different results it receives to produce a single, coherent answer, which includes image samples of the camera's capabilities, detailed technical specifications, and highlights from user reviews.

[0213] 6. Return of results

[0214] Finally, the server sends the integrated results back to the user device, which then displays them to the user. For example, in response to a prompt entered by the user, a comprehensive answer such as "This smartphone's camera performance is excellent, and it takes clear photos even at night. In addition, many users rate its ease of use highly."

[0215] Program processing explanation

[0216] The system for implementing the present invention uses the following hardware and software. User terminals include smartphones, tablets, and PCs, and the server uses software such as Flask (a Python web framework), OpenAI (registered trademark) API, PyTorch, and Hugging Face Transformers. Users enter questions about products in the virtual store, and the prompts are sent to the server.

[0217] The server analyzes the prompts and sends sub-prompts to the appropriate artificial intelligence models. The image generation model generates images based on the input text prompt, the sentence generation model generates sentences based on the text prompt, and the review analysis model analyzes product reviews and summarizes or rates their content. The server consolidates these results and returns them to the user.

[0218] Specific examples

[0219] For example, a user enters the prompt "How good is the camera on this phone?" The prompt is split into:

[0220] "Generate a sample image from the camera"

[0221] "Please explain the technical specifications of the camera."

[0222] "Please summarize user reviews of the camera's performance."

[0223] The results of the generative AI model based on these sub-prompts are integrated and sent back to the user, allowing the user to obtain comprehensive information with a single prompt input.

[0224] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0225] Step 1:

[0226] A user enters a question about a product in a virtual store

[0227] A user inputs a prompt such as "How is the camera performance of this smartphone?" into the user terminal and sends the question to the server. The input is a text prompt, and the output is the prompt data transferred to the server.

[0228] Step 2:

[0229] The server receives a prompt

[0230] The server receives prompts sent from the user terminal and stores them as input data for parsing. The input is the raw prompt text sent by the user, and the output is the prompt to be parsed.

[0231] Step 3:

[0232] The server parses the prompt and splits it into subprompts

[0233] The server analyzes the received prompt and divides it into multiple sub-prompts based on different tasks, such as "analyzing camera image samples," "presenting camera specifications," and "summarizing user reviews." The input is the prompt text, and the output is a set of sub-prompts corresponding to each task.

[0234] Step 4:

[0235] Send sub-prompts to each AI model

[0236] The server sends the generated sub-prompts to each AI model. For example, it sends a sub-prompt such as "Generate an image sample of the camera" to the image generation model, and a sub-prompt such as "Describe the technical specifications of the camera" to the sentence generation model. The input is the sub-prompt, and the output is the task sent to each model.

[0237] Step 5:

[0238] The server receives the results from each AI model.

[0239] The server receives the output from each AI model. The image generation model generates sample images of the camera, the sentence generation model generates technical specifications for the camera, and the review analysis model generates summaries of user reviews. The input is the generated results based on the sub-prompts, and the output is the resulting data from each model.

[0240] Step 6:

[0241] The server aggregates the results

[0242] The server then combines the results it receives to generate a single, coherent answer. For example, it might combine the generated camera image, technical specifications, and review summary into a single response. The input is the result data from each model, and the output is the combined answer.

[0243] Step 7:

[0244] The server sends the consolidated results back to the user

[0245] The server sends the integrated results to the user's device. The user's device displays the results to the user, allowing the user to obtain comprehensive information about the product. For example, the user's device might display a message such as, "This smartphone's camera performance is excellent, and it takes clear photos even at night. Many users also rate its ease of use highly." The input is the integrated result data, and the output is the content displayed on the user's device.

[0246] 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.

[0247] This invention provides a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and influences the analysis results.

[0248] Program processing

[0249] server

[0250] 1. Accepting the prompt

[0251] The server receives the prompt sent from the user terminal.

[0252] Example: A user enters the prompt "Generate a poem to comfort me when I'm sad."

[0253] 2. Emotion Analysis

[0254] The server is equipped with an emotion engine that analyzes text data from the user's input prompts to recognize the user's emotions.

[0255] Example: The emotion "sad" is recognized from the input prompt.

[0256] 3. Parsing the prompt

[0257] The server parses the received prompt, understands its content and splits it into appropriate tasks.

[0258] Example: Identify the task "Generate a poem."

[0259] 4. Generating a subprompt

[0260] The server generates sub-prompts for each task based on the analysis results, taking into account the analysis results of the emotion engine.

[0261] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[0262] 5. Sending a sub-prompt

[0263] The server sends the generated sub-prompts to the appropriate artificial intelligence model.

[0264] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[0265] 6. Acceptance of Results

[0266] The server receives the output from the artificial intelligence model.

[0267] Example: Accepting a poem generated from a poetry generation model.

[0268] 7. Synthesis of results

[0269] The server aggregates the results it receives.

[0270] Example: Compiling the generated poems into a single data package.

[0271] 8. Return of results

[0272] The server transmits the integrated results to the user terminal.

[0273] User terminal

[0274] 1. Enter the prompt

[0275] The user enters a prompt into the interface.

[0276] Example: User types "Generate a comforting poem for when I'm sad."

[0277] 2. Sending a prompt

[0278] The user terminal sends the entered prompt to the server.

[0279] 3. Displaying the results

[0280] The results returned by the server are displayed to the user.

[0281] Example: A poem to comfort you when you're feeling sad is displayed.

[0282] Specific examples

[0283] Example 1: Emotion-based poetry generation process

[0284] The user enters the prompt "Generate a comforting poem for sad feelings."

[0285] The user terminal sends a prompt to the server.

[0286] The server receives the prompt and recognizes the emotion "sad" using an emotion engine.

[0287] The server parses the prompt and identifies the task "Generate Poetry."

[0288] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[0289] The server receives the poem generated from the poetry generation model.

[0290] The server aggregates the generated poems and transmits them to the user's terminal.

[0291] The user terminal displays the received poem to the user.

[0292] The user checks the generated poem and experiences changes in emotion.

[0293] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling users to easily complete complex tasks.

[0294] The processing flow will be explained below.

[0295] Step 1:

[0296] The user enters a prompt into the interface.

[0297] Example: Type "Generate a poem to comfort me when I'm sad."

[0298] Step 2:

[0299] The terminal sends the prompt entered by the user to the server.

[0300] For example: A prompt is sent to the server.

[0301] Step 3:

[0302] The server receives the prompt.

[0303] Example: Accept the prompt "Generate a poem to comfort someone who is sad."

[0304] Step 4:

[0305] The server uses an emotion engine to analyze the user's emotion from the text data of the prompt.

[0306] Example: Analyze keywords such as "sad" in the prompt and recognize that the user's emotion is "sadness."

[0307] Step 5:

[0308] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[0309] Example: Parse the prompt and identify the task "Generate a poem."

[0310] Step 6:

[0311] The server generates sub-prompts corresponding to each task, taking into account the emotional information obtained from the emotion engine.

[0312] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[0313] Step 7:

[0314] The server sends the generated sub-prompts to the selected artificial intelligence model.

[0315] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[0316] Step 8:

[0317] The server receives the output from the artificial intelligence model.

[0318] Example: Accepting a poem generated from a poetry generation model.

[0319] Step 9:

[0320] The server consolidates the results it receives.

[0321] Example: Compiling the generated poems into a single data package.

[0322] Step 10:

[0323] The server transmits the integrated results to the user terminal.

[0324] Example: Sending a consolidated poetry data package to a user device.

[0325] Step 11:

[0326] The terminal displays the results received from the server to the user.

[0327] Example: "Poems to comfort sad feelings" is displayed on the user's interface.

[0328] Step 12:

[0329] The user checks the displayed results.

[0330] Example: Read the generated poem and experience the changes in emotions.

[0331] This system analyzes the user's emotions from the prompt text data and uses the results to generate more personalized results so that the user can get appropriate content that takes their emotions into consideration.

[0332] Example 2

[0333] 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."

[0334] Currently, many artificial intelligence systems exist, but they generally use a single model to respond to prompts entered by the user. However, it is difficult for a single model to generate outputs that adequately reflect the user's emotions. This creates a need for systems that can provide a truly personalized experience for users. Furthermore, they lack the ability to automatically identify multiple tasks and generate appropriate sub-prompts. A system that can solve these problems is needed.

[0335] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0336] In this invention, the server includes means including an emotion engine that recognizes a user's emotions and influences the analysis result, means for receiving a prompt from the user, means for analyzing the received prompt and dividing it into multiple sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, and means for returning the integrated results to the user. This makes it possible to provide a truly personalized output by taking the user's emotions into consideration and appropriately processing multiple tasks.

[0337] A "prompt" is text that a user enters into a system as an instruction or question.

[0338] A "subprompt" is a more specific instruction for performing a particular task that is generated from a parsed prompt.

[0339] An "artificial intelligence model" is a machine learning algorithm that has been trained to perform a specific task automatically.

[0340] An "emotion engine" is a part of a system that analyzes user input text and identifies the emotions contained in that text.

[0341] A "task" is an individual task or operation that a system must perform.

[0342] "Results synthesis" is the process of combining the individual outputs from multiple artificial intelligence models into a single, coherent result.

[0343] This invention is a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and affects the analysis results.

[0344] Hardware and software configuration

[0345] The system of the present invention uses the following hardware and software.

[0346] Server: A server computer equipped with a high-performance processor and a large amount of memory. Specific examples include cloud-based servers (AWS (registered trademark), Google (registered trademark), Microsoft (registered trademark), Azure (registered trademark), etc.).

[0347] User terminal: A device used by a user to access a system. Examples include personal computers, smartphones, and tablets.

[0348] Emotion engine: Software that uses natural language processing techniques to analyze user emotions, for example, by using open source NLP libraries (such as Hugging Face's Transformers).

[0349] Artificial intelligence model: As a generative AI model, we use machine learning models (e.g., GPT-3®, BERT) that are trained to perform specific tasks.

[0350] System Operation

[0351] 1. User prompt input

[0352] The user enters a prompt into the interface, for example, text such as "Generate a poem to comfort me when I'm feeling sad."

[0353] 2. Sending a prompt

[0354] The user terminal sends the input prompt to the server using an HTTP POST request.

[0355] 3. Accepting and parsing prompts

[0356] The server receives a prompt sent from a user terminal, identifies the emotion of the prompt using an emotion engine, and then analyzes the prompt and divides it into multiple tasks, for example, identifying a task called "poetry creation."

[0357] 4. Generating a subprompt

[0358] Based on the analysis results, the server generates sub-prompts for each task, taking into account the results of the emotion engine. Specifically, it generates specific sub-prompts such as "Please generate a poem that will comfort sad feelings."

[0359] 5. Submitting a subprompt and accepting the results

[0360] The server sends the generated sub-prompts to an artificial intelligence model, such as a poetry generation model, and receives output from the model.

[0361] 6. Consolidating and returning results

[0362] The server integrates the results obtained from multiple tasks and sends them back to the user's device, which receives them and displays them on the screen.

[0363] Specific examples

[0364] Example 1: Emotion-based poetry generation process

[0365] The user enters the prompt "Generate a comforting poem for sad feelings."

[0366] The user terminal sends a prompt to the server.

[0367] The server receives the prompt and identifies the emotion "sad" through an emotion engine.

[0368] The server parses the prompt and identifies the task "Generate Poetry."

[0369] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[0370] The server receives the poem generated from the poetry generation model.

[0371] The server aggregates the generated poems and transmits them to the user's terminal.

[0372] The user terminal displays the received poem to the user.

[0373] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling complex tasks to be completed easily.

[0374] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0375] Step 1:

[0376] The user terminal sends a prompt to the server. Specifically, when the user enters "Create a poem to comfort me when I'm sad" into the interface and presses the submit button, the system sends an HTTP POST request to the server. The input is a text prompt, and the output is a request sent to the server.

[0377] Step 2:

[0378] The server receives a prompt sent from the user terminal. The server receives an HTTP POST request and parses the text prompt included in the request body. At this time, the server extracts the prompt text and passes it on to subsequent processing. The input is the prompt text sent by the user, and the output is the prompt data to be parsed.

[0379] Step 3:

[0380] The emotion engine on the server analyzes the prompt and recognizes the user's emotion. It uses a natural language processing (NLP) library to tokenize the prompt and then runs it through an emotion classifier to extract an emotion tag. Specifically, it recognizes the emotion "sad." In this process, the input is the prompt text and the output is an emotion tag (e.g., "sad").

[0381] Step 4:

[0382] The server parses the prompt and splits it into tasks. NLP techniques are used to parse the text and identify tasks (e.g., "Generate Poetry"). The input is the prompt text with emotion tags, and the output is the identified tasks.

[0383] Step 5:

[0384] Based on the analysis results, the server generates sub-prompts corresponding to each task, taking into account the results of the emotion engine. For example, it generates a sub-prompt such as "Please create a poem that will comfort me when I'm feeling sad." In this process, the input is the identified task and emotion tag, and the output is the sub-prompt.

[0385] Step 6:

[0386] The server sends the generated sub-prompt to the appropriate AI model. It sends the sub-prompt via an HTTP request to the poetry generation model's API endpoint and waits for a response. The input is the sub-prompt, and the output is a request sent to the poetry generation model.

[0387] Step 7:

[0388] The server receives the output from the AI ​​model, analyzes the HTTP response from the poetry generation model, and obtains the generated poem. The input is the response data from the poetry generation model, and the output is the text of the generated poem.

[0389] Step 8:

[0390] The server aggregates the results from multiple tasks, compiles the generated poems into a single data package, and formats it in a format that can be sent back to the user. The input is the multiple output results, and the output is the aggregated result data.

[0391] Step 9:

[0392] The server returns the integrated results to the user terminal. An HTTP response is generated and the integrated result data is sent to the user terminal. The input is the integrated result data, and the output is the response sent to the user terminal.

[0393] Step 10:

[0394] The user terminal displays the results returned from the server to the user. The HTTP response is received and its contents are displayed on the screen. As a concrete example, the generated poem is displayed on the user's interface. The input is the response data from the server, and the output is the result displayed to the user.

[0395] (Application example 2)

[0396] 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."

[0397] Existing food delivery systems can respond quickly to user orders, but they cannot provide personalized suggestions that take into account the user's emotions and mood. This limits their ability to improve user satisfaction. In particular, there is a need for systems that provide meal suggestions that take into account emotions such as stress and fatigue.

[0398] The specification process 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 receiving a prompt from the user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to each AI model, means for analyzing the user's emotions, and means for generating sub-prompts based on the emotion analysis results and sending them to the AI ​​model. This enables personalized meal suggestions tailored to the user's emotions.

[0399] A "prompt" refers to the text of a request or question entered by a user into the system.

[0400] A "sub-prompt" is a prompt that you receive broken down into multiple smaller tasks.

[0401] "Artificial intelligence model" refers to a machine learning algorithm for performing a specific task, including generative and propositional models.

[0402] "Sentiment analysis" refers to the process of identifying and analyzing emotions from a user's input prompts.

[0403] A "generative model" refers to an AI algorithm that automatically generates new text or images based on input requests.

[0404] "Proposal model" refers to an AI algorithm that provides appropriate suggestions and options based on the analysis results.

[0405] "Synthesis" refers to the process of combining the results of multiple sub-prompts into one.

[0406] The present invention realizes a system that analyzes prompts from a user, generates multiple sub-prompts based on the emotional analysis results, and sends them to multiple artificial intelligence models to provide personalized results to the user. To implement this invention, a server, a user terminal, and various software are required.

[0407] Hardware and software used

[0408] Hardware

[0409] server

[0410] User device (smartphone)

[0411] software

[0412] Sentiment analysis engines (e.g., Python-based NLP libraries such as spaCy and NLTK)

[0413] Artificial intelligence models (e.g., GPT-4® or Transformers for running generative and propositional models)

[0414] A database (for temporarily storing results)

[0415] System configuration

[0416] 1. Prompt input from the user terminal

[0417] A user uses a food delivery application to enter a prompt, which can be entered as text or speech. For example, "I'm feeling stressed today, can you recommend a comforting meal?"

[0418] 2. Accepting the prompt

[0419] The user terminal sends the input prompt to the server, which receives it and inputs it into the emotion analysis engine.

[0420] 3. Emotion Analysis

[0421] The server uses an emotion analysis engine (such as spaCy or NLTK) to analyze the text in the prompt and identify the user's emotion. For example, it analyzes the emotion "stress" from the prompt "I'm stressed."

[0422] 4. Parsing the prompt

[0423] The server splits the prompt into multiple sub-prompts, taking into account the results of the sentiment analysis, using a text analysis library or similar.

[0424] 5. Generating Subprompts

[0425] Based on the results of the sentiment analysis, the server generates an appropriate sub-prompt, such as "Please suggest some dishes that will help relieve stress."

[0426] 6. Sending sub-prompts

[0427] The server then sends the generated sub-prompts to an appropriate artificial intelligence model, such as a generative model that suggests dishes.

[0428] 7. Acceptance and Integration of Results

[0429] The server receives and integrates the results from the artificial intelligence models, a process that combines multiple results into a single data package.

[0430] 8. Return of results

[0431] The server sends the integrated results to the user's device, which then displays, for example, a list of recommended dishes for relieving stress.

[0432] Specific examples

[0433] User: Enters the prompt "I'm feeling stressed today, what's a comforting dish?"

[0434] User terminal: Sends the prompt to the server.

[0435] Server: Receives the prompt, performs emotion analysis, and identifies the emotion "stress."

[0436] Server: Based on the analysis results, generate a sub-prompt such as "Please suggest a dish that will help relieve stress" and send it to the dish suggestion model.

[0437] Server: Receives the resulting proposals, integrates them, and sends them back to the user device.

[0438] User terminal: The suggested dishes are displayed to the user.

[0439] This process allows users to receive personalized meal suggestions that match their emotions, realizing a food delivery service that meets a variety of needs.

[0440] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0441] Step 1:

[0442] A user opens a food delivery application on a user device such as a smartphone and inputs a prompt such as, "I'm feeling stressed today, so please recommend some soothing dishes." This user input becomes the initial prompt data.

[0443] Step 2:

[0444] The user terminal sends the input prompt to the server. The prompt data sent to the server is "I'm feeling stressed today, so please tell me some soothing dishes."

[0445] Step 3:

[0446] The server receives the prompt and uses a sentiment analysis engine (e.g., spaCy or NLTK) to analyze the text in the prompt and identify the user's sentiment. In this case, the resulting sentiment is "stress."

[0447] Step 4:

[0448] The server analyzes the prompt based on the results of the sentiment analysis and divides it into appropriate sub-prompts. In this case, the sub-prompt generated is "Please suggest a dish that will help relieve stress."

[0449] Step 5:

[0450] The server sends the generated sub-prompts to a dish suggestion model (e.g., GPT-4 or Transformers), which receives the sub-prompts as input and then outputs dish suggestions based on them.

[0451] Step 6:

[0452] The server receives the results of the dish recommendation model, such as a list of dishes like "herbal tea," "banana smoothie," and "yogurt parfait."

[0453] Step 7:

[0454] The server aggregates the multiple recommendations into a single data package, in this case, a list of recommended dishes.

[0455] Step 8:

[0456] The server transmits the integrated results to the user terminal, and the transmitted data includes the integrated recipe list.

[0457] Step 9:

[0458] The user device displays the received list of dishes to the user, who can then check suggested dishes such as "herbal tea," "banana smoothie," and "yogurt parfait" on the application.

[0459] The above steps realize a system that allows users to get personalized recipe suggestions based on their emotions.

[0460] 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.

[0461] 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.

[0462] 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.

[0463] [Second embodiment]

[0464] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0465] 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.

[0466] 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).

[0467] 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.

[0468] 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.

[0469] 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).

[0470] 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.

[0471] 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.

[0472] 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.

[0473] 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.

[0474] 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.

[0475] 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."

[0476] The present invention provides a system that allows a user to easily order a primary prompt and coordinates the prompt across multiple artificial intelligence models, including means for receiving a prompt from a user, means for parsing the prompt and dividing it into multiple sub-prompts, means for sending the sub-prompts to respective artificial intelligence models, means for receiving results from the artificial intelligence models, means for aggregating the results, and means for returning the aggregated results to the user.

[0477] Program processing

[0478] server

[0479] 1. Accepting the prompt

[0480] The server receives the prompt sent from the user terminal.

[0481] Example: A user enters the prompt "Generate an image of a beautiful landscape and add a description."

[0482] 2. Parsing the prompt

[0483] The server analyzes the received prompts and splits them into different tasks (e.g., image generation, description generation).

[0484] 3. Generating a subprompt

[0485] The server generates sub-prompts for each task based on the analysis results.

[0486] Example: Generate a sub-prompt for "Generate an image of a beautiful landscape" and another for "Generate a description of the image."

[0487] 4. Sending a sub-prompt

[0488] The server sends the generated sub-prompt to the appropriate artificial intelligence model, for example, sending a sub-prompt such as "Generate an image of a beautiful landscape" to an image generation model, and a sub-prompt such as "Generate a description of the image" to a text generation model.

[0489] 5. Acceptance of Results

[0490] The server receives the output results from each artificial intelligence model.

[0491] Example: Receive a landscape image generated from an image generation model and a description generated from a text generation model.

[0492] 6. Synthesis of results

[0493] The server combines the received images and descriptions into a single consolidated result.

[0494] Example: The integrated results include a landscape image and its description.

[0495] 7. Return of results

[0496] The server transmits the integrated results to the user terminal.

[0497] User terminal

[0498] 1. Enter the prompt

[0499] The user enters a prompt into the interface.

[0500] Example: User types "Generate an image of a beautiful landscape and add a description."

[0501] 2. Sending a prompt

[0502] The user terminal sends the entered prompt to the server.

[0503] 3. Displaying the results

[0504] The results returned by the server are displayed to the user.

[0505] Example: A beautiful landscape image and its description are displayed.

[0506] Specific examples

[0507] Example 1: Image and description generation process

[0508] The user enters the prompt "Create an image of a beautiful landscape and add a description."

[0509] The user terminal sends a prompt to the server.

[0510] The server receives the prompt and splits it into an image generation task and a description generation task.

[0511] The server sends a sub-prompt to the image generation model saying, "Generate an image of a beautiful landscape," and a sub-prompt to the text generation model saying, "Generate a description of the image."

[0512] The server receives the landscape image generated from the image generation model and the explanatory text generated from the text generation model.

[0513] The server aggregates these results and sends them to the user terminal.

[0514] The user terminal displays the received landscape image and explanatory text on the interface.

[0515] The user checks the generated image and description.

[0516] This invention provides a system that allows users to seamlessly utilize multiple artificial intelligence models and efficiently process complex tasks, enabling users to easily obtain high-quality results.

[0517] The processing flow will be explained below.

[0518] Step 1:

[0519] The user enters a prompt into the interface.

[0520] For example, type "Generate an image of a beautiful landscape and add a description."

[0521] Step 2:

[0522] The terminal sends the prompt entered by the user to the server.

[0523] For example: A prompt is sent to the server.

[0524] Step 3:

[0525] The server receives the prompt.

[0526] Example: Accept the prompt "Generate an image of a beautiful landscape and add a description."

[0527] Step 4:

[0528] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[0529] Example: Split into tasks "image generation" and "description generation".

[0530] Step 5:

[0531] The server generates sub-prompts corresponding to each task.

[0532] Example: Generate the sub-prompts "Generate an image of a beautiful landscape" and "Generate a description of the generated image."

[0533] Step 6:

[0534] The server sends the generated sub-prompts to the selected artificial intelligence model.

[0535] Example: Send an image generation model the sub-prompt "Generate an image of a beautiful landscape," and a text generation model the sub-prompt "Generate a description of the generated image."

[0536] Step 7:

[0537] The server receives the output results from each artificial intelligence model.

[0538] Example: Accepts beautiful landscape images generated from an image generation model and explanatory text generated from a text generation model.

[0539] Step 8:

[0540] The server consolidates the results it receives.

[0541] Example: A landscape image and its description are packaged together as a single data package.

[0542] Step 9:

[0543] The server transmits the integrated results to the user terminal.

[0544] Example: Sending a synthesized landscape image and its description to a user device.

[0545] Step 10:

[0546] The terminal displays the results received from the server to the user.

[0547] Example: A user interface displays an image of a beautiful landscape and a description of it.

[0548] Step 11:

[0549] The user checks the displayed results.

[0550] Example: View and evaluate the generated landscape images and their descriptions.

[0551] Example 1

[0552] 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."

[0553] In today's world, users are required to efficiently process complex tasks using multiple AI models, but existing systems lack a simple way to achieve this. A method is needed that allows users to seamlessly execute different tasks, such as image generation and text generation, within a single system, and then integrate and efficiently use the results.

[0554] 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.

[0555] In this invention, the server includes means for receiving a prompt sentence from a user terminal, means for analyzing the received prompt sentence and dividing it into multiple sub-prompt sentences based on the prompt sentence, means for sending the sub-prompt sentences to each generative AI model, means for receiving results from the generative AI models, means for integrating the received results, and means for returning the integrated results to the user terminal, thereby enabling a user to efficiently process complex tasks using multiple AI models and easily obtain the integrated results.

[0556] A "prompt" is a textual instruction for a user to enter a specific task or request.

[0557] A "user terminal" is a device (e.g., a computer, smartphone, tablet, etc.) that a user uses to input prompt sentences and communicate with a server.

[0558] The "server" is a central computer system that receives prompts, analyzes them, distributes tasks to generative AI models, and synthesizes and returns the results to the user device.

[0559] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task (e.g., image generation or text generation).

[0560] An "image generation model" is an artificial intelligence model that is trained to generate images based on input prompts.

[0561] A "text generation model" is an artificial intelligence model that is trained to generate natural language text based on an input prompt.

[0562] The "means for parsing a prompt sentence" is a software component that has the function of parsing a prompt sentence and dividing it into multiple sub-prompt sentences.

[0563] A "sub-prompt sentence" is a detailed instruction for a specific task that is extracted as a result of parsing a prompt sentence.

[0564] A "means for aggregating results" is a software component that has the ability to combine individual results from a generative AI model into a single aggregated result.

[0565] The present invention provides a system that allows users to easily input prompt sentences and process complex tasks using multiple generative AI models. This system consists of a server and a user terminal.

[0566] server

[0567] The server has the functions of receiving prompt sentences from the user device, analyzing the received prompt sentences and dividing them into multiple sub-prompts, sending the sub-prompts to the appropriate generative AI model, and receiving the results from the generative AI model, integrating them, and sending them back to the user device.

[0568] User terminal

[0569] The user terminal provides an interface for the user to input a prompt sentence, which is then sent to the server, which processes the sentence and displays the result.

[0570] Specific implementation example

[0571] Entering a prompt statement

[0572] The user inputs a prompt sentence into the interface of the user terminal, such as "Create an image of a beautiful landscape and add a description of it." This prompt sentence is sent to the server as an HTTP request.

[0573] Parsing the prompt statement

[0574] The server analyzes the received prompt sentence. It uses natural language processing (NLP) technology to divide the prompt sentence into an image generation task and a text generation task. Specifically, it generates the following sub-prompt sentences:

[0575] Image generation subprompt: "Generate an image of a beautiful landscape."

[0576] Text generation subprompt: "Generate a description for that image."

[0577] Sending a subprompt statement

[0578] The server sends the image generation subprompt to the image generation model and the text generation subprompt to the text generation model. These models are generative AI models trained specifically for each task. The server sends the subprompt sentence as an API request and waits for a response from the model.

[0579] Acceptance and integration of results

[0580] The server receives the results from the generative AI models. For example, an image generation model might return an image of a beautiful landscape, and a text generation model might return a description of that image. It then combines these results and stores them in a single data format (e.g., JSON).

[0581] Returning the results

[0582] The server sends the integrated results to the user's device, which then displays the results on its interface. The user can then view the beautiful landscape images and their descriptions.

[0583] This system allows users to efficiently process complex tasks using multiple generative AI models and easily obtain integrated results. By clarifying what prompts will produce what results, the user experience is greatly improved.

[0584] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0585] Step 1:

[0586] Entering a prompt statement

[0587] The user inputs a prompt sentence through the interface of the user terminal.

[0588] Input: The prompt text entered by the user (e.g., "Generate an image of a beautiful landscape and add a description.")

[0589] Output: The user terminal generates an HTTP request to send the prompt text to the server.

[0590] Step 2:

[0591] Receiving a prompt

[0592] The server receives the prompt sentence sent from the user terminal.

[0593] Input: Prompt sent from the user's terminal to the server

[0594] Output: Stores the prompt text in an internal data structure.

[0595] Step 3:

[0596] Parsing the prompt statement

[0597] The server analyzes the received prompt sentence using natural language processing technology and divides it into multiple sub-prompt sentences.

[0598] Input: The received prompt (e.g., "Generate an image of a beautiful landscape and add a description.")

[0599] Output: Split sub-prompt sentences (e.g. "Generate an image of a beautiful landscape", "Generate a description of the image")

[0600] Step 4:

[0601] Sending a subprompt statement

[0602] The server sends the split subprompt sentences to the appropriate generative AI model. For example, an image generation subprompt is sent to an image generation model, and an explanatory sentence generation subprompt is sent to a text generation model.

[0603] Input: Separated subprompt sentence

[0604] Output: API request to the generative AI model

[0605] Step 5:

[0606] Acceptance of results

[0607] The server receives the results returned by the generative AI models: generated images from the image generation model, and generated explanatory text from the text generation model.

[0608] Input: Responses from each generative AI model (e.g., generated landscape images, generated descriptions)

[0609] Output: Stores the resulting data in an internal data structure.

[0610] Step 6:

[0611] Integration of results

[0612] The server then combines the results it receives, for example combining images and descriptions into a single data format (e.g., a JSON object).

[0613] Input: Accepted result data (e.g., generated landscape image, generated description)

[0614] Output: Consolidated results (e.g., a JSON object containing image URLs and descriptions)

[0615] Step 7:

[0616] Returning the results

[0617] The server returns the consolidated results to the user terminal.

[0618] Input: Consolidated results (e.g., JSON format data)

[0619] Output: HTTP response to the user's device

[0620] Step 8:

[0621] Displaying the results

[0622] The user terminal receives the results returned from the server and displays them to the user.

[0623] Input: Integration results sent from the server (e.g., JSON format data)

[0624] Output: The result shown to the user (e.g., a landscape image and its description)

[0625] Through the above process, users can efficiently obtain the results of complex tasks that go through multiple generative AI models.

[0626] (Application example 1)

[0627] 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."

[0628] Conventional customer support systems for virtual stores have been limited to using individual models or simple FAQ databases to answer user questions, making it difficult to provide integrated, advanced answers. As a result, there has been a problem in that it takes a lot of time and effort for users to obtain the information they are looking for.

[0629] 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.

[0630] In this invention, the server includes means for receiving a prompt from a user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, means for returning the integrated results to the user, and means for analyzing questions about products in a virtual store and providing related information, thereby enabling the user to obtain detailed and comprehensive information about products by inputting a prompt once.

[0631] "Means for receiving prompts from a user" refers to a device or software function that allows the system to receive questions or requests entered by a user.

[0632] The "means for analyzing a prompt and dividing it into a plurality of sub-prompts based on the prompt" refers to the function of a device or software that understands a received prompt and breaks it down into a plurality of specific tasks.

[0633] The "means for sending sub-prompts to respective artificial intelligence models" refers to a device or software function that sends the divided sub-prompts to the appropriate artificial intelligence models so that they can be processed.

[0634] "Means for receiving results from an artificial intelligence model" refers to a device or software function for receiving results generated by an artificial intelligence model.

[0635] "Means for integrating received results" refers to the function of a device or software that combines the results received from each artificial intelligence model into a single, consistent result.

[0636] A "means for returning the consolidated results to the user" is a device or software function that displays or presents the consolidated results to the user.

[0637] "Means for analyzing product-related questions within a virtual store and providing related information" refers to a device or software function for analyzing product-related questions entered by users in a virtual store and providing related information in response to those questions.

[0638] An "image generation model" is a type of artificial intelligence that generates images based on input text prompts.

[0639] A "sentence generation model" is a type of artificial intelligence that generates sentences based on input text prompts.

[0640] A "review analysis model" is a type of artificial intelligence that analyzes product reviews and summarizes or evaluates their content.

[0641] The present invention is a system that allows users to easily order key prompts and coordinates prompts across multiple artificial intelligence models. This system aims to enable multiple artificial intelligence models to cooperate and generate answers to questions about products asked by users, particularly in virtual stores.

[0642] The main components of the system are as follows:

[0643] 1. Accepting the prompt

[0644] The user inputs a question about a product in the virtual store. For example, the user inputs a prompt such as "How is the camera performance of this smartphone?" This prompt is sent from the user terminal to the server.

[0645] 2. Parsing the prompt

[0646] The server parses the received prompt and divides it into multiple sub-prompts based on different tasks, for example, "analyzing image samples from a camera," "presenting technical specifications of a camera," and "summarizing user reviews."

[0647] 3. Generate and send a subprompt

[0648] Based on the analysis results, the server generates sub-prompts appropriate for each task and sends them to each AI model, such as "Generate an image sample from the camera," "Explain the camera's technical specifications," or "Summarize user reviews about the camera's performance."

[0649] 4. Acceptance of Results

[0650] The server receives the output results from each artificial intelligence model (image generation model, sentence generation model, review analysis model).

[0651] 5. Synthesis of results

[0652] The server then combines the different results it receives to produce a single, coherent answer, which includes image samples of the camera's capabilities, detailed technical specifications, and highlights from user reviews.

[0653] 6. Return of results

[0654] Finally, the server sends the integrated results back to the user device, which then displays them to the user. For example, in response to a prompt entered by the user, a comprehensive answer such as "This smartphone's camera performance is excellent, and it takes clear photos even at night. In addition, many users rate its ease of use highly."

[0655] Program processing explanation

[0656] The system for implementing the present invention uses the following hardware and software. User terminals include smartphones, tablets, and PCs, and the server uses software such as Flask (a Python web framework), OpenAI API, PyTorch, and Hugging Face Transformers. Users enter questions about products in the virtual store, and the prompts are sent to the server.

[0657] The server analyzes the prompts and sends sub-prompts to the appropriate artificial intelligence models. The image generation model generates images based on the input text prompt, the sentence generation model generates sentences based on the text prompt, and the review analysis model analyzes product reviews and summarizes or rates their content. The server consolidates these results and returns them to the user.

[0658] Specific examples

[0659] For example, a user enters the prompt "How good is the camera on this phone?" The prompt is split into:

[0660] "Generate a sample image from the camera"

[0661] "Please explain the technical specifications of the camera."

[0662] "Please summarize user reviews of the camera's performance."

[0663] The results of the generative AI model based on these sub-prompts are integrated and sent back to the user, allowing the user to obtain comprehensive information with a single prompt input.

[0664] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0665] Step 1:

[0666] A user enters a question about a product in a virtual store

[0667] A user inputs a prompt such as "How is the camera performance of this smartphone?" into the user terminal and sends the question to the server. The input is a text prompt, and the output is the prompt data transferred to the server.

[0668] Step 2:

[0669] The server receives a prompt

[0670] The server receives prompts sent from the user terminal and stores them as input data for parsing. The input is the raw prompt text sent by the user, and the output is the prompt to be parsed.

[0671] Step 3:

[0672] The server parses the prompt and splits it into subprompts

[0673] The server analyzes the received prompt and divides it into multiple sub-prompts based on different tasks, such as "analyzing camera image samples," "presenting camera specifications," and "summarizing user reviews." The input is the prompt text, and the output is a set of sub-prompts corresponding to each task.

[0674] Step 4:

[0675] Send sub-prompts to each AI model

[0676] The server sends the generated sub-prompts to each AI model. For example, it sends a sub-prompt such as "Generate an image sample of the camera" to the image generation model, and a sub-prompt such as "Describe the technical specifications of the camera" to the sentence generation model. The input is the sub-prompt, and the output is the task sent to each model.

[0677] Step 5:

[0678] The server receives the results from each AI model.

[0679] The server receives the output from each AI model. The image generation model generates sample images of the camera, the sentence generation model generates technical specifications for the camera, and the review analysis model generates summaries of user reviews. The input is the generated results based on the sub-prompts, and the output is the resulting data from each model.

[0680] Step 6:

[0681] The server aggregates the results

[0682] The server then combines the results it receives to generate a single, coherent answer. For example, it might combine the generated camera image, technical specifications, and review summary into a single response. The input is the result data from each model, and the output is the combined answer.

[0683] Step 7:

[0684] The server sends the consolidated results back to the user

[0685] The server sends the integrated results to the user's device. The user's device displays the results to the user, allowing the user to obtain comprehensive information about the product. For example, the user's device might display a message such as, "This smartphone's camera performance is excellent, and it takes clear photos even at night. Many users also rate its ease of use highly." The input is the integrated result data, and the output is the content displayed on the user's device.

[0686] 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.

[0687] This invention provides a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and influences the analysis results.

[0688] Program processing

[0689] server

[0690] 1. Accepting the prompt

[0691] The server receives the prompt sent from the user terminal.

[0692] Example: A user enters the prompt "Generate a poem to comfort me when I'm sad."

[0693] 2. Emotion Analysis

[0694] The server is equipped with an emotion engine that analyzes text data from the user's input prompts to recognize the user's emotions.

[0695] Example: The emotion "sad" is recognized from the input prompt.

[0696] 3. Parsing the prompt

[0697] The server parses the received prompt, understands its content and splits it into appropriate tasks.

[0698] Example: Identify the task "Generate a poem."

[0699] 4. Generating a subprompt

[0700] The server generates sub-prompts for each task based on the analysis results, taking into account the analysis results of the emotion engine.

[0701] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[0702] 5. Sending a sub-prompt

[0703] The server sends the generated sub-prompts to the appropriate artificial intelligence model.

[0704] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[0705] 6. Acceptance of Results

[0706] The server receives the output from the artificial intelligence model.

[0707] Example: Accepting a poem generated from a poetry generation model.

[0708] 7. Synthesis of results

[0709] The server aggregates the results it receives.

[0710] Example: Compiling the generated poems into a single data package.

[0711] 8. Return of results

[0712] The server transmits the integrated results to the user terminal.

[0713] User terminal

[0714] 1. Enter the prompt

[0715] The user enters a prompt into the interface.

[0716] Example: User types "Generate a comforting poem for when I'm sad."

[0717] 2. Sending a prompt

[0718] The user terminal sends the entered prompt to the server.

[0719] 3. Displaying the results

[0720] The results returned by the server are displayed to the user.

[0721] Example: A poem to comfort you when you're feeling sad is displayed.

[0722] Specific examples

[0723] Example 1: Emotion-based poetry generation process

[0724] The user enters the prompt "Generate a comforting poem for sad feelings."

[0725] The user terminal sends a prompt to the server.

[0726] The server receives the prompt and recognizes the emotion "sad" using an emotion engine.

[0727] The server parses the prompt and identifies the task "Generate Poetry."

[0728] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[0729] The server receives the poem generated from the poetry generation model.

[0730] The server aggregates the generated poems and transmits them to the user's terminal.

[0731] The user terminal displays the received poem to the user.

[0732] The user checks the generated poem and experiences changes in emotion.

[0733] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling users to easily complete complex tasks.

[0734] The processing flow will be explained below.

[0735] Step 1:

[0736] The user enters a prompt into the interface.

[0737] Example: Type "Generate a poem to comfort me when I'm sad."

[0738] Step 2:

[0739] The terminal sends the prompt entered by the user to the server.

[0740] For example: A prompt is sent to the server.

[0741] Step 3:

[0742] The server receives the prompt.

[0743] Example: Accept the prompt "Generate a poem to comfort someone who is sad."

[0744] Step 4:

[0745] The server uses an emotion engine to analyze the user's emotion from the text data of the prompt.

[0746] Example: Analyze keywords such as "sad" in the prompt and recognize that the user's emotion is "sadness."

[0747] Step 5:

[0748] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[0749] Example: Parse the prompt and identify the task "Generate a poem."

[0750] Step 6:

[0751] The server generates sub-prompts corresponding to each task, taking into account the emotional information obtained from the emotion engine.

[0752] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[0753] Step 7:

[0754] The server sends the generated sub-prompts to the selected artificial intelligence model.

[0755] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[0756] Step 8:

[0757] The server receives the output from the artificial intelligence model.

[0758] Example: Accepting a poem generated from a poetry generation model.

[0759] Step 9:

[0760] The server consolidates the results it receives.

[0761] Example: Compiling the generated poems into a single data package.

[0762] Step 10:

[0763] The server transmits the integrated results to the user terminal.

[0764] Example: Sending a consolidated poetry data package to a user device.

[0765] Step 11:

[0766] The terminal displays the results received from the server to the user.

[0767] Example: "Poems to comfort sad feelings" is displayed on the user's interface.

[0768] Step 12:

[0769] The user checks the displayed results.

[0770] Example: Read the generated poem and experience the changes in emotions.

[0771] This system analyzes the user's emotions from the prompt text data and uses the results to generate more personalized results so that the user can get appropriate content that takes their emotions into consideration.

[0772] Example 2

[0773] 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."

[0774] Currently, many artificial intelligence systems exist, but they generally use a single model to respond to prompts entered by the user. However, it is difficult for a single model to generate outputs that adequately reflect the user's emotions. This creates a need for systems that can provide a truly personalized experience for users. Furthermore, they lack the ability to automatically identify multiple tasks and generate appropriate sub-prompts. A system that can solve these problems is needed.

[0775] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0776] In this invention, the server includes means including an emotion engine that recognizes a user's emotions and influences the analysis result, means for receiving a prompt from the user, means for analyzing the received prompt and dividing it into multiple sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, and means for returning the integrated results to the user. This makes it possible to provide a truly personalized output by taking the user's emotions into consideration and appropriately processing multiple tasks.

[0777] A "prompt" is text that a user enters into a system as an instruction or question.

[0778] A "subprompt" is a more specific instruction for performing a particular task that is generated from a parsed prompt.

[0779] An "artificial intelligence model" is a machine learning algorithm that has been trained to perform a specific task automatically.

[0780] An "emotion engine" is a part of a system that analyzes user input text and identifies the emotions contained in that text.

[0781] A "task" is an individual task or operation that a system must perform.

[0782] "Results synthesis" is the process of combining the individual outputs from multiple artificial intelligence models into a single, coherent result.

[0783] This invention is a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and affects the analysis results.

[0784] Hardware and software configuration

[0785] The system of the present invention uses the following hardware and software.

[0786] Server: A server computer with a high-performance processor and a large amount of memory. For example, you can use a cloud-based server (AWS, Google Cloud, Microsoft Azure, etc.).

[0787] User terminal: A device used by a user to access a system. Examples include personal computers, smartphones, and tablets.

[0788] Emotion engine: Software that uses natural language processing techniques to analyze user emotions, for example, by using open source NLP libraries (such as Hugging Face's Transformers).

[0789] Artificial intelligence models: Generative AI models use machine learning models (e.g., GPT-3, BERT) that are trained to perform specific tasks.

[0790] System Operation

[0791] 1. User prompt input

[0792] The user enters a prompt into the interface, for example, text such as "Generate a poem to comfort me when I'm feeling sad."

[0793] 2. Sending a prompt

[0794] The user terminal sends the input prompt to the server using an HTTP POST request.

[0795] 3. Accepting and parsing prompts

[0796] The server receives a prompt sent from a user terminal, identifies the emotion of the prompt using an emotion engine, and then analyzes the prompt and divides it into multiple tasks, for example, identifying a task called "poetry creation."

[0797] 4. Generating a subprompt

[0798] Based on the analysis results, the server generates sub-prompts for each task, taking into account the results of the emotion engine. Specifically, it generates specific sub-prompts such as "Please generate a poem that will comfort sad feelings."

[0799] 5. Submitting a subprompt and accepting the results

[0800] The server sends the generated sub-prompts to an artificial intelligence model, such as a poetry generation model, and receives output from the model.

[0801] 6. Consolidating and returning results

[0802] The server integrates the results obtained from multiple tasks and sends them back to the user's device, which receives them and displays them on the screen.

[0803] Specific examples

[0804] Example 1: Emotion-based poetry generation process

[0805] The user enters the prompt "Generate a comforting poem for sad feelings."

[0806] The user terminal sends a prompt to the server.

[0807] The server receives the prompt and identifies the emotion "sad" through an emotion engine.

[0808] The server parses the prompt and identifies the task "Generate Poetry."

[0809] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[0810] The server receives the poem generated from the poetry generation model.

[0811] The server aggregates the generated poems and transmits them to the user's terminal.

[0812] The user terminal displays the received poem to the user.

[0813] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling complex tasks to be completed easily.

[0814] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0815] Step 1:

[0816] The user terminal sends a prompt to the server. Specifically, when the user enters "Create a poem to comfort me when I'm sad" into the interface and presses the submit button, the system sends an HTTP POST request to the server. The input is a text prompt, and the output is a request sent to the server.

[0817] Step 2:

[0818] The server receives a prompt sent from the user terminal. The server receives an HTTP POST request and parses the text prompt included in the request body. At this time, the server extracts the prompt text and passes it on to subsequent processing. The input is the prompt text sent by the user, and the output is the prompt data to be parsed.

[0819] Step 3:

[0820] The emotion engine on the server analyzes the prompt and recognizes the user's emotion. It uses a natural language processing (NLP) library to tokenize the prompt and then runs it through an emotion classifier to extract an emotion tag. Specifically, it recognizes the emotion "sad." In this process, the input is the prompt text and the output is an emotion tag (e.g., "sad").

[0821] Step 4:

[0822] The server parses the prompt and splits it into tasks. NLP techniques are used to parse the text and identify tasks (e.g., "Generate Poetry"). The input is the prompt text with emotion tags, and the output is the identified tasks.

[0823] Step 5:

[0824] Based on the analysis results, the server generates sub-prompts corresponding to each task, taking into account the results of the emotion engine. For example, it generates a sub-prompt such as "Please create a poem that will comfort me when I'm feeling sad." In this process, the input is the identified task and emotion tag, and the output is the sub-prompt.

[0825] Step 6:

[0826] The server sends the generated sub-prompt to the appropriate AI model. It sends the sub-prompt via an HTTP request to the poetry generation model's API endpoint and waits for a response. The input is the sub-prompt, and the output is a request sent to the poetry generation model.

[0827] Step 7:

[0828] The server receives the output from the AI ​​model, analyzes the HTTP response from the poetry generation model, and obtains the generated poem. The input is the response data from the poetry generation model, and the output is the text of the generated poem.

[0829] Step 8:

[0830] The server aggregates the results from multiple tasks, compiles the generated poems into a single data package, and formats it in a format that can be sent back to the user. The input is the multiple output results, and the output is the aggregated result data.

[0831] Step 9:

[0832] The server returns the integrated results to the user terminal. An HTTP response is generated and the integrated result data is sent to the user terminal. The input is the integrated result data, and the output is the response sent to the user terminal.

[0833] Step 10:

[0834] The user terminal displays the results returned from the server to the user. The HTTP response is received and its contents are displayed on the screen. As a concrete example, the generated poem is displayed on the user's interface. The input is the response data from the server, and the output is the result displayed to the user.

[0835] (Application example 2)

[0836] 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."

[0837] Existing food delivery systems can respond quickly to user orders, but they cannot provide personalized suggestions that take into account the user's emotions and mood. This limits their ability to improve user satisfaction. In particular, there is a need for systems that provide meal suggestions that take into account emotions such as stress and fatigue.

[0838] The specification process 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 receiving a prompt from the user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to each AI model, means for analyzing the user's emotions, and means for generating sub-prompts based on the emotion analysis results and sending them to the AI ​​model. This enables personalized meal suggestions tailored to the user's emotions.

[0839] A "prompt" refers to the text of a request or question entered by a user into the system.

[0840] A "sub-prompt" is a prompt that you receive broken down into multiple smaller tasks.

[0841] "Artificial intelligence model" refers to a machine learning algorithm for performing a specific task, including generative and propositional models.

[0842] "Sentiment analysis" refers to the process of identifying and analyzing emotions from a user's input prompts.

[0843] A "generative model" refers to an AI algorithm that automatically generates new text or images based on input requests.

[0844] "Proposal model" refers to an AI algorithm that provides appropriate suggestions and options based on the analysis results.

[0845] "Synthesis" refers to the process of combining the results of multiple sub-prompts into one.

[0846] The present invention realizes a system that analyzes prompts from a user, generates multiple sub-prompts based on the emotional analysis results, and sends them to multiple artificial intelligence models to provide personalized results to the user. To implement this invention, a server, a user terminal, and various software are required.

[0847] Hardware and software used

[0848] Hardware

[0849] server

[0850] User device (smartphone)

[0851] software

[0852] Sentiment analysis engines (e.g., Python-based NLP libraries such as spaCy and NLTK)

[0853] Artificial intelligence models (e.g., GPT-4 or Transformers for running generative and propositional models)

[0854] A database (for temporarily storing results)

[0855] System configuration

[0856] 1. Prompt input from the user terminal

[0857] A user uses a food delivery application to enter a prompt, which can be entered as text or speech. For example, "I'm feeling stressed today, can you recommend a comforting meal?"

[0858] 2. Accepting the prompt

[0859] The user terminal sends the input prompt to the server, which receives it and inputs it into the emotion analysis engine.

[0860] 3. Emotion Analysis

[0861] The server uses an emotion analysis engine (such as spaCy or NLTK) to analyze the text in the prompt and identify the user's emotion. For example, it analyzes the emotion "stress" from the prompt "I'm stressed."

[0862] 4. Parsing the prompt

[0863] The server splits the prompt into multiple sub-prompts, taking into account the results of the sentiment analysis, using a text analysis library or similar.

[0864] 5. Generating Subprompts

[0865] Based on the results of the sentiment analysis, the server generates an appropriate sub-prompt, such as "Please suggest some dishes that will help relieve stress."

[0866] 6. Sending sub-prompts

[0867] The server then sends the generated sub-prompts to an appropriate artificial intelligence model, such as a generative model that suggests dishes.

[0868] 7. Acceptance and Integration of Results

[0869] The server receives and integrates the results from the artificial intelligence models, a process that combines multiple results into a single data package.

[0870] 8. Return of results

[0871] The server sends the integrated results to the user's device, which then displays, for example, a list of recommended dishes for relieving stress.

[0872] Specific examples

[0873] User: Enters the prompt "I'm feeling stressed today, what's a comforting dish?"

[0874] User terminal: Sends the prompt to the server.

[0875] Server: Receives the prompt, performs emotion analysis, and identifies the emotion "stress."

[0876] Server: Based on the analysis results, generate a sub-prompt such as "Please suggest a dish that will help relieve stress" and send it to the dish suggestion model.

[0877] Server: Receives the resulting proposals, integrates them, and sends them back to the user device.

[0878] User terminal: The suggested dishes are displayed to the user.

[0879] This process allows users to receive personalized meal suggestions that match their emotions, realizing a food delivery service that meets a variety of needs.

[0880] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0881] Step 1:

[0882] A user opens a food delivery application on a user device such as a smartphone and inputs a prompt such as, "I'm feeling stressed today, so please recommend some soothing dishes." This user input becomes the initial prompt data.

[0883] Step 2:

[0884] The user terminal sends the input prompt to the server. The prompt data sent to the server is "I'm feeling stressed today, so please tell me some soothing dishes."

[0885] Step 3:

[0886] The server receives the prompt and uses a sentiment analysis engine (e.g., spaCy or NLTK) to analyze the text in the prompt and identify the user's sentiment. In this case, the resulting sentiment is "stress."

[0887] Step 4:

[0888] The server analyzes the prompt based on the results of the sentiment analysis and divides it into appropriate sub-prompts. In this case, the sub-prompt generated is "Please suggest a dish that will help relieve stress."

[0889] Step 5:

[0890] The server sends the generated sub-prompts to a dish suggestion model (e.g., GPT-4 or Transformers), which receives the sub-prompts as input and then outputs dish suggestions based on them.

[0891] Step 6:

[0892] The server receives the results of the dish recommendation model, such as a list of dishes like "herbal tea," "banana smoothie," and "yogurt parfait."

[0893] Step 7:

[0894] The server aggregates the multiple recommendations into a single data package, in this case, a list of recommended dishes.

[0895] Step 8:

[0896] The server transmits the integrated results to the user terminal, and the transmitted data includes the integrated recipe list.

[0897] Step 9:

[0898] The user device displays the received list of dishes to the user, who can then check suggested dishes such as "herbal tea," "banana smoothie," and "yogurt parfait" on the application.

[0899] The above steps realize a system that allows users to get personalized recipe suggestions based on their emotions.

[0900] 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.

[0901] 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.

[0902] 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.

[0903] [Third embodiment]

[0904] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0905] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[0906] 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).

[0907] 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.

[0908] 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.

[0909] 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).

[0910] 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.

[0911] 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.

[0912] 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.

[0913] 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.

[0914] 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.

[0915] 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."

[0916] The present invention provides a system that allows a user to easily order a primary prompt and coordinates the prompt across multiple artificial intelligence models, including means for receiving a prompt from a user, means for parsing the prompt and dividing it into multiple sub-prompts, means for sending the sub-prompts to respective artificial intelligence models, means for receiving results from the artificial intelligence models, means for aggregating the results, and means for returning the aggregated results to the user.

[0917] Program processing

[0918] server

[0919] 1. Accepting the prompt

[0920] The server receives the prompt sent from the user terminal.

[0921] Example: A user enters the prompt "Generate an image of a beautiful landscape and add a description."

[0922] 2. Parsing the prompt

[0923] The server analyzes the received prompts and splits them into different tasks (e.g., image generation, description generation).

[0924] 3. Generating a subprompt

[0925] The server generates sub-prompts for each task based on the analysis results.

[0926] Example: Generate a sub-prompt for "Generate an image of a beautiful landscape" and another for "Generate a description of the image."

[0927] 4. Sending a sub-prompt

[0928] The server sends the generated sub-prompt to the appropriate artificial intelligence model, for example, sending a sub-prompt such as "Generate an image of a beautiful landscape" to an image generation model, and a sub-prompt such as "Generate a description of the image" to a text generation model.

[0929] 5. Acceptance of Results

[0930] The server receives the output results from each artificial intelligence model.

[0931] Example: Receive a landscape image generated from an image generation model and a description generated from a text generation model.

[0932] 6. Synthesis of results

[0933] The server combines the received images and descriptions into a single consolidated result.

[0934] Example: The integrated results include a landscape image and its description.

[0935] 7. Return of results

[0936] The server transmits the integrated results to the user terminal.

[0937] User terminal

[0938] 1. Enter the prompt

[0939] The user enters a prompt into the interface.

[0940] Example: User types "Generate an image of a beautiful landscape and add a description."

[0941] 2. Sending a prompt

[0942] The user terminal sends the entered prompt to the server.

[0943] 3. Displaying the results

[0944] The results returned by the server are displayed to the user.

[0945] Example: A beautiful landscape image and its description are displayed.

[0946] Specific examples

[0947] Example 1: Image and description generation process

[0948] The user enters the prompt "Create an image of a beautiful landscape and add a description."

[0949] The user terminal sends a prompt to the server.

[0950] The server receives the prompt and splits it into an image generation task and a description generation task.

[0951] The server sends a sub-prompt to the image generation model saying, "Generate an image of a beautiful landscape," and a sub-prompt to the text generation model saying, "Generate a description of the image."

[0952] The server receives the landscape image generated from the image generation model and the explanatory text generated from the text generation model.

[0953] The server aggregates these results and sends them to the user terminal.

[0954] The user terminal displays the received landscape image and explanatory text on the interface.

[0955] The user checks the generated image and description.

[0956] This invention provides a system that allows users to seamlessly utilize multiple artificial intelligence models and efficiently process complex tasks, enabling users to easily obtain high-quality results.

[0957] The processing flow will be explained below.

[0958] Step 1:

[0959] The user enters a prompt into the interface.

[0960] For example, type "Generate an image of a beautiful landscape and add a description."

[0961] Step 2:

[0962] The terminal sends the prompt entered by the user to the server.

[0963] For example: A prompt is sent to the server.

[0964] Step 3:

[0965] The server receives the prompt.

[0966] Example: Accept the prompt "Generate an image of a beautiful landscape and add a description."

[0967] Step 4:

[0968] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[0969] Example: Split into tasks "image generation" and "description generation".

[0970] Step 5:

[0971] The server generates sub-prompts corresponding to each task.

[0972] Example: Generate the sub-prompts "Generate an image of a beautiful landscape" and "Generate a description of the generated image."

[0973] Step 6:

[0974] The server sends the generated sub-prompts to the selected artificial intelligence model.

[0975] Example: Send an image generation model the sub-prompt "Generate an image of a beautiful landscape," and a text generation model the sub-prompt "Generate a description of the generated image."

[0976] Step 7:

[0977] The server receives the output results from each artificial intelligence model.

[0978] Example: Accepts beautiful landscape images generated from an image generation model and explanatory text generated from a text generation model.

[0979] Step 8:

[0980] The server consolidates the results it receives.

[0981] Example: A landscape image and its description are packaged together as a single data package.

[0982] Step 9:

[0983] The server transmits the integrated results to the user terminal.

[0984] Example: Sending a synthesized landscape image and its description to a user device.

[0985] Step 10:

[0986] The terminal displays the results received from the server to the user.

[0987] Example: A user interface displays an image of a beautiful landscape and a description of it.

[0988] Step 11:

[0989] The user checks the displayed results.

[0990] Example: View and evaluate the generated landscape images and their descriptions.

[0991] Example 1

[0992] 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."

[0993] In today's world, users are required to efficiently process complex tasks using multiple AI models, but existing systems lack a simple way to achieve this. A method is needed that allows users to seamlessly execute different tasks, such as image generation and text generation, within a single system, and then integrate and efficiently use the results.

[0994] 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.

[0995] In this invention, the server includes means for receiving a prompt sentence from a user terminal, means for analyzing the received prompt sentence and dividing it into multiple sub-prompt sentences based on the prompt sentence, means for sending the sub-prompt sentences to each generative AI model, means for receiving results from the generative AI models, means for integrating the received results, and means for returning the integrated results to the user terminal, thereby enabling a user to efficiently process complex tasks using multiple AI models and easily obtain the integrated results.

[0996] A "prompt" is a textual instruction for a user to enter a specific task or request.

[0997] A "user terminal" is a device (e.g., a computer, smartphone, tablet, etc.) that a user uses to input prompt sentences and communicate with a server.

[0998] The "server" is a central computer system that receives prompts, analyzes them, distributes tasks to generative AI models, and synthesizes and returns the results to the user device.

[0999] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task (e.g., image generation or text generation).

[1000] An "image generation model" is an artificial intelligence model that is trained to generate images based on input prompts.

[1001] A "text generation model" is an artificial intelligence model that is trained to generate natural language text based on an input prompt.

[1002] The "means for parsing a prompt sentence" is a software component that has the function of parsing a prompt sentence and dividing it into multiple sub-prompt sentences.

[1003] A "sub-prompt sentence" is a detailed instruction for a specific task that is extracted as a result of parsing a prompt sentence.

[1004] A "means for aggregating results" is a software component that has the ability to combine individual results from a generative AI model into a single aggregated result.

[1005] The present invention provides a system that allows users to easily input prompt sentences and process complex tasks using multiple generative AI models. This system consists of a server and a user terminal.

[1006] server

[1007] The server has the functions of receiving prompt sentences from the user device, analyzing the received prompt sentences and dividing them into multiple sub-prompts, sending the sub-prompts to the appropriate generative AI model, and receiving the results from the generative AI model, integrating them, and sending them back to the user device.

[1008] User terminal

[1009] The user terminal provides an interface for the user to input a prompt sentence, which is then sent to the server, which processes the sentence and displays the result.

[1010] Specific implementation example

[1011] Entering a prompt statement

[1012] The user inputs a prompt sentence into the interface of the user terminal, such as "Create an image of a beautiful landscape and add a description of it." This prompt sentence is sent to the server as an HTTP request.

[1013] Parsing the prompt statement

[1014] The server analyzes the received prompt sentence. It uses natural language processing (NLP) technology to divide the prompt sentence into an image generation task and a text generation task. Specifically, it generates the following sub-prompt sentences:

[1015] Image generation subprompt: "Generate an image of a beautiful landscape."

[1016] Text generation subprompt: "Generate a description for that image."

[1017] Sending a subprompt statement

[1018] The server sends the image generation subprompt to the image generation model and the text generation subprompt to the text generation model. These models are generative AI models trained specifically for each task. The server sends the subprompt sentence as an API request and waits for a response from the model.

[1019] Acceptance and integration of results

[1020] The server receives the results from the generative AI models. For example, an image generation model might return an image of a beautiful landscape, and a text generation model might return a description of that image. It then combines these results and stores them in a single data format (e.g., JSON).

[1021] Returning the results

[1022] The server sends the integrated results to the user's device, which then displays the results on its interface. The user can then view the beautiful landscape images and their descriptions.

[1023] This system allows users to efficiently process complex tasks using multiple generative AI models and easily obtain integrated results. By clarifying what prompts will produce what results, the user experience is greatly improved.

[1024] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1025] Step 1:

[1026] Entering a prompt statement

[1027] The user inputs a prompt sentence through the interface of the user terminal.

[1028] Input: The prompt text entered by the user (e.g., "Generate an image of a beautiful landscape and add a description.")

[1029] Output: The user terminal generates an HTTP request to send the prompt text to the server.

[1030] Step 2:

[1031] Receiving a prompt

[1032] The server receives the prompt sentence sent from the user terminal.

[1033] Input: Prompt sent from the user's terminal to the server

[1034] Output: Stores the prompt text in an internal data structure.

[1035] Step 3:

[1036] Parsing the prompt statement

[1037] The server analyzes the received prompt sentence using natural language processing technology and divides it into multiple sub-prompt sentences.

[1038] Input: The received prompt (e.g., "Generate an image of a beautiful landscape and add a description.")

[1039] Output: Split sub-prompt sentences (e.g. "Generate an image of a beautiful landscape", "Generate a description of the image")

[1040] Step 4:

[1041] Sending a subprompt statement

[1042] The server sends the split subprompt sentences to the appropriate generative AI model. For example, an image generation subprompt is sent to an image generation model, and an explanatory sentence generation subprompt is sent to a text generation model.

[1043] Input: Separated subprompt sentence

[1044] Output: API request to the generative AI model

[1045] Step 5:

[1046] Acceptance of results

[1047] The server receives the results returned by the generative AI models: generated images from the image generation model, and generated explanatory text from the text generation model.

[1048] Input: Responses from each generative AI model (e.g., generated landscape images, generated descriptions)

[1049] Output: Stores the resulting data in an internal data structure.

[1050] Step 6:

[1051] Integration of results

[1052] The server then combines the results it receives, for example combining images and descriptions into a single data format (e.g., a JSON object).

[1053] Input: Accepted result data (e.g., generated landscape image, generated description)

[1054] Output: Consolidated results (e.g., a JSON object containing image URLs and descriptions)

[1055] Step 7:

[1056] Returning the results

[1057] The server returns the consolidated results to the user terminal.

[1058] Input: Consolidated results (e.g., JSON format data)

[1059] Output: HTTP response to the user's device

[1060] Step 8:

[1061] Displaying the results

[1062] The user terminal receives the results returned from the server and displays them to the user.

[1063] Input: Integration results sent from the server (e.g., JSON format data)

[1064] Output: The result shown to the user (e.g., a landscape image and its description)

[1065] Through the above process, users can efficiently obtain the results of complex tasks that go through multiple generative AI models.

[1066] (Application example 1)

[1067] 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."

[1068] Conventional customer support systems for virtual stores have been limited to using individual models or simple FAQ databases to answer user questions, making it difficult to provide integrated, advanced answers. As a result, there has been a problem in that it takes a lot of time and effort for users to obtain the information they are looking for.

[1069] 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.

[1070] In this invention, the server includes means for receiving a prompt from a user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, means for returning the integrated results to the user, and means for analyzing questions about products in a virtual store and providing related information, thereby enabling the user to obtain detailed and comprehensive information about products by inputting a prompt once.

[1071] "Means for receiving prompts from a user" refers to a device or software function that allows the system to receive questions or requests entered by a user.

[1072] The "means for analyzing a prompt and dividing it into a plurality of sub-prompts based on the prompt" refers to the function of a device or software that understands a received prompt and breaks it down into a plurality of specific tasks.

[1073] The "means for sending sub-prompts to respective artificial intelligence models" refers to a device or software function that sends the divided sub-prompts to the appropriate artificial intelligence models so that they can be processed.

[1074] "Means for receiving results from an artificial intelligence model" refers to a device or software function for receiving results generated by an artificial intelligence model.

[1075] "Means for integrating received results" refers to the function of a device or software that combines the results received from each artificial intelligence model into a single, consistent result.

[1076] A "means for returning the consolidated results to the user" is a device or software function that displays or presents the consolidated results to the user.

[1077] "Means for analyzing product-related questions within a virtual store and providing related information" refers to a device or software function for analyzing product-related questions entered by users in a virtual store and providing related information in response to those questions.

[1078] An "image generation model" is a type of artificial intelligence that generates images based on input text prompts.

[1079] A "sentence generation model" is a type of artificial intelligence that generates sentences based on input text prompts.

[1080] A "review analysis model" is a type of artificial intelligence that analyzes product reviews and summarizes or evaluates their content.

[1081] The present invention is a system that allows users to easily order key prompts and coordinates prompts across multiple artificial intelligence models. This system aims to enable multiple artificial intelligence models to cooperate and generate answers to questions about products asked by users, particularly in virtual stores.

[1082] The main components of the system are as follows:

[1083] 1. Accepting the prompt

[1084] The user inputs a question about a product in the virtual store. For example, the user inputs a prompt such as "How is the camera performance of this smartphone?" This prompt is sent from the user terminal to the server.

[1085] 2. Parsing the prompt

[1086] The server parses the received prompt and divides it into multiple sub-prompts based on different tasks, for example, "analyzing image samples from a camera," "presenting technical specifications of a camera," and "summarizing user reviews."

[1087] 3. Generate and send a subprompt

[1088] Based on the analysis results, the server generates sub-prompts appropriate for each task and sends them to each AI model, such as "Generate an image sample from the camera," "Explain the camera's technical specifications," or "Summarize user reviews about the camera's performance."

[1089] 4. Acceptance of Results

[1090] The server receives the output results from each artificial intelligence model (image generation model, sentence generation model, review analysis model).

[1091] 5. Synthesis of results

[1092] The server then combines the different results it receives to produce a single, coherent answer, which includes image samples of the camera's capabilities, detailed technical specifications, and highlights from user reviews.

[1093] 6. Return of results

[1094] Finally, the server sends the integrated results back to the user device, which then displays them to the user. For example, in response to a prompt entered by the user, a comprehensive answer such as "This smartphone's camera performance is excellent, and it takes clear photos even at night. In addition, many users rate its ease of use highly."

[1095] Program processing explanation

[1096] The system for implementing the present invention uses the following hardware and software. User terminals include smartphones, tablets, and PCs, and the server uses software such as Flask (a Python web framework), OpenAI API, PyTorch, and Hugging Face Transformers. Users enter questions about products in the virtual store, and the prompts are sent to the server.

[1097] The server analyzes the prompts and sends sub-prompts to the appropriate artificial intelligence models. The image generation model generates images based on the input text prompt, the sentence generation model generates sentences based on the text prompt, and the review analysis model analyzes product reviews and summarizes or rates their content. The server consolidates these results and returns them to the user.

[1098] Specific examples

[1099] For example, a user enters the prompt "How good is the camera on this phone?" The prompt is split into:

[1100] "Generate a sample image from the camera"

[1101] "Please explain the technical specifications of the camera."

[1102] "Please summarize user reviews of the camera's performance."

[1103] The results of the generative AI model based on these sub-prompts are integrated and sent back to the user, allowing the user to obtain comprehensive information with a single prompt input.

[1104] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1105] Step 1:

[1106] A user enters a question about a product in a virtual store

[1107] A user inputs a prompt such as "How is the camera performance of this smartphone?" into the user terminal and sends the question to the server. The input is a text prompt, and the output is the prompt data transferred to the server.

[1108] Step 2:

[1109] The server receives a prompt

[1110] The server receives prompts sent from the user terminal and stores them as input data for parsing. The input is the raw prompt text sent by the user, and the output is the prompt to be parsed.

[1111] Step 3:

[1112] The server parses the prompt and splits it into subprompts

[1113] The server analyzes the received prompt and divides it into multiple sub-prompts based on different tasks, such as "analyzing camera image samples," "presenting camera specifications," and "summarizing user reviews." The input is the prompt text, and the output is a set of sub-prompts corresponding to each task.

[1114] Step 4:

[1115] Send sub-prompts to each AI model

[1116] The server sends the generated sub-prompts to each AI model. For example, it sends a sub-prompt such as "Generate an image sample of the camera" to the image generation model, and a sub-prompt such as "Describe the technical specifications of the camera" to the sentence generation model. The input is the sub-prompt, and the output is the task sent to each model.

[1117] Step 5:

[1118] The server receives the results from each AI model.

[1119] The server receives the output from each AI model. The image generation model generates sample images of the camera, the sentence generation model generates technical specifications for the camera, and the review analysis model generates summaries of user reviews. The input is the generated results based on the sub-prompts, and the output is the resulting data from each model.

[1120] Step 6:

[1121] The server aggregates the results

[1122] The server then combines the results it receives to generate a single, coherent answer. For example, it might combine the generated camera image, technical specifications, and review summary into a single response. The input is the result data from each model, and the output is the combined answer.

[1123] Step 7:

[1124] The server sends the consolidated results back to the user

[1125] The server sends the integrated results to the user's device. The user's device displays the results to the user, allowing the user to obtain comprehensive information about the product. For example, the user's device might display a message such as, "This smartphone's camera performance is excellent, and it takes clear photos even at night. Many users also rate its ease of use highly." The input is the integrated result data, and the output is the content displayed on the user's device.

[1126] 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.

[1127] This invention provides a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and influences the analysis results.

[1128] Program processing

[1129] server

[1130] 1. Accepting the prompt

[1131] The server receives the prompt sent from the user terminal.

[1132] Example: A user enters the prompt "Generate a poem to comfort me when I'm sad."

[1133] 2. Emotion Analysis

[1134] The server is equipped with an emotion engine that analyzes text data from the user's input prompts to recognize the user's emotions.

[1135] Example: The emotion "sad" is recognized from the input prompt.

[1136] 3. Parsing the prompt

[1137] The server parses the received prompt, understands its content and splits it into appropriate tasks.

[1138] Example: Identify the task "Generate a poem."

[1139] 4. Generating a subprompt

[1140] The server generates sub-prompts for each task based on the analysis results, taking into account the analysis results of the emotion engine.

[1141] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[1142] 5. Sending a sub-prompt

[1143] The server sends the generated sub-prompts to the appropriate artificial intelligence model.

[1144] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[1145] 6. Acceptance of Results

[1146] The server receives the output from the artificial intelligence model.

[1147] Example: Accepting a poem generated from a poetry generation model.

[1148] 7. Synthesis of results

[1149] The server aggregates the results it receives.

[1150] Example: Compiling the generated poems into a single data package.

[1151] 8. Return of results

[1152] The server transmits the integrated results to the user terminal.

[1153] User terminal

[1154] 1. Enter the prompt

[1155] The user enters a prompt into the interface.

[1156] Example: User types "Generate a comforting poem for when I'm sad."

[1157] 2. Sending a prompt

[1158] The user terminal sends the entered prompt to the server.

[1159] 3. Displaying the results

[1160] The results returned by the server are displayed to the user.

[1161] Example: A poem to comfort you when you're feeling sad is displayed.

[1162] Specific examples

[1163] Example 1: Emotion-based poetry generation process

[1164] The user enters the prompt "Generate a comforting poem for sad feelings."

[1165] The user terminal sends a prompt to the server.

[1166] The server receives the prompt and recognizes the emotion "sad" using an emotion engine.

[1167] The server parses the prompt and identifies the task "Generate Poetry."

[1168] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[1169] The server receives the poem generated from the poetry generation model.

[1170] The server aggregates the generated poems and transmits them to the user's terminal.

[1171] The user terminal displays the received poem to the user.

[1172] The user checks the generated poem and experiences changes in emotion.

[1173] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling users to easily complete complex tasks.

[1174] The processing flow will be explained below.

[1175] Step 1:

[1176] The user enters a prompt into the interface.

[1177] Example: Type "Generate a poem to comfort me when I'm sad."

[1178] Step 2:

[1179] The terminal sends the prompt entered by the user to the server.

[1180] For example: A prompt is sent to the server.

[1181] Step 3:

[1182] The server receives the prompt.

[1183] Example: Accept the prompt "Generate a poem to comfort someone who is sad."

[1184] Step 4:

[1185] The server uses an emotion engine to analyze the user's emotion from the text data of the prompt.

[1186] Example: Analyze keywords such as "sad" in the prompt and recognize that the user's emotion is "sadness."

[1187] Step 5:

[1188] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[1189] Example: Parse the prompt and identify the task "Generate a poem."

[1190] Step 6:

[1191] The server generates sub-prompts corresponding to each task, taking into account the emotional information obtained from the emotion engine.

[1192] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[1193] Step 7:

[1194] The server sends the generated sub-prompts to the selected artificial intelligence model.

[1195] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[1196] Step 8:

[1197] The server receives the output from the artificial intelligence model.

[1198] Example: Accepting a poem generated from a poetry generation model.

[1199] Step 9:

[1200] The server consolidates the results it receives.

[1201] Example: Compiling the generated poems into a single data package.

[1202] Step 10:

[1203] The server transmits the integrated results to the user terminal.

[1204] Example: Sending a consolidated poetry data package to a user device.

[1205] Step 11:

[1206] The terminal displays the results received from the server to the user.

[1207] Example: "Poems to comfort sad feelings" is displayed on the user's interface.

[1208] Step 12:

[1209] The user checks the displayed results.

[1210] Example: Read the generated poem and experience the changes in emotions.

[1211] This system analyzes the user's emotions from the prompt text data and uses the results to generate more personalized results so that the user can get appropriate content that takes their emotions into consideration.

[1212] Example 2

[1213] 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."

[1214] Currently, many artificial intelligence systems exist, but they generally use a single model to respond to prompts entered by the user. However, it is difficult for a single model to generate outputs that adequately reflect the user's emotions. This creates a need for systems that can provide a truly personalized experience for users. Furthermore, they lack the ability to automatically identify multiple tasks and generate appropriate sub-prompts. A system that can solve these problems is needed.

[1215] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1216] In this invention, the server includes means including an emotion engine that recognizes a user's emotions and influences the analysis result, means for receiving a prompt from the user, means for analyzing the received prompt and dividing it into multiple sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, and means for returning the integrated results to the user. This makes it possible to provide a truly personalized output by taking the user's emotions into consideration and appropriately processing multiple tasks.

[1217] A "prompt" is text that a user enters into a system as an instruction or question.

[1218] A "subprompt" is a more specific instruction for performing a particular task that is generated from a parsed prompt.

[1219] An "artificial intelligence model" is a machine learning algorithm that has been trained to perform a specific task automatically.

[1220] An "emotion engine" is a part of a system that analyzes user input text and identifies the emotions contained in that text.

[1221] A "task" is an individual task or operation that a system must perform.

[1222] "Results synthesis" is the process of combining the individual outputs from multiple artificial intelligence models into a single, coherent result.

[1223] This invention is a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and affects the analysis results.

[1224] Hardware and software configuration

[1225] The system of the present invention uses the following hardware and software.

[1226] Server: A server computer with a high-performance processor and a large amount of memory. For example, you can use a cloud-based server (AWS, Google Cloud, Microsoft Azure, etc.).

[1227] User terminal: A device used by a user to access a system. Examples include personal computers, smartphones, and tablets.

[1228] Emotion engine: Software that uses natural language processing techniques to analyze user emotions, for example, by using open source NLP libraries (such as Hugging Face's Transformers).

[1229] Artificial intelligence models: Generative AI models use machine learning models (e.g., GPT-3, BERT) that are trained to perform specific tasks.

[1230] System Operation

[1231] 1. User prompt input

[1232] The user enters a prompt into the interface, for example, text such as "Generate a poem to comfort me when I'm feeling sad."

[1233] 2. Sending a prompt

[1234] The user terminal sends the input prompt to the server using an HTTP POST request.

[1235] 3. Accepting and parsing prompts

[1236] The server receives a prompt sent from a user terminal, identifies the emotion of the prompt using an emotion engine, and then analyzes the prompt and divides it into multiple tasks, for example, identifying a task called "poetry creation."

[1237] 4. Generating a subprompt

[1238] Based on the analysis results, the server generates sub-prompts for each task, taking into account the results of the emotion engine. Specifically, it generates specific sub-prompts such as "Please generate a poem that will comfort sad feelings."

[1239] 5. Submitting a subprompt and accepting the results

[1240] The server sends the generated sub-prompts to an artificial intelligence model, such as a poetry generation model, and receives output from the model.

[1241] 6. Consolidating and returning results

[1242] The server integrates the results obtained from multiple tasks and sends them back to the user's device, which receives them and displays them on the screen.

[1243] Specific examples

[1244] Example 1: Emotion-based poetry generation process

[1245] The user enters the prompt "Generate a comforting poem for sad feelings."

[1246] The user terminal sends a prompt to the server.

[1247] The server receives the prompt and identifies the emotion "sad" through an emotion engine.

[1248] The server parses the prompt and identifies the task "Generate Poetry."

[1249] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[1250] The server receives the poem generated from the poetry generation model.

[1251] The server aggregates the generated poems and transmits them to the user's terminal.

[1252] The user terminal displays the received poem to the user.

[1253] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling complex tasks to be completed easily.

[1254] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1255] Step 1:

[1256] The user terminal sends a prompt to the server. Specifically, when the user enters "Create a poem to comfort me when I'm sad" into the interface and presses the submit button, the system sends an HTTP POST request to the server. The input is a text prompt, and the output is a request sent to the server.

[1257] Step 2:

[1258] The server receives a prompt sent from the user terminal. The server receives an HTTP POST request and parses the text prompt included in the request body. At this time, the server extracts the prompt text and passes it on to subsequent processing. The input is the prompt text sent by the user, and the output is the prompt data to be parsed.

[1259] Step 3:

[1260] The emotion engine on the server analyzes the prompt and recognizes the user's emotion. It uses a natural language processing (NLP) library to tokenize the prompt and then runs it through an emotion classifier to extract an emotion tag. Specifically, it recognizes the emotion "sad." In this process, the input is the prompt text and the output is an emotion tag (e.g., "sad").

[1261] Step 4:

[1262] The server parses the prompt and splits it into tasks. NLP techniques are used to parse the text and identify tasks (e.g., "Generate Poetry"). The input is the prompt text with emotion tags, and the output is the identified tasks.

[1263] Step 5:

[1264] Based on the analysis results, the server generates sub-prompts corresponding to each task, taking into account the results of the emotion engine. For example, it generates a sub-prompt such as "Please create a poem that will comfort me when I'm feeling sad." In this process, the input is the identified task and emotion tag, and the output is the sub-prompt.

[1265] Step 6:

[1266] The server sends the generated sub-prompt to the appropriate AI model. It sends the sub-prompt via an HTTP request to the poetry generation model's API endpoint and waits for a response. The input is the sub-prompt, and the output is a request sent to the poetry generation model.

[1267] Step 7:

[1268] The server receives the output from the AI ​​model, analyzes the HTTP response from the poetry generation model, and obtains the generated poem. The input is the response data from the poetry generation model, and the output is the text of the generated poem.

[1269] Step 8:

[1270] The server aggregates the results from multiple tasks, compiles the generated poems into a single data package, and formats it in a format that can be sent back to the user. The input is the multiple output results, and the output is the aggregated result data.

[1271] Step 9:

[1272] The server returns the integrated results to the user terminal. An HTTP response is generated and the integrated result data is sent to the user terminal. The input is the integrated result data, and the output is the response sent to the user terminal.

[1273] Step 10:

[1274] The user terminal displays the results returned from the server to the user. The HTTP response is received and its contents are displayed on the screen. As a concrete example, the generated poem is displayed on the user's interface. The input is the response data from the server, and the output is the result displayed to the user.

[1275] (Application example 2)

[1276] 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."

[1277] Existing food delivery systems can respond quickly to user orders, but they cannot provide personalized suggestions that take into account the user's emotions and mood. This limits their ability to improve user satisfaction. In particular, there is a need for systems that provide meal suggestions that take into account emotions such as stress and fatigue.

[1278] The specification process 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 receiving a prompt from the user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to each AI model, means for analyzing the user's emotions, and means for generating sub-prompts based on the emotion analysis results and sending them to the AI ​​model. This enables personalized meal suggestions tailored to the user's emotions.

[1279] A "prompt" refers to the text of a request or question entered by a user into the system.

[1280] A "sub-prompt" is a prompt that you receive broken down into multiple smaller tasks.

[1281] "Artificial intelligence model" refers to a machine learning algorithm for performing a specific task, including generative and propositional models.

[1282] "Sentiment analysis" refers to the process of identifying and analyzing emotions from a user's input prompts.

[1283] A "generative model" refers to an AI algorithm that automatically generates new text or images based on input requests.

[1284] "Proposal model" refers to an AI algorithm that provides appropriate suggestions and options based on the analysis results.

[1285] "Synthesis" refers to the process of combining the results of multiple sub-prompts into one.

[1286] The present invention realizes a system that analyzes prompts from a user, generates multiple sub-prompts based on the emotional analysis results, and sends them to multiple artificial intelligence models to provide personalized results to the user. To implement this invention, a server, a user terminal, and various software are required.

[1287] Hardware and software used

[1288] Hardware

[1289] server

[1290] User device (smartphone)

[1291] software

[1292] Sentiment analysis engines (e.g., Python-based NLP libraries such as spaCy and NLTK)

[1293] Artificial intelligence models (e.g., GPT-4 or Transformers for running generative and propositional models)

[1294] A database (for temporarily storing results)

[1295] System configuration

[1296] 1. Prompt input from the user terminal

[1297] A user uses a food delivery application to enter a prompt, which can be entered as text or speech. For example, "I'm feeling stressed today, can you recommend a comforting meal?"

[1298] 2. Accepting the prompt

[1299] The user terminal sends the input prompt to the server, which receives it and inputs it into the emotion analysis engine.

[1300] 3. Emotion Analysis

[1301] The server uses an emotion analysis engine (such as spaCy or NLTK) to analyze the text in the prompt and identify the user's emotion. For example, it analyzes the emotion "stress" from the prompt "I'm stressed."

[1302] 4. Parsing the prompt

[1303] The server splits the prompt into multiple sub-prompts, taking into account the results of the sentiment analysis, using a text analysis library or similar.

[1304] 5. Generating Subprompts

[1305] Based on the results of the sentiment analysis, the server generates an appropriate sub-prompt, such as "Please suggest some dishes that will help relieve stress."

[1306] 6. Sending sub-prompts

[1307] The server then sends the generated sub-prompts to an appropriate artificial intelligence model, such as a generative model that suggests dishes.

[1308] 7. Acceptance and Integration of Results

[1309] The server receives and integrates the results from the artificial intelligence models, a process that combines multiple results into a single data package.

[1310] 8. Return of results

[1311] The server sends the integrated results to the user's device, which then displays, for example, a list of recommended dishes for relieving stress.

[1312] Specific examples

[1313] User: Enters the prompt "I'm feeling stressed today, what's a comforting dish?"

[1314] User terminal: Sends the prompt to the server.

[1315] Server: Receives the prompt, performs emotion analysis, and identifies the emotion "stress."

[1316] Server: Based on the analysis results, generate a sub-prompt such as "Please suggest a dish that will help relieve stress" and send it to the dish suggestion model.

[1317] Server: Receives the resulting proposals, integrates them, and sends them back to the user device.

[1318] User terminal: The suggested dishes are displayed to the user.

[1319] This process allows users to receive personalized meal suggestions that match their emotions, realizing a food delivery service that meets a variety of needs.

[1320] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1321] Step 1:

[1322] A user opens a food delivery application on a user device such as a smartphone and inputs a prompt such as, "I'm feeling stressed today, so please recommend some soothing dishes." This user input becomes the initial prompt data.

[1323] Step 2:

[1324] The user terminal sends the input prompt to the server. The prompt data sent to the server is "I'm feeling stressed today, so please tell me some soothing dishes."

[1325] Step 3:

[1326] The server receives the prompt and uses a sentiment analysis engine (e.g., spaCy or NLTK) to analyze the text in the prompt and identify the user's sentiment. In this case, the resulting sentiment is "stress."

[1327] Step 4:

[1328] The server analyzes the prompt based on the results of the sentiment analysis and divides it into appropriate sub-prompts. In this case, the sub-prompt generated is "Please suggest a dish that will help relieve stress."

[1329] Step 5:

[1330] The server sends the generated sub-prompts to a dish suggestion model (e.g., GPT-4 or Transformers), which receives the sub-prompts as input and then outputs dish suggestions based on them.

[1331] Step 6:

[1332] The server receives the results of the dish recommendation model, such as a list of dishes like "herbal tea," "banana smoothie," and "yogurt parfait."

[1333] Step 7:

[1334] The server aggregates the multiple recommendations into a single data package, in this case, a list of recommended dishes.

[1335] Step 8:

[1336] The server transmits the integrated results to the user terminal, and the transmitted data includes the integrated recipe list.

[1337] Step 9:

[1338] The user device displays the received list of dishes to the user, who can then check suggested dishes such as "herbal tea," "banana smoothie," and "yogurt parfait" on the application.

[1339] The above steps realize a system that allows users to get personalized recipe suggestions based on their emotions.

[1340] 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.

[1341] 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.

[1342] 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.

[1343] [Fourth embodiment]

[1344] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1345] 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.

[1346] 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).

[1347] 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.

[1348] 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.

[1349] 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).

[1350] 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.

[1351] 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.

[1352] 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.

[1353] 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.

[1354] 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.

[1355] 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.

[1356] 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."

[1357] The present invention provides a system that allows a user to easily order a primary prompt and coordinates the prompt across multiple artificial intelligence models, including means for receiving a prompt from a user, means for parsing the prompt and dividing it into multiple sub-prompts, means for sending the sub-prompts to respective artificial intelligence models, means for receiving results from the artificial intelligence models, means for aggregating the results, and means for returning the aggregated results to the user.

[1358] Program processing

[1359] server

[1360] 1. Accepting the prompt

[1361] The server receives the prompt sent from the user terminal.

[1362] Example: A user enters the prompt "Generate an image of a beautiful landscape and add a description."

[1363] 2. Parsing the prompt

[1364] The server analyzes the received prompts and splits them into different tasks (e.g., image generation, description generation).

[1365] 3. Generating a subprompt

[1366] The server generates sub-prompts for each task based on the analysis results.

[1367] Example: Generate a sub-prompt for "Generate an image of a beautiful landscape" and another for "Generate a description of the image."

[1368] 4. Sending a sub-prompt

[1369] The server sends the generated sub-prompt to the appropriate artificial intelligence model, for example, sending a sub-prompt such as "Generate an image of a beautiful landscape" to an image generation model, and a sub-prompt such as "Generate a description of the image" to a text generation model.

[1370] 5. Acceptance of Results

[1371] The server receives the output results from each artificial intelligence model.

[1372] Example: Receive a landscape image generated from an image generation model and a description generated from a text generation model.

[1373] 6. Synthesis of results

[1374] The server combines the received images and descriptions into a single consolidated result.

[1375] Example: The integrated results include a landscape image and its description.

[1376] 7. Return of results

[1377] The server transmits the integrated results to the user terminal.

[1378] User terminal

[1379] 1. Enter the prompt

[1380] The user enters a prompt into the interface.

[1381] Example: User types "Generate an image of a beautiful landscape and add a description."

[1382] 2. Sending a prompt

[1383] The user terminal sends the entered prompt to the server.

[1384] 3. Displaying the results

[1385] The results returned by the server are displayed to the user.

[1386] Example: A beautiful landscape image and its description are displayed.

[1387] Specific examples

[1388] Example 1: Image and description generation process

[1389] The user enters the prompt "Create an image of a beautiful landscape and add a description."

[1390] The user terminal sends a prompt to the server.

[1391] The server receives the prompt and splits it into an image generation task and a description generation task.

[1392] The server sends a sub-prompt to the image generation model saying, "Generate an image of a beautiful landscape," and a sub-prompt to the text generation model saying, "Generate a description of the image."

[1393] The server receives the landscape image generated from the image generation model and the explanatory text generated from the text generation model.

[1394] The server aggregates these results and sends them to the user terminal.

[1395] The user terminal displays the received landscape image and explanatory text on the interface.

[1396] The user checks the generated image and description.

[1397] This invention provides a system that allows users to seamlessly utilize multiple artificial intelligence models and efficiently process complex tasks, enabling users to easily obtain high-quality results.

[1398] The processing flow will be explained below.

[1399] Step 1:

[1400] The user enters a prompt into the interface.

[1401] For example, type "Generate an image of a beautiful landscape and add a description."

[1402] Step 2:

[1403] The terminal sends the prompt entered by the user to the server.

[1404] For example: A prompt is sent to the server.

[1405] Step 3:

[1406] The server receives the prompt.

[1407] Example: Accept the prompt "Generate an image of a beautiful landscape and add a description."

[1408] Step 4:

[1409] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[1410] Example: Split into tasks "image generation" and "description generation".

[1411] Step 5:

[1412] The server generates sub-prompts corresponding to each task.

[1413] Example: Generate the sub-prompts "Generate an image of a beautiful landscape" and "Generate a description of the generated image."

[1414] Step 6:

[1415] The server sends the generated sub-prompts to the selected artificial intelligence model.

[1416] Example: Send an image generation model the sub-prompt "Generate an image of a beautiful landscape," and a text generation model the sub-prompt "Generate a description of the generated image."

[1417] Step 7:

[1418] The server receives the output results from each artificial intelligence model.

[1419] Example: Accepts beautiful landscape images generated from an image generation model and explanatory text generated from a text generation model.

[1420] Step 8:

[1421] The server consolidates the results it receives.

[1422] Example: A landscape image and its description are packaged together as a single data package.

[1423] Step 9:

[1424] The server transmits the integrated results to the user terminal.

[1425] Example: Sending a synthesized landscape image and its description to a user device.

[1426] Step 10:

[1427] The terminal displays the results received from the server to the user.

[1428] Example: A user interface displays an image of a beautiful landscape and a description of it.

[1429] Step 11:

[1430] The user checks the displayed results.

[1431] Example: View and evaluate the generated landscape images and their descriptions.

[1432] Example 1

[1433] 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."

[1434] In today's world, users are required to efficiently process complex tasks using multiple AI models, but existing systems lack a simple way to achieve this. A method is needed that allows users to seamlessly execute different tasks, such as image generation and text generation, within a single system, and then integrate and efficiently use the results.

[1435] 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.

[1436] In this invention, the server includes means for receiving a prompt sentence from a user terminal, means for analyzing the received prompt sentence and dividing it into multiple sub-prompt sentences based on the prompt sentence, means for sending the sub-prompt sentences to each generative AI model, means for receiving results from the generative AI models, means for integrating the received results, and means for returning the integrated results to the user terminal, thereby enabling a user to efficiently process complex tasks using multiple AI models and easily obtain the integrated results.

[1437] A "prompt" is a textual instruction for a user to enter a specific task or request.

[1438] A "user terminal" is a device (e.g., a computer, smartphone, tablet, etc.) that a user uses to input prompt sentences and communicate with a server.

[1439] The "server" is a central computer system that receives prompts, analyzes them, distributes tasks to generative AI models, and synthesizes and returns the results to the user device.

[1440] A "generative AI model" is an artificial intelligence model that is trained to perform a specific task (e.g., image generation or text generation).

[1441] An "image generation model" is an artificial intelligence model that is trained to generate images based on input prompts.

[1442] A "text generation model" is an artificial intelligence model that is trained to generate natural language text based on an input prompt.

[1443] The "means for parsing a prompt sentence" is a software component that has the function of parsing a prompt sentence and dividing it into multiple sub-prompt sentences.

[1444] A "sub-prompt sentence" is a detailed instruction for a specific task that is extracted as a result of parsing a prompt sentence.

[1445] A "means for aggregating results" is a software component that has the ability to combine individual results from a generative AI model into a single aggregated result.

[1446] The present invention provides a system that allows users to easily input prompt sentences and process complex tasks using multiple generative AI models. This system consists of a server and a user terminal.

[1447] server

[1448] The server has the functions of receiving prompt sentences from the user device, analyzing the received prompt sentences and dividing them into multiple sub-prompts, sending the sub-prompts to the appropriate generative AI model, and receiving the results from the generative AI model, integrating them, and sending them back to the user device.

[1449] User terminal

[1450] The user terminal provides an interface for the user to input a prompt sentence, which is then sent to the server, which processes the sentence and displays the result.

[1451] Specific implementation example

[1452] Entering a prompt statement

[1453] The user inputs a prompt sentence into the interface of the user terminal, such as "Create an image of a beautiful landscape and add a description of it." This prompt sentence is sent to the server as an HTTP request.

[1454] Parsing the prompt statement

[1455] The server analyzes the received prompt sentence. It uses natural language processing (NLP) technology to divide the prompt sentence into an image generation task and a text generation task. Specifically, it generates the following sub-prompt sentences:

[1456] Image generation subprompt: "Generate an image of a beautiful landscape."

[1457] Text generation subprompt: "Generate a description for that image."

[1458] Sending a subprompt statement

[1459] The server sends the image generation subprompt to the image generation model and the text generation subprompt to the text generation model. These models are generative AI models trained specifically for each task. The server sends the subprompt sentence as an API request and waits for a response from the model.

[1460] Acceptance and integration of results

[1461] The server receives the results from the generative AI models. For example, an image generation model might return an image of a beautiful landscape, and a text generation model might return a description of that image. It then combines these results and stores them in a single data format (e.g., JSON).

[1462] Returning the results

[1463] The server sends the integrated results to the user's device, which then displays the results on its interface. The user can then view the beautiful landscape images and their descriptions.

[1464] This system allows users to efficiently process complex tasks using multiple generative AI models and easily obtain integrated results. By clarifying what prompts will produce what results, the user experience is greatly improved.

[1465] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1466] Step 1:

[1467] Entering a prompt statement

[1468] The user inputs a prompt sentence through the interface of the user terminal.

[1469] Input: The prompt text entered by the user (e.g., "Generate an image of a beautiful landscape and add a description.")

[1470] Output: The user terminal generates an HTTP request to send the prompt text to the server.

[1471] Step 2:

[1472] Receiving a prompt

[1473] The server receives the prompt sentence sent from the user terminal.

[1474] Input: Prompt sent from the user's terminal to the server

[1475] Output: Stores the prompt text in an internal data structure.

[1476] Step 3:

[1477] Parsing the prompt statement

[1478] The server analyzes the received prompt sentence using natural language processing technology and divides it into multiple sub-prompt sentences.

[1479] Input: The received prompt (e.g., "Generate an image of a beautiful landscape and add a description.")

[1480] Output: Split sub-prompt sentences (e.g. "Generate an image of a beautiful landscape", "Generate a description of the image")

[1481] Step 4:

[1482] Sending a subprompt statement

[1483] The server sends the split subprompt sentences to the appropriate generative AI model. For example, an image generation subprompt is sent to an image generation model, and an explanatory sentence generation subprompt is sent to a text generation model.

[1484] Input: Separated subprompt sentence

[1485] Output: API request to the generative AI model

[1486] Step 5:

[1487] Acceptance of results

[1488] The server receives the results returned by the generative AI models: generated images from the image generation model, and generated explanatory text from the text generation model.

[1489] Input: Responses from each generative AI model (e.g., generated landscape images, generated descriptions)

[1490] Output: Stores the resulting data in an internal data structure.

[1491] Step 6:

[1492] Integration of results

[1493] The server then combines the results it receives, for example combining images and descriptions into a single data format (e.g., a JSON object).

[1494] Input: Accepted result data (e.g., generated landscape image, generated description)

[1495] Output: Consolidated results (e.g., a JSON object containing image URLs and descriptions)

[1496] Step 7:

[1497] Returning the results

[1498] The server returns the consolidated results to the user terminal.

[1499] Input: Consolidated results (e.g., JSON format data)

[1500] Output: HTTP response to the user's device

[1501] Step 8:

[1502] Displaying the results

[1503] The user terminal receives the results returned from the server and displays them to the user.

[1504] Input: Integration results sent from the server (e.g., JSON format data)

[1505] Output: The result shown to the user (e.g., a landscape image and its description)

[1506] Through the above process, users can efficiently obtain the results of complex tasks that go through multiple generative AI models.

[1507] (Application example 1)

[1508] 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."

[1509] Conventional customer support systems for virtual stores have been limited to using individual models or simple FAQ databases to answer user questions, making it difficult to provide integrated, advanced answers. As a result, there has been a problem in that it takes a lot of time and effort for users to obtain the information they are looking for.

[1510] 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.

[1511] In this invention, the server includes means for receiving a prompt from a user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, means for returning the integrated results to the user, and means for analyzing questions about products in a virtual store and providing related information, thereby enabling the user to obtain detailed and comprehensive information about products by inputting a prompt once.

[1512] "Means for receiving prompts from a user" refers to a device or software function that allows the system to receive questions or requests entered by a user.

[1513] The "means for analyzing a prompt and dividing it into a plurality of sub-prompts based on the prompt" refers to the function of a device or software that understands a received prompt and breaks it down into a plurality of specific tasks.

[1514] The "means for sending sub-prompts to respective artificial intelligence models" refers to a device or software function that sends the divided sub-prompts to the appropriate artificial intelligence models so that they can be processed.

[1515] "Means for receiving results from an artificial intelligence model" refers to a device or software function for receiving results generated by an artificial intelligence model.

[1516] "Means for integrating received results" refers to the function of a device or software that combines the results received from each artificial intelligence model into a single, consistent result.

[1517] A "means for returning the consolidated results to the user" is a device or software function that displays or presents the consolidated results to the user.

[1518] "Means for analyzing product-related questions within a virtual store and providing related information" refers to a device or software function for analyzing product-related questions entered by users in a virtual store and providing related information in response to those questions.

[1519] An "image generation model" is a type of artificial intelligence that generates images based on input text prompts.

[1520] A "sentence generation model" is a type of artificial intelligence that generates sentences based on input text prompts.

[1521] A "review analysis model" is a type of artificial intelligence that analyzes product reviews and summarizes or evaluates their content.

[1522] The present invention is a system that allows users to easily order key prompts and coordinates prompts across multiple artificial intelligence models. This system aims to enable multiple artificial intelligence models to cooperate and generate answers to questions about products asked by users, particularly in virtual stores.

[1523] The main components of the system are as follows:

[1524] 1. Accepting the prompt

[1525] The user inputs a question about a product in the virtual store. For example, the user inputs a prompt such as "How is the camera performance of this smartphone?" This prompt is sent from the user terminal to the server.

[1526] 2. Parsing the prompt

[1527] The server parses the received prompt and divides it into multiple sub-prompts based on different tasks, for example, "analyzing image samples from a camera," "presenting technical specifications of a camera," and "summarizing user reviews."

[1528] 3. Generate and send a subprompt

[1529] Based on the analysis results, the server generates sub-prompts appropriate for each task and sends them to each AI model, such as "Generate an image sample from the camera," "Explain the camera's technical specifications," or "Summarize user reviews about the camera's performance."

[1530] 4. Acceptance of Results

[1531] The server receives the output results from each artificial intelligence model (image generation model, sentence generation model, review analysis model).

[1532] 5. Synthesis of results

[1533] The server then combines the different results it receives to produce a single, coherent answer, which includes image samples of the camera's capabilities, detailed technical specifications, and highlights from user reviews.

[1534] 6. Return of results

[1535] Finally, the server sends the integrated results back to the user device, which then displays them to the user. For example, in response to a prompt entered by the user, a comprehensive answer such as "This smartphone's camera performance is excellent, and it takes clear photos even at night. In addition, many users rate its ease of use highly."

[1536] Program processing explanation

[1537] The system for implementing the present invention uses the following hardware and software. User terminals include smartphones, tablets, and PCs, and the server uses software such as Flask (a Python web framework), OpenAI API, PyTorch, and Hugging Face Transformers. Users enter questions about products in the virtual store, and the prompts are sent to the server.

[1538] The server analyzes the prompts and sends sub-prompts to the appropriate artificial intelligence models. The image generation model generates images based on the input text prompt, the sentence generation model generates sentences based on the text prompt, and the review analysis model analyzes product reviews and summarizes or rates their content. The server consolidates these results and returns them to the user.

[1539] Specific examples

[1540] For example, a user enters the prompt "How good is the camera on this phone?" The prompt is split into:

[1541] "Generate a sample image from the camera"

[1542] "Please explain the technical specifications of the camera."

[1543] "Please summarize user reviews of the camera's performance."

[1544] The results of the generative AI model based on these sub-prompts are integrated and sent back to the user, allowing the user to obtain comprehensive information with a single prompt input.

[1545] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1546] Step 1:

[1547] A user enters a question about a product in a virtual store

[1548] A user inputs a prompt such as "How is the camera performance of this smartphone?" into the user terminal and sends the question to the server. The input is a text prompt, and the output is the prompt data transferred to the server.

[1549] Step 2:

[1550] The server receives a prompt

[1551] The server receives prompts sent from the user terminal and stores them as input data for parsing. The input is the raw prompt text sent by the user, and the output is the prompt to be parsed.

[1552] Step 3:

[1553] The server parses the prompt and splits it into subprompts

[1554] The server analyzes the received prompt and divides it into multiple sub-prompts based on different tasks, such as "analyzing camera image samples," "presenting camera specifications," and "summarizing user reviews." The input is the prompt text, and the output is a set of sub-prompts corresponding to each task.

[1555] Step 4:

[1556] Send sub-prompts to each AI model

[1557] The server sends the generated sub-prompts to each AI model. For example, it sends a sub-prompt such as "Generate an image sample of the camera" to the image generation model, and a sub-prompt such as "Describe the technical specifications of the camera" to the sentence generation model. The input is the sub-prompt, and the output is the task sent to each model.

[1558] Step 5:

[1559] The server receives the results from each AI model.

[1560] The server receives the output from each AI model. The image generation model generates sample images of the camera, the sentence generation model generates technical specifications for the camera, and the review analysis model generates summaries of user reviews. The input is the generated results based on the sub-prompts, and the output is the resulting data from each model.

[1561] Step 6:

[1562] The server aggregates the results

[1563] The server then combines the results it receives to generate a single, coherent answer. For example, it might combine the generated camera image, technical specifications, and review summary into a single response. The input is the result data from each model, and the output is the combined answer.

[1564] Step 7:

[1565] The server sends the consolidated results back to the user

[1566] The server sends the integrated results to the user's device. The user's device displays the results to the user, allowing the user to obtain comprehensive information about the product. For example, the user's device might display a message such as, "This smartphone's camera performance is excellent, and it takes clear photos even at night. Many users also rate its ease of use highly." The input is the integrated result data, and the output is the content displayed on the user's device.

[1567] 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.

[1568] This invention provides a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and influences the analysis results.

[1569] Program processing

[1570] server

[1571] 1. Accepting the prompt

[1572] The server receives the prompt sent from the user terminal.

[1573] Example: A user enters the prompt "Generate a poem to comfort me when I'm sad."

[1574] 2. Emotion Analysis

[1575] The server is equipped with an emotion engine that analyzes text data from the user's input prompts to recognize the user's emotions.

[1576] Example: The emotion "sad" is recognized from the input prompt.

[1577] 3. Parsing the prompt

[1578] The server parses the received prompt, understands its content and splits it into appropriate tasks.

[1579] Example: Identify the task "Generate a poem."

[1580] 4. Generating a subprompt

[1581] The server generates sub-prompts for each task based on the analysis results, taking into account the analysis results of the emotion engine.

[1582] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[1583] 5. Sending a sub-prompt

[1584] The server sends the generated sub-prompts to the appropriate artificial intelligence model.

[1585] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[1586] 6. Acceptance of Results

[1587] The server receives the output from the artificial intelligence model.

[1588] Example: Accepting a poem generated from a poetry generation model.

[1589] 7. Synthesis of results

[1590] The server aggregates the results it receives.

[1591] Example: Compiling the generated poems into a single data package.

[1592] 8. Return of results

[1593] The server transmits the integrated results to the user terminal.

[1594] User terminal

[1595] 1. Enter the prompt

[1596] The user enters a prompt into the interface.

[1597] Example: User types "Generate a comforting poem for when I'm sad."

[1598] 2. Sending a prompt

[1599] The user terminal sends the entered prompt to the server.

[1600] 3. Displaying the results

[1601] The results returned by the server are displayed to the user.

[1602] Example: A poem to comfort you when you're feeling sad is displayed.

[1603] Specific examples

[1604] Example 1: Emotion-based poetry generation process

[1605] The user enters the prompt "Generate a comforting poem for sad feelings."

[1606] The user terminal sends a prompt to the server.

[1607] The server receives the prompt and recognizes the emotion "sad" using an emotion engine.

[1608] The server parses the prompt and identifies the task "Generate Poetry."

[1609] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[1610] The server receives the poem generated from the poetry generation model.

[1611] The server aggregates the generated poems and transmits them to the user's terminal.

[1612] The user terminal displays the received poem to the user.

[1613] The user checks the generated poem and experiences changes in emotion.

[1614] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling users to easily complete complex tasks.

[1615] The processing flow will be explained below.

[1616] Step 1:

[1617] The user enters a prompt into the interface.

[1618] Example: Type "Generate a poem to comfort me when I'm sad."

[1619] Step 2:

[1620] The terminal sends the prompt entered by the user to the server.

[1621] For example: A prompt is sent to the server.

[1622] Step 3:

[1623] The server receives the prompt.

[1624] Example: Accept the prompt "Generate a poem to comfort someone who is sad."

[1625] Step 4:

[1626] The server uses an emotion engine to analyze the user's emotion from the text data of the prompt.

[1627] Example: Analyze keywords such as "sad" in the prompt and recognize that the user's emotion is "sadness."

[1628] Step 5:

[1629] The server parses the prompt, understands its content and splits it into the appropriate tasks.

[1630] Example: Parse the prompt and identify the task "Generate a poem."

[1631] Step 6:

[1632] The server generates sub-prompts corresponding to each task, taking into account the emotional information obtained from the emotion engine.

[1633] For example, generate a specific sub-prompt such as, "Generate a poem that comforts sadness."

[1634] Step 7:

[1635] The server sends the generated sub-prompts to the selected artificial intelligence model.

[1636] Example: Sending a poetry generation model the sub-prompt "Generate a poem that comforts sadness."

[1637] Step 8:

[1638] The server receives the output from the artificial intelligence model.

[1639] Example: Accepting a poem generated from a poetry generation model.

[1640] Step 9:

[1641] The server consolidates the results it receives.

[1642] Example: Compiling the generated poems into a single data package.

[1643] Step 10:

[1644] The server transmits the integrated results to the user terminal.

[1645] Example: Sending a consolidated poetry data package to a user device.

[1646] Step 11:

[1647] The terminal displays the results received from the server to the user.

[1648] Example: "Poems to comfort sad feelings" is displayed on the user's interface.

[1649] Step 12:

[1650] The user checks the displayed results.

[1651] Example: Read the generated poem and experience the changes in emotions.

[1652] This system analyzes the user's emotions from the prompt text data and uses the results to generate more personalized results so that the user can get appropriate content that takes their emotions into consideration.

[1653] Example 2

[1654] 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."

[1655] Currently, many artificial intelligence systems exist, but they generally use a single model to respond to prompts entered by the user. However, it is difficult for a single model to generate outputs that adequately reflect the user's emotions. This creates a need for systems that can provide a truly personalized experience for users. Furthermore, they lack the ability to automatically identify multiple tasks and generate appropriate sub-prompts. A system that can solve these problems is needed.

[1656] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1657] In this invention, the server includes means including an emotion engine that recognizes a user's emotions and influences the analysis result, means for receiving a prompt from the user, means for analyzing the received prompt and dividing it into multiple sub-prompts based on the prompt, means for sending the sub-prompts to respective AI models, means for receiving results from the AI ​​models, means for integrating the received results, and means for returning the integrated results to the user. This makes it possible to provide a truly personalized output by taking the user's emotions into consideration and appropriately processing multiple tasks.

[1658] A "prompt" is text that a user enters into a system as an instruction or question.

[1659] A "subprompt" is a more specific instruction for performing a particular task that is generated from a parsed prompt.

[1660] An "artificial intelligence model" is a machine learning algorithm that has been trained to perform a specific task automatically.

[1661] An "emotion engine" is a part of a system that analyzes user input text and identifies the emotions contained in that text.

[1662] A "task" is an individual task or operation that a system must perform.

[1663] "Results synthesis" is the process of combining the individual outputs from multiple artificial intelligence models into a single, coherent result.

[1664] This invention is a system that analyzes a prompt from a user, generates multiple sub-prompts based on the prompt, sends them to multiple artificial intelligence models, and then integrates the results obtained from each model and returns them to the user. Furthermore, this invention incorporates an emotion engine that recognizes the user's emotions and affects the analysis results.

[1665] Hardware and software configuration

[1666] The system of the present invention uses the following hardware and software.

[1667] Server: A server computer with a high-performance processor and a large amount of memory. For example, you can use a cloud-based server (AWS, Google Cloud, Microsoft Azure, etc.).

[1668] User terminal: A device used by a user to access a system. Examples include personal computers, smartphones, and tablets.

[1669] Emotion engine: Software that uses natural language processing techniques to analyze user emotions, for example, by using open source NLP libraries (such as Hugging Face's Transformers).

[1670] Artificial intelligence models: Generative AI models use machine learning models (e.g., GPT-3, BERT) that are trained to perform specific tasks.

[1671] System Operation

[1672] 1. User prompt input

[1673] The user enters a prompt into the interface, for example, text such as "Generate a poem to comfort me when I'm feeling sad."

[1674] 2. Sending a prompt

[1675] The user terminal sends the input prompt to the server using an HTTP POST request.

[1676] 3. Accepting and parsing prompts

[1677] The server receives a prompt sent from a user terminal, identifies the emotion of the prompt using an emotion engine, and then analyzes the prompt and divides it into multiple tasks, for example, identifying a task called "poetry creation."

[1678] 4. Generating a subprompt

[1679] Based on the analysis results, the server generates sub-prompts for each task, taking into account the results of the emotion engine. Specifically, it generates specific sub-prompts such as "Please generate a poem that will comfort sad feelings."

[1680] 5. Submitting a subprompt and accepting the results

[1681] The server sends the generated sub-prompts to an artificial intelligence model, such as a poetry generation model, and receives output from the model.

[1682] 6. Consolidating and returning results

[1683] The server integrates the results obtained from multiple tasks and sends them back to the user's device, which receives them and displays them on the screen.

[1684] Specific examples

[1685] Example 1: Emotion-based poetry generation process

[1686] The user enters the prompt "Generate a comforting poem for sad feelings."

[1687] The user terminal sends a prompt to the server.

[1688] The server receives the prompt and identifies the emotion "sad" through an emotion engine.

[1689] The server parses the prompt and identifies the task "Generate Poetry."

[1690] The server sends a sub-prompt to the poetry generation model saying, "Please generate a poem that comforts sad feelings."

[1691] The server receives the poem generated from the poetry generation model.

[1692] The server aggregates the generated poems and transmits them to the user's terminal.

[1693] The user terminal displays the received poem to the user.

[1694] This invention allows users to easily obtain emotion-aware prompt analysis and generated results, providing a more personalized experience, improving user satisfaction and enabling complex tasks to be completed easily.

[1695] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1696] Step 1:

[1697] The user terminal sends a prompt to the server. Specifically, when the user enters "Create a poem to comfort me when I'm sad" into the interface and presses the submit button, the system sends an HTTP POST request to the server. The input is a text prompt, and the output is a request sent to the server.

[1698] Step 2:

[1699] The server receives a prompt sent from the user terminal. The server receives an HTTP POST request and parses the text prompt included in the request body. At this time, the server extracts the prompt text and passes it on to subsequent processing. The input is the prompt text sent by the user, and the output is the prompt data to be parsed.

[1700] Step 3:

[1701] The emotion engine on the server analyzes the prompt and recognizes the user's emotion. It uses a natural language processing (NLP) library to tokenize the prompt and then runs it through an emotion classifier to extract an emotion tag. Specifically, it recognizes the emotion "sad." In this process, the input is the prompt text and the output is an emotion tag (e.g., "sad").

[1702] Step 4:

[1703] The server parses the prompt and splits it into tasks. NLP techniques are used to parse the text and identify tasks (e.g., "Generate Poetry"). The input is the prompt text with emotion tags, and the output is the identified tasks.

[1704] Step 5:

[1705] Based on the analysis results, the server generates sub-prompts corresponding to each task, taking into account the results of the emotion engine. For example, it generates a sub-prompt such as "Please create a poem that will comfort me when I'm feeling sad." In this process, the input is the identified task and emotion tag, and the output is the sub-prompt.

[1706] Step 6:

[1707] The server sends the generated sub-prompt to the appropriate AI model. It sends the sub-prompt via an HTTP request to the poetry generation model's API endpoint and waits for a response. The input is the sub-prompt, and the output is a request sent to the poetry generation model.

[1708] Step 7:

[1709] The server receives the output from the AI ​​model, analyzes the HTTP response from the poetry generation model, and obtains the generated poem. The input is the response data from the poetry generation model, and the output is the text of the generated poem.

[1710] Step 8:

[1711] The server aggregates the results from multiple tasks, compiles the generated poems into a single data package, and formats it in a format that can be sent back to the user. The input is the multiple output results, and the output is the aggregated result data.

[1712] Step 9:

[1713] The server returns the integrated results to the user terminal. An HTTP response is generated and the integrated result data is sent to the user terminal. The input is the integrated result data, and the output is the response sent to the user terminal.

[1714] Step 10:

[1715] The user terminal displays the results returned from the server to the user. The HTTP response is received and its contents are displayed on the screen. As a concrete example, the generated poem is displayed on the user's interface. The input is the response data from the server, and the output is the result displayed to the user.

[1716] (Application example 2)

[1717] 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."

[1718] Existing food delivery systems can respond quickly to user orders, but they cannot provide personalized suggestions that take into account the user's emotions and mood. This limits their ability to improve user satisfaction. In particular, there is a need for systems that provide meal suggestions that take into account emotions such as stress and fatigue.

[1719] The specification process 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 receiving a prompt from the user, means for analyzing the received prompt and dividing it into a plurality of sub-prompts based on the prompt, means for sending the sub-prompts to each AI model, means for analyzing the user's emotions, and means for generating sub-prompts based on the emotion analysis results and sending them to the AI ​​model. This enables personalized meal suggestions tailored to the user's emotions.

[1720] A "prompt" refers to the text of a request or question entered by a user into the system.

[1721] A "sub-prompt" is a prompt that you receive broken down into multiple smaller tasks.

[1722] "Artificial intelligence model" refers to a machine learning algorithm for performing a specific task, including generative and propositional models.

[1723] "Sentiment analysis" refers to the process of identifying and analyzing emotions from a user's input prompts.

[1724] A "generative model" refers to an AI algorithm that automatically generates new text or images based on input requests.

[1725] "Proposal model" refers to an AI algorithm that provides appropriate suggestions and options based on the analysis results.

[1726] "Synthesis" refers to the process of combining the results of multiple sub-prompts into one.

[1727] The present invention realizes a system that analyzes prompts from a user, generates multiple sub-prompts based on the emotional analysis results, and sends them to multiple artificial intelligence models to provide personalized results to the user. To implement this invention, a server, a user terminal, and various software are required.

[1728] Hardware and software used

[1729] Hardware

[1730] server

[1731] User device (smartphone)

[1732] software

[1733] Sentiment analysis engines (e.g., Python-based NLP libraries such as spaCy and NLTK)

[1734] Artificial intelligence models (e.g., GPT-4 or Transformers for running generative and propositional models)

[1735] A database (for temporarily storing results)

[1736] System configuration

[1737] 1. Prompt input from the user terminal

[1738] A user uses a food delivery application to enter a prompt, which can be entered as text or speech. For example, "I'm feeling stressed today, can you recommend a comforting meal?"

[1739] 2. Accepting the prompt

[1740] The user terminal sends the input prompt to the server, which receives it and inputs it into the emotion analysis engine.

[1741] 3. Emotion Analysis

[1742] The server uses an emotion analysis engine (such as spaCy or NLTK) to analyze the text in the prompt and identify the user's emotion. For example, it analyzes the emotion "stress" from the prompt "I'm stressed."

[1743] 4. Parsing the prompt

[1744] The server splits the prompt into multiple sub-prompts, taking into account the results of the sentiment analysis, using a text analysis library or similar.

[1745] 5. Generating Subprompts

[1746] Based on the results of the sentiment analysis, the server generates an appropriate sub-prompt, such as "Please suggest some dishes that will help relieve stress."

[1747] 6. Sending sub-prompts

[1748] The server then sends the generated sub-prompts to an appropriate artificial intelligence model, such as a generative model that suggests dishes.

[1749] 7. Acceptance and Integration of Results

[1750] The server receives and integrates the results from the artificial intelligence models, a process that combines multiple results into a single data package.

[1751] 8. Return of results

[1752] The server sends the integrated results to the user's device, which then displays, for example, a list of recommended dishes for relieving stress.

[1753] Specific examples

[1754] User: Enters the prompt "I'm feeling stressed today, what's a comforting dish?"

[1755] User terminal: Sends the prompt to the server.

[1756] Server: Receives the prompt, performs emotion analysis, and identifies the emotion "stress."

[1757] Server: Based on the analysis results, generate a sub-prompt such as "Please suggest a dish that will help relieve stress" and send it to the dish suggestion model.

[1758] Server: Receives the resulting proposals, integrates them, and sends them back to the user device.

[1759] User terminal: The suggested dishes are displayed to the user.

[1760] This process allows users to receive personalized meal suggestions that match their emotions, realizing a food delivery service that meets a variety of needs.

[1761] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1762] Step 1:

[1763] A user opens a food delivery application on a user device such as a smartphone and inputs a prompt such as, "I'm feeling stressed today, so please recommend some soothing dishes." This user input becomes the initial prompt data.

[1764] Step 2:

[1765] The user terminal sends the input prompt to the server. The prompt data sent to the server is "I'm feeling stressed today, so please tell me some soothing dishes."

[1766] Step 3:

[1767] The server receives the prompt and uses a sentiment analysis engine (e.g., spaCy or NLTK) to analyze the text in the prompt and identify the user's sentiment. In this case, the resulting sentiment is "stress."

[1768] Step 4:

[1769] The server analyzes the prompt based on the results of the sentiment analysis and divides it into appropriate sub-prompts. In this case, the sub-prompt generated is "Please suggest a dish that will help relieve stress."

[1770] Step 5:

[1771] The server sends the generated sub-prompts to a dish suggestion model (e.g., GPT-4 or Transformers), which receives the sub-prompts as input and then outputs dish suggestions based on them.

[1772] Step 6:

[1773] The server receives the results of the dish recommendation model, such as a list of dishes like "herbal tea," "banana smoothie," and "yogurt parfait."

[1774] Step 7:

[1775] The server aggregates the multiple recommendations into a single data package, in this case, a list of recommended dishes.

[1776] Step 8:

[1777] The server transmits the integrated results to the user terminal, and the transmitted data includes the integrated recipe list.

[1778] Step 9:

[1779] The user device displays the received list of dishes to the user, who can then check suggested dishes such as "herbal tea," "banana smoothie," and "yogurt parfait" on the application.

[1780] The above steps realize a system that allows users to get personalized recipe suggestions based on their emotions.

[1781] 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.

[1782] 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.

[1783] 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.

[1784] 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.

[1785] 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.

[1786] 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.

[1787] 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).

[1788] 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.

[1789] 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."

[1790] 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.

[1791] 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).

[1792] 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.

[1793] 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.

[1794] 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.

[1795] 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.

[1796] 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.

[1797] 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.

[1798] 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.

[1799] 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.

[1800] 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.

[1801] 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.

[1802] The following is further disclosed regarding the above embodiment.

[1803] (Claim 1)

[1804] means for receiving a prompt from a user;

[1805] means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt;

[1806] means for transmitting the sub-prompts to respective artificial intelligence models;

[1807] means for receiving results from said artificial intelligence model;

[1808] a means for integrating the results received; and

[1809] means for returning the consolidated results to the user;

[1810] A system including:

[1811] (Claim 2)

[1812] 2. The system of claim 1, wherein the prompt parsing means includes means for identifying different tasks within the prompt.

[1813] (Claim 3)

[1814] The system of claim 1 , wherein the artificial intelligence models include an image generation model and a sentence generation model.

[1815]

[1816] "Example 1"

[1817] (Claim 1)

[1818] means for receiving a prompt sentence from a user terminal;

[1819] means for analyzing a received prompt sentence and dividing the prompt sentence into a plurality of sub-prompt sentences based on the prompt sentence;

[1820] means for transmitting the sub-prompt sentence to each generative AI model;

[1821] means for receiving results from the generative AI model;

[1822] a means for integrating the results received; and

[1823] means for returning the integrated results to the user terminal;

[1824] A system including:

[1825] (Claim 2)

[1826] 2. The system of claim 1, wherein the prompt sentence parsing means includes means for identifying different tasks within the prompt sentence.

[1827] (Claim 3)

[1828] 10. The system of claim 1, wherein the generative AI models include an image generation model and a text generation model.

[1829] "Application Example 1"

[1830] (Claim 1)

[1831] means for receiving a prompt from a user;

[1832] means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt;

[1833] means for transmitting the sub-prompts to respective artificial intelligence models;

[1834] means for receiving results from said artificial intelligence model;

[1835] a means for integrating the results received; and

[1836] means for returning the consolidated results to the user;

[1837] A means for analyzing product-related questions in the virtual store and providing relevant information;

[1838] A system including:

[1839] (Claim 2)

[1840] 2. The system of claim 1, wherein the prompt parsing means includes means for identifying different tasks within the prompt.

[1841] (Claim 3)

[1842] 2. The system of claim 1, wherein the artificial intelligence models include an image generation model, a sentence generation model, and a review analysis model.

[1843] "Example 2: Combining Emotion Engines"

[1844] (Claim 1)

[1845] means for receiving a prompt from a user;

[1846] means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt;

[1847] means for transmitting the sub-prompts to respective artificial intelligence models;

[1848] means for receiving results from said artificial intelligence model;

[1849] a means for integrating the results received; and

[1850] means for returning the consolidated results to the user;

[1851] A system that includes an emotion engine that recognizes user emotions and influences analysis results.

[1852] (Claim 2)

[1853] 2. The system of claim 1, wherein the prompt parsing means includes means for identifying different tasks within the prompt.

[1854] (Claim 3)

[1855] The system of claim 1 , wherein the artificial intelligence models include an image generation model and a sentence generation model.

[1856] "Application example 2 when combining emotion engines"

[1857] (Claim 1)

[1858] means for receiving a prompt from a user;

[1859] means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt;

[1860] means for transmitting the sub-prompts to respective artificial intelligence models;

[1861] means for receiving results from said artificial intelligence model;

[1862] a means for integrating the results received; and

[1863] means for returning the consolidated results to the user;

[1864] means for analyzing user emotions;

[1865] means for generating sub-prompts based on the sentiment analysis results and sending the sub-prompts to the artificial intelligence model;

[1866] a means by which the integrated results are adjusted to the user's emotions;

[1867] A system including:

[1868] (Claim 2)

[1869] 2. The system of claim 1, wherein the prompt parsing means includes means for identifying different tasks within the prompt.

[1870] (Claim 3)

[1871] The system of claim 1 , wherein the artificial intelligence models include generative models and propositional models. [Explanation of symbols]

[1872] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for receiving a prompt from a user; means for parsing a received prompt and dividing the prompt into a plurality of sub-prompts based on the prompt; means for transmitting the sub-prompts to respective artificial intelligence models; means for receiving results from said artificial intelligence model; a means of integrating the results received; and means for returning the consolidated results to the user; A system including:

2. 2. The system of claim 1, wherein the prompt parsing means includes means for identifying different tasks within the prompt.

3. The system of claim 1 , wherein the artificial intelligence models include an image generation model and a sentence generation model.

Images