System

Abstract

A system is provided.SOLUTION: A system comprising: means for receiving a search query from a user; means for obtaining a response from a natural language generation model using the search query; and means for returning the obtained response to the user.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

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

[0004] In business, it is important to be able to search past sales announcements quickly and efficiently and to respond to users immediately regarding their content. However, many companies lack systems for efficiently managing and searching past sales announcements, making it difficult to quickly obtain the necessary information. They also lack the ability to respond immediately to inquiries about detailed content. There is a need for a system that addresses these issues and enables more efficient and prompt search of sales announcements and response to inquiries. [Means for solving the problem]

[0005] The present invention provides a system including a means for receiving a search query from a user, a means for obtaining a response from a natural language generation model using the search query, and a means for returning the obtained response to the user. The system also includes a means for receiving a detailed request from the user including a specific ID, a means for obtaining corresponding content from a database based on the ID, a means for obtaining a response from the natural language generation model using the obtained content, and a means for returning the obtained response to the user. Furthermore, the system further includes a means for dynamically generating prompts to be sent to the natural language generation model when obtaining a response from the natural language generation model using the search query and detailed request, thereby enabling more accurate and efficient responses to user requests. This enables quick and accurate responses to business searches and detailed inquiries.

[0006] "User" means any person or entity that utilizes the System to enter search queries or detailed requests and receive results.

[0007] "Search query" refers to a string of characters or a set of keywords that a user submits to a system to retrieve specific information.

[0008] A "natural language generation model" refers to an artificial intelligence algorithm that generates responses that simulate human language based on given input data.

[0009] A "response" refers to information or an answer generated by a natural language generation model and provided to a user.

[0010] "Detail Request" refers to a request sent by a user to the system to obtain more detailed information about data containing a specific ID.

[0011] "ID" refers to a number or code that uniquely identifies a particular item within a database.

[0012] A "database" refers to a collection of data in which specific information is systematically stored and designed to be easily searchable and accessible.

[0013] A "prompt" refers to input data or instructions given to a natural language generation model.

[0014] "Dynamic generation means" refers to methods or techniques for constructing prompts in real time based on a user's search query or detailed request and passing them to a natural language generation model.

[0015] "Search" refers to the act of finding or obtaining specific information.

[0016] "Detailed inquiry" refers to the act of requesting more detailed information about a specific item. [Brief explanation of the drawings]

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

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

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

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

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

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

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

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

[0025] [First embodiment]

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

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

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

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

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

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

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

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

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

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

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

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

[0038] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and responding to users immediately regarding their content. The system includes a terminal through which a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[0039] Search function embodiment

[0040] User enters and submits a search query

[0041] The user inputs a search query using the device interface. For example, they input the keyword "Business Announcement Contents for 2023" and press the search button. This operation sends the search query from the device to the server.

[0042] Server receives and processes search queries

[0043] The server receives a search query sent from a user's device, analyzes the received query, and generates and sends a prompt to a natural language generation model (e.g., OpenAI's API) based on the query's content. The natural language generation model generates a response related to the query and sends it back to the server.

[0044] Response generation and forwarding

[0045] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's device is based on the user's query, providing information quickly and accurately.

[0046] Detailed inquiry function embodiment

[0047] User enters and submits detailed request

[0048] To find out more about a specific business announcement, the user inputs its ID and sends a details request. For example, if the user wants to know the details of the business announcement with ID 1, the user inputs the ID and presses the details button. This operation sends a details request from the terminal to the server.

[0049] Server receives and processes detailed requests

[0050] The server receives a detailed request sent from the user's device. It analyzes the ID included in the request and retrieves the relevant sales notice information from the database. It generates and sends a prompt to the natural language generation model based on the relevant information. The natural language generation model generates a detailed response and sends it back to the server.

[0051] Response generation and forwarding

[0052] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's terminal is based on the detailed request, providing information quickly and accurately.

[0053] Specific examples

[0054] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[0055] {

[0056] "query": "2023 Business Announcement"

[0057] }

[0058] The server receives this query and generates and sends prompts to the natural language generation model, which then generates a response and returns it to the user.

[0059] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[0060] GET / detail / 1

[0061] The server receives this request, retrieves the relevant information from a database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user.

[0062] This system allows users to quickly and efficiently search for previously distributed business announcements and get immediate responses to their detailed information, significantly improving the efficiency of managing and using business announcements.

[0063] The processing flow will be explained below.

[0064] Search function processing steps

[0065] Step 1:

[0066] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[0067] Step 2:

[0068] The user presses the search button to send a search request including the query from the terminal to the server.

[0069] Step 3:

[0070] The server receives the search request, parses the query string, and generates a prompt to send to the OpenAI API.

[0071] Step 4:

[0072] The server sends the generated prompt to the OpenAI API and requests a response.

[0073] Step 5:

[0074] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[0075] Step 6:

[0076] The server compiles the response received from the OpenAI API and sends it back to the user.

[0077] Step 7:

[0078] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[0079] Processing steps for detailed inquiry function

[0080] Step 1:

[0081] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[0082] Step 2:

[0083] The user presses the details button, and a details request including the ID is sent from the terminal to the server.

[0084] Step 3:

[0085] The server receives the detailed request, extracts the ID included in the request, and searches the database based on that ID to retrieve the relevant business announcement.

[0086] Step 4:

[0087] The server uses the acquired business notice to generate a prompt to send to the OpenAI API.

[0088] Step 5:

[0089] The server sends the generated prompt to the OpenAI API and requests a response.

[0090] Step 6:

[0091] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[0092] Step 7:

[0093] The server compiles the response received from the OpenAI API and sends it back to the user.

[0094] Step 8:

[0095] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[0096] Example 1

[0097] 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."

[0098] Conventional information retrieval systems have difficulty generating fast and accurate responses to specific keywords or detailed requests. Furthermore, more flexible and efficient information processing methods are needed to respond to diverse user requests. In particular, there is a need for systems that can accurately search past information and respond to users immediately, but there are not enough systems available that can meet these demands.

[0099] 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.

[0100] In this invention, the server includes means for receiving input from a user, means for dynamically generating prompts for a generative model based on the input, and means for obtaining responses from the generative model using the dynamically generated prompts, thereby enabling the server to respond to a variety of user requests and provide responses quickly and accurately.

[0101] "Means for receiving input from a user" refers to the functionality or interface that allows a user to input information and for the system to receive it.

[0102] "Means for dynamically generating prompts for a generative model" refers to a function that automatically generates questions (prompts) that enable the generative model to return an appropriate response based on input information.

[0103] The "means for obtaining a response from a generative model" refers to a function for receiving a response returned from a generative model based on a generated prompt.

[0104] "Means for returning to the user" refers to a function within the system that sends information to the user's terminal in order to display the obtained response to the user.

[0105] The term "means for receiving a request including a specific identifier" refers to a function whereby a user provides an identifier for specifying specific information and the system receives the identifier.

[0106] "Means for obtaining corresponding content from information storage means" refers to the function of searching and retrieving information based on a specified identifier from a database or storage within the system.

[0107] The "means for dynamically generating a prompt based on specific conditions" refers to a function for dynamically generating an appropriate prompt in response to a user's input or request.

[0108] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and instantly responding to the user about their content. The system includes a terminal through which the user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[0109] Search function embodiment

[0110] User enters a search query

[0111] The user uses their own device to input the keywords they want to search for into the interface. For example, they can input "2023 business announcements" and press the search button to confirm the query input.

[0112] Sending queries by terminal

[0113] After pressing the search button, the terminal sends the entered query to the server. The data sent is as follows:

[0114] {

[0115] "query": "2023 Business Announcement"

[0116] }

[0117] Server receives and processes search queries

[0118] The server receives the search query sent from the device, analyzes the query, and dynamically generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[0119] "Please tell me what the business announcement for 2023 will be."

[0120] Prompting a Natural Language Generation Model

[0121] The server sends the generated prompts to a natural language generation model (e.g., a popular generative AI API), which generates a response based on the prompts and sends the result back to the server.

[0122] Receiving and formatting the response by the server

[0123] The server receives the response sent back by the generative model and formats it appropriately, which may include converting it to HTML or JSON format.

[0124] Terminal display of responses

[0125] The terminal receives the formatted response from the server and displays it to the user, allowing the user to quickly see the results of their query.

[0126] Detailed inquiry function embodiment

[0127] User enters detailed request

[0128] If the user wants to know the details of a specific business announcement, he or she inputs its ID and presses the Details button. For example, the user inputs "Details of the business announcement with ID 1."

[0129] Sending a details request by terminal

[0130] After pressing the Details button, the device will send the input request to the server, which will have the following format:

[0131] GET / detail / 1

[0132] Server receives and processes detailed requests

[0133] The server receives the detailed request sent from the device, analyzes the ID included in the request, and searches the database. Based on the information obtained from the database, it generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[0134] "Please tell me the details of the business announcement with ID 1."

[0135] Prompting a Natural Language Generation Model

[0136] The server sends the generated prompt to a natural language generation model to generate a detailed response, which is then sent back to the server.

[0137] Server response formatting and forwarding

[0138] The server receives the response from the generative model, formats it in an appropriate format, and sends it back to the device. Examples of returned data are as follows:

[0139] {

[0140] "response": "The details of the sales announcement with ID 1 are as follows..."

[0141] }

[0142] Terminal display of responses

[0143] The terminal displays the response received from the server on the screen, allowing the user to quickly check the correct information for the detailed request.

[0144] The present invention allows users to quickly and efficiently search past business announcements and get immediate responses to detailed information, greatly improving the efficiency of business announcement management and utilization.

[0145] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0146] Search function embodiment

[0147] Step 1:

[0148] User enters a search query

[0149] The user inputs the keywords they want to search for using the interface of their device. For example, they input "Business Notices for 2023" and press the search button. This operation confirms the query they entered.

[0150] Input: "Business Notice for 2023"

[0151] Output: The query is typed into the terminal

[0152] Step 2:

[0153] Sending queries by terminal

[0154] When the search button is pressed, the device sends the entered query to the server. Specifically, the query is packaged in JSON format and sent to the server as an HTTP request.

[0155] Input: "Business Notice for 2023"

[0156] Output: The query is sent to the server as an HTTP request.

[0157] Step 3:

[0158] Server receives search query

[0159] The server receives the query sent from the device, parses the received data in JSON format, and extracts the search keywords.

[0160] Input: Query from terminal

[0161] Output: Parsed search keywords

[0162] Step 4:

[0163] The server parses the query and generates prompts

[0164] The server analyzes the received query and dynamically generates a prompt based on its content, which becomes the question sent to the generative AI model.

[0165] How it works: Builds a prompt from search keywords.

[0166] Input: Parsed search keywords

[0167] Output: Generated prompt: "What is the content of your 2023 sales announcement?"

[0168] Step 5:

[0169] Prompting a Natural Language Generation Model

[0170] The server sends the generated prompt to a generative AI model (e.g., a general generative AI API).

[0171] What it does: Sends the prompt to the generative AI model as an HTTP request.

[0172] Input: Generated prompt

[0173] Output: A request to the NLG model

[0174] Step 6:

[0175] Receiving responses from generative AI models

[0176] The server receives a response from the generative AI model, which is a natural language answer generated based on the prompt.

[0177] Behavior: Receives the response as an HTTP response and parses it.

[0178] Input: The response returned by the generative AI model

[0179] Output: Parsed response content

[0180] Step 7:

[0181] Server response formatting

[0182] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display to the user.

[0183] Behavior: Formats the response content.

[0184] Input: Response from the generative AI model

[0185] Output: Formatted response data

[0186] Step 8:

[0187] Response from the server to the device

[0188] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[0189] Behavior: Sends response data as an HTTP response.

[0190] Input: Formatted response data

[0191] Output: Send response data to the terminal

[0192] Step 9:

[0193] Terminal display of responses

[0194] The terminal displays the response received from the server to the user, allowing the user to check the search results.

[0195] Behavior: Displays the response data in the interface.

[0196] Input: Response data from the server

[0197] Output: What is displayed to the user

[0198] Detailed inquiry function embodiment

[0199] Step 1:

[0200] User enters detailed request

[0201] If the user wants to know the details of a specific business announcement, he or she enters its ID and presses the Details button. For example, the user might enter "Details of the business announcement with ID 1." This operation confirms the request for details.

[0202] Input: "Details of business announcement with ID 1"

[0203] Output: Detailed request typed into terminal

[0204] Step 2:

[0205] Sending a details request by terminal

[0206] When the details button is pressed, the terminal transmits the input request to the server. Specifically, the request is transmitted to the server as an HTTP GET request.

[0207] Input: "Details of business announcement with ID 1"

[0208] Output: The detailed request is sent to the server as an HTTP GET request.

[0209] Step 3:

[0210] Server receives detailed request

[0211] The server receives a detailed request containing the ID sent by the device, which is then parsed to extract the ID.

[0212] Input: Detailed request from terminal

[0213] Output: Parsed ID

[0214] Step 4:

[0215] Server parses the request and performs a database lookup

[0216] The server searches the database based on the ID included in the detailed request and obtains the relevant business notice information.

[0217] What it does: Searches the database based on the ID and retrieves the relevant information.

[0218] Input: Parsed ID

[0219] Output: Business announcement information retrieved from the database

[0220] Step 5:

[0221] Server-generated prompts

[0222] Based on the information obtained, the server generates a prompt that is sent to the generative AI model, asking for more information.

[0223] How it works: Construct a prompt from the information obtained.

[0224] Input: Business announcement information retrieved from the database

[0225] Output: Generated prompt: "Please tell me the details of the sales announcement with ID 1."

[0226] Step 6:

[0227] Prompting a Natural Language Generation Model

[0228] The server sends the generated prompts to the generative AI model.

[0229] What it does: Sends the prompt to the generative AI model as an HTTP request.

[0230] Input: Generated prompt

[0231] Output: A request to the NLG model

[0232] Step 7:

[0233] Receiving responses from generative AI models

[0234] The server receives a response from the generative AI model, which is a detailed answer generated based on the prompt.

[0235] Behavior: Receives the response as an HTTP response and parses it.

[0236] Input: The response returned by the generative AI model

[0237] Output: Parsed response content

[0238] Step 8:

[0239] Server response formatting

[0240] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display on the device.

[0241] Behavior: Formats the response content.

[0242] Input: Response from the generative AI model

[0243] Output: Formatted response data

[0244] Step 9:

[0245] Response from the server to the device

[0246] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[0247] Behavior: Sends response data as an HTTP response.

[0248] Input: Formatted response data

[0249] Output: Send response data to the terminal

[0250] Step 10:

[0251] Terminal display of responses

[0252] The terminal displays the response received from the server to the user, allowing the user to quickly check the information for the detailed request.

[0253] Behavior: Displays the response data in the interface.

[0254] Input: Response data from the server

[0255] Output: What is displayed to the user

[0256] (Application example 1)

[0257] 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."

[0258] On current online shopping sites, it can take a significant amount of time for users to quickly search and obtain information about past campaigns or detailed product information. Furthermore, there is a lack of systems that accurately provide the information users are looking for, resulting in a poor user experience. The present invention aims to solve these problems and provide a system that allows users to efficiently and quickly obtain the information they need.

[0259] 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.

[0260] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, means for providing a database for efficiently managing information searched in the past, and means for inputting a search query through an interface on a smartphone, thereby enabling users to quickly and accurately search and obtain information on past campaigns and detailed product information.

[0261] A "search query" is a keyword or phrase that a user enters to locate specific information.

[0262] A "natural language generation model" is an artificial intelligence technique that generates relevant sentences and responses based on input text.

[0263] A "response" is the information or answer provided by a system in response to a user's search query.

[0264] A "database" is a collection of data that stores previously searched information and detailed information in an organized manner so that it can be efficiently searched and retrieved later.

[0265] A "smartphone" is a type of mobile phone that has internet connectivity and a variety of applications, allowing users to perform various searches and obtain information.

[0266] An "interface" is a screen or operating means that a user uses to interact with a system.

[0267] A "detail request" is a request including a specific ID that a user enters when they want to know more about a specific piece of information.

[0268] The present invention is a system for quickly retrieving information and detailed information that a user has previously searched for. This system is operated via an interface installed on the user's smartphone. The main components of the system include the following:

[0269] System Configuration

[0270] Hardware elements

[0271] 1. Smartphone: A device where users enter search queries and detailed requests. It communicates with the server through an interface.

[0272] 2. Server: Receives queries and requests, retrieves information from the database, and generates and sends prompts to the natural language generation model.

[0273] Software Elements

[0274] 1. Interface: A smartphone application built using React Native that provides an interface for users to enter search queries and detailed requests.

[0275] 2. Server application: Built using Node.js and the Express framework, it handles user queries and requests, generates prompts, and sends them to the natural language generation model.

[0276] 3. Database Management System: MongoDB is used to store and manage past search queries and detailed information.

[0277] 4. Natural language generation models: Generative AI models provided as APIs (e.g., OpenAI's API) are used to generate relevant responses based on input queries.

[0278] Processing flow

[0279] Query Processing

[0280] A user uses the smartphone interface to enter a search query, for example, "Black Friday deals 2022," and presses the search button. This query is then sent from the smartphone to the server.

[0281] The server analyzes the received query and generates and sends the following prompt to the natural language generation model:

[0282] Search past sales announcements: Black Friday 2022 campaigns

[0283] The response returned by the generative AI model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain past campaign information.

[0284] Detailed Request Processing

[0285] If a user wants to know more information about a specific product or campaign, they enter its ID into the smartphone interface. For example, if they want to know the details of the sales announcement with ID 1, they enter the ID and press the details button. The details request is sent from the smartphone to the server.

[0286] The server receives the request, retrieves the relevant information from the database, and then generates and sends the following prompt to the natural language generation model:

[0287] Detailed information about business notice: [Content of information obtained]

[0288] The detailed response returned by the natural language generation model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain the required details.

[0289] This invention enables users to quickly and efficiently search and obtain information about past campaigns and detailed product information. The collaboration between the database and the natural language generation model improves the user experience and enables efficient information acquisition.

[0290] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0291] Step 1:

[0292] A user uses a smartphone interface to enter a search query and presses the submit button, where the input is a keyword such as "Black Friday deals 2022." As an output, the smartphone sends this query to a server.

[0293] Step 2:

[0294] The server receives the search query sent by the user. The input is the user's search query, and the server analyzes the received query as output. As a result of the analysis, it generates a prompt such as "Search past sales announcements: Black Friday campaigns 2022."

[0295] Step 3:

[0296] The server sends the generated prompt to the natural language generation model. The input is the generated prompt, and the output is a response from the natural language generation model. The specific operation is to send the prompt via an API request and obtain the response as an API response.

[0297] Step 4:

[0298] The server analyzes the response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it processes the response text to format it for the user.

[0299] Step 5:

[0300] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[0301] Step 6:

[0302] The user inputs a detailed request including a specific ID into the smartphone interface and presses the send button. The input is "Details of the business announcement with ID 1." This request is sent from the smartphone to the server as an output.

[0303] Step 7:

[0304] The server receives the detailed request sent by the user. The input is the detailed request from the user, and the received request is analyzed as the output. As a result of the analysis, the information is retrieved from the database using the corresponding ID.

[0305] Step 8:

[0306] The server generates a prompt based on the information retrieved from the database and sends it to the natural language generation model. The input is the information retrieved from the database, and the generated prompt is sent as output. Specifically, it creates a prompt called "Sales Notice Details: [Content of retrieved information]" and sends it via an API request.

[0307] Step 9:

[0308] The server receives the detailed response returned by the NLG model. The input is the generated prompt and the output is the response from the NLG model. The action is to get the response as an API response.

[0309] Step 10:

[0310] The server parses the detailed response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it performs a process to format the response text.

[0311] Step 11:

[0312] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[0313] Step 12:

[0314] Detailed information will be displayed on the user's smartphone, allowing the user to quickly and accurately obtain the information they need.

[0315] 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.

[0316] The present invention provides a system for quickly and efficiently searching for previously distributed sales announcements and instantly responding to users about their content. The system includes a terminal where a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model. The system also combines an emotion engine that recognizes the user's emotions and adjusts the response based on those emotions to provide more personalized responses.

[0317] Search function embodiment

[0318] User enters and submits a search query

[0319] The user inputs a search query using the device interface, for example, "sales announcements for 2023." The device then identifies the emotion from the user's input and generates emotion data using an emotion engine.

[0320] Server receives and processes search queries

[0321] The server receives search queries and emotion data sent from the user's device, analyzes the received queries and emotion data, and generates prompts to send to the OpenAI API, which include adjustments based on the user's emotion.

[0322] Response generation and forwarding

[0323] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[0324] Detailed inquiry function embodiment

[0325] User enters and submits detailed request

[0326] The user inputs the ID of a specific business announcement and sends a request for details. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[0327] Server receives and processes detailed requests

[0328] The server receives the detailed request and emotion data sent from the user's device, extracts the ID included in the request, and retrieves the corresponding sales announcement information from the database. Based on the retrieved information, it generates a prompt to send to the OpenAI API. This prompt also includes adjustments based on the user's emotion.

[0329] Response generation and forwarding

[0330] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[0331] Specific examples

[0332] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[0333] {

[0334] "query": "Business Notice 2023",

[0335] "emotion": "neutral"

[0336] }

[0337] The server receives the query and emotion data, generates prompts for the natural language generation model, and sends the resulting responses back to the user. The emotion engine is tuned to generate responses with a "neutral" tone.

[0338] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[0339] GET / detail / 1

[0340] {

[0341] "emotion": "curious"

[0342] }

[0343] The server receives the request and emotion data, retrieves the relevant information from the database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user. The emotion engine is tuned to generate responses with a "curious" tone.

[0344] The present invention enables a user to quickly and efficiently search for business announcements distributed in the past and to receive an immediate response regarding the detailed contents thereof, and furthermore, by providing a response adapted to the user's emotions, a more satisfying service can be realized.

[0345] The processing flow will be explained below.

[0346] Search function processing steps

[0347] Step 1:

[0348] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[0349] Step 2:

[0350] When a user presses the search button, a search request including a query is sent from the terminal to the server. At this time, the terminal identifies the emotion from the user's input and generates emotion data using an emotion engine.

[0351] Step 3:

[0352] The server receives the search request and sentiment data, parses the query string and sentiment data, and generates a prompt to send to the natural language generation model.

[0353] Step 4:

[0354] The server sends the generated prompt to a natural language generation model for a response, where the prompt includes a modulation that reflects the user's sentiment.

[0355] Step 5:

[0356] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[0357] Step 6:

[0358] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[0359] Step 7:

[0360] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[0361] Processing steps for detailed inquiry function

[0362] Step 1:

[0363] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[0364] Step 2:

[0365] When the user presses the details button, a detailed request including the ID is sent from the device to the server. At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[0366] Step 3:

[0367] The server receives the detailed request and emotion data. Specifically, it extracts the ID included in the request and searches the database based on that ID to retrieve the relevant sales announcement.

[0368] Step 4:

[0369] The server uses the acquired sales pitch to generate prompts for transmission to a natural language generation model, which also include adjustments based on the user's sentiment.

[0370] Step 5:

[0371] The server sends the generated prompt to a natural language generation model and requests a response.

[0372] Step 6:

[0373] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[0374] Step 7:

[0375] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[0376] Step 8:

[0377] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[0378] Example 2

[0379] 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."

[0380] Conventional systems have been unable to consider user emotions when quickly and accurately responding to users' search queries and detailed requests. This can result in mechanical, unpersonalized responses, potentially reducing user satisfaction. Furthermore, prompts are generated statically, making it difficult to provide intuitive responses. Therefore, a system that takes user emotions into account and generates more personalized responses quickly and efficiently is needed.

[0381] 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.

[0382] In this invention, the server includes a means for receiving a search query from a user, a means for generating a prompt sentence using the search query, and a means for obtaining a response from a generative AI model using the prompt sentence. This makes it possible to dynamically generate a prompt sentence in response to the user's search query and quickly and efficiently obtain a response from the generative AI model. Furthermore, by including a means for identifying the user's emotion and generating emotion data, and a means for adjusting the tone or content of the response based on the user's emotion data, it is possible to provide a personalized response based on the user's emotion.

[0383] "User" means a person who utilizes the system to enter a search query or detailed request.

[0384] A "search query" is text that a user enters into a system to search for information.

[0385] A "prompt" is a text instruction generated from a search query or detail request and sent to a generative AI model.

[0386] A "generative AI model" is an artificial intelligence that automatically generates a response based on an input prompt.

[0387] "Emotion data" is data indicating a user's emotion identified from the user's search query or detail request.

[0388] The "emotion engine" is the part of the system that identifies emotions from user input and generates emotion data.

[0389] A "response" is information that a generative AI model generates based on a prompt and sends back to the user.

[0390] A "detail request" is a user request for more information based on specific identifying information.

[0391] "Identification information" refers to information that uniquely identifies specific information within a database.

[0392] A "database" is a collection of information used to retrieve information in response to a user's request.

[0393] "Tone" refers to the emotional tone or mood of the generated response, which is adjusted based on the user's emotions.

[0394] The present invention is a system that quickly and efficiently generates responses based on search queries and detailed requests entered by users. The system includes a terminal where users enter queries, a server for receiving and processing queries, and a function for generating responses using a generative AI model. It also combines an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions to provide more personalized responses.

[0395] Processing search queries

[0396] User enters and submits a search query

[0397] The user inputs a search query using the device interface. For example, they input "Business announcements for 2023." The device then identifies emotions from the user's input and generates emotion data using an emotion engine. The emotion data and the search query are then sent to the server.

[0398] Server receives and processes search queries

[0399] The server receives the search query and emotion data sent from the user's device. It analyzes the received query and emotion data and generates a prompt sentence to send to the generative AI model. This prompt includes adjustments based on the user's emotion. For example, it might be in the form of "Please provide detailed information about the sales announcements for 2023 in a neutral tone." The server then sends this prompt sentence to the generative AI model.

[0400] Response generation and forwarding

[0401] The generative AI model generates a response based on the prompt received from the server and sends it back to the server, which then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user.

[0402] Detail Request Processing

[0403] User enters and submits detailed request

[0404] The user inputs and sends a detailed request including specific identification information. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies emotions from the user's input and generates emotion data using an emotion engine. This emotion data and the detailed request are then sent to the server.

[0405] Server receives and processes detailed requests

[0406] The server receives the detailed request and emotion data sent from the user's device. It extracts the identification information included in the request and retrieves the corresponding sales announcement information from the database. For example, it retrieves information such as "ID: 1, Title: New Product Announcement, Content: Details about New Products for 2023." Based on this, it generates a prompt to send to the generative AI model. This prompt also includes adjustments based on the user's emotion. The server then sends this prompt to the generative AI model.

[0407] Response generation and forwarding

[0408] The generative AI model generates a response based on the prompt received from the server and sends it back to the server. The server then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user. For example, a response such as "The sales announcement with ID 1 is about a new product announcement. Details of the new product to be introduced in 2023 are..." may be generated.

[0409] In this way, the present invention can quickly and efficiently generate personalized responses that take into account the user's emotions. Specific examples of prompt sentences are shown below.

[0410] Prompt Sentence Examples

[0411] 1. Prompt based on search query:

[0412] "Please provide detailed information in a neutral tone regarding the 2023 business announcement."

[0413] 2. Prompt sentence based on detailed request:

[0414] "Please explain the details of the business announcement with ID 1 in an interesting tone."

[0415] In this way, responses to user inputs are provided in appropriate tone and content.

[0416] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0417] Step 1: User enters and submits a search query

[0418] The user inputs a search query using the device interface. For example, the user inputs "sales announcements for 2023." The device then generates emotion data using an emotion engine to identify emotions from the user's input. The device then sends the search query together with this emotion data to the server.

[0419] Input: Search query ("Sales announcements for 2023"), emotion data ("neutral")

[0420] Output: Query data sent to the server

[0421] Specific operation: The user enters "Sales announcement for 2023" into the device and clicks the send button. The emotion engine analyzes this input and generates emotion data called "neutral." The device then sends the query data to the server.

[0422] Step 2: The server receives the search query and generates a prompt.

[0423] The server receives the search query and emotion data sent from the user's device. After receiving the query, it analyzes the query content and generates a prompt to send to the generative AI model. This prompt includes adjustments based on the user's emotion, such as "Please provide detailed information in a neutral tone about the sales announcements for 2023."

[0424] Input: Query data received by the server (search query, sentiment data)

[0425] Output: A prompt to send to the generative AI model

[0426] Specific operation: The server analyzes the received query data and generates a prompt, such as "Please provide detailed information about the sales announcements for 2023 in a neutral tone," and prepares to send the prompt to the generation AI model.

[0427] Step 3: The server sends a prompt to the generative AI model

[0428] The server sends the generated prompt to the generative AI model, which then generates a response based on the prompt.

[0429] Input: A prompt to send to the generative AI model

[0430] Output: The response sent back from the generative AI model

[0431] Specific operation: The server generates a prompt and sends it to the generative AI model, which receives it and generates a response.

[0432] Step 4: Response generation and organization and coordination by the server

[0433] The server receives the responses sent back by the generative AI model and organizes them, adjusting the tone and content based on the user's emotions, sometimes again using the emotion engine.

[0434] Input: The response returned by the generative AI model

[0435] Output: The adjusted response that is sent back to the user

[0436] What it does: The response sent back to the server is processed by a natural language analysis module, and the emotion engine is used to adjust the tone and content, so the response might be formatted as something like, "Here's what our 2023 sales announcement will include..."

[0437] Step 5: Server sends back response

[0438] The server sends the adjusted response back to the user's terminal.

[0439] Input: Adjusted response

[0440] Output: The response that is displayed to the user

[0441] Specific operation: The server sends the adjusted response to the user's terminal, and the user confirms the response through the terminal interface.

[0442] (Application example 2)

[0443] 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."

[0444] Conventionally, when users search for trailers or reviews of previously distributed content (movies or TV dramas), they lack a means to quickly and efficiently obtain relevant information. Furthermore, there is a lack of systems that can provide personalized responses based on the user's emotions, creating a demand for an improved user experience. In particular, it is difficult to generate responses with an appropriate tone based on the user's emotions.

[0445] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0446] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, and means for recognizing the user's emotion using emotion recognition means and adjusting the response based on the emotion, thereby enabling the user to quickly and efficiently search for trailers and reviews of previously distributed content (movies and dramas) and further provide personalized responses according to the user's emotion.

[0447] A "search query" refers to a word or phrase that a user enters to obtain specific information.

[0448] A "natural language generation model" refers to an algorithm or system that is trained on a huge amount of text data to generate natural-sounding sentences similar to those written by humans.

[0449] "Response" refers to the content generated and returned by the system in response to a user's search query.

[0450] "Emotion recognition means" refers to technology or systems that identify emotions from user input, facial expressions, voice, etc., and extract them as specific emotional data.

[0451] A "prompt" refers to an input sentence that instructs a natural language generation model what kind of response to generate.

[0452] A "detail request" refers to a request from a user that includes a specific identifier and asks for a more detailed response about a particular piece of information.

[0453] "Database" refers to a system or location for organizing, storing, and managing information.

[0454] "Identifier" refers to a string of symbols, numbers, or letters that uniquely identifies specific information.

[0455] This invention is a system that allows users to quickly and efficiently search for trailers and reviews of previously distributed content (movies and TV dramas) and receive instant responses regarding the details. It also includes various means for providing personalized responses according to the user's emotions. Specific methods for implementing this invention are described below.

[0456] System Configuration

[0457] Hardware

[0458] Device: Smartphone

[0459] Server: High-performance cloud server

[0460] Emotion Recognition Device: A module for recognizing emotions from voice and text data

[0461] software

[0462] Natural Language Generation model: OpenAI API (text-davinci-003)

[0463] Emotion Recognition Software: Emotion Recognition Model (e.g., Emotion AI SDK)

[0464] Program processing

[0465] Query Reception and Emotion Recognition

[0466] A user inputs a search query using a smartphone interface, for example, "Reviews of the 2021 movie 'Inception.'" The device receives the query and uses an emotion recognition device to identify the user's emotion from the input query.

[0467] Prompt Generation

[0468] The server generates a prompt to send to the natural language generation model based on the received search query and sentiment data. An example of a generated prompt would be, "I'm excited. I'm asking about a review of the 2021 film 'Inception.' I'd appreciate your information and support."

[0469] Generating and returning a response

[0470] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the prompt and sends it back to the server. The server then organizes the response, adjusts the tone and content to match the user's emotions, and sends it back to the user's smartphone.

[0471] Specific examples

[0472] For example, if a user searches for "2021 movie 'Inception' reviews," they'll get the following:

[0473] A user types "Review of the 2021 film Inception" into their device.

[0474] The terminal receives this query and obtains the emotion data "excited" from the emotion recognition device.

[0475] The server generates a prompt based on the query and emotion data and sends it to the OpenAI API in the form, "Emotion is excited. I'm asking about a review of the 2021 movie 'Inception'. Please provide information and support."

[0476] The OpenAI API generates a response and sends it back to the server.

[0477] The server sorts the responses, adjusts the tone appropriately based on the emotion of "excitement," and sends them back to the user's terminal.

[0478] The system allows users to quickly learn about previously distributed content and receive personalized, emotion-based responses.

[0479] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0480] Step 1:

[0481] A user enters a search query into a device.

[0482] Specifically, a user opens the application on their smartphone and types "reviews of the 2021 film 'Inception'" into the search bar.

[0483] Input: Search query (e.g., "2021 movie 'Inception' reviews")

[0484] Output: The search query entered by the user

[0485] Step 2:

[0486] The terminal receives the search query and recognizes the user's emotion using the emotion recognition means.

[0487] For example, the device analyzes the voice and word choice from the input text to identify the emotion "excitement."

[0488] Input: Search query from user

[0489] Data processing: Emotion recognition software analyzes text and extracts emotional data

[0490] Output: Search query and sentiment data (e.g., "excited")

[0491] Step 3:

[0492] The device sends the search query and emotion data to the server.

[0493] Specifically, the device generates a data packet to send the search query "2021 movie 'Inception' reviews" and the emotion "excited" to the server.

[0494] Input: Search query and sentiment data

[0495] Output: Data packet sent to the server

[0496] Step 4:

[0497] The server generates prompts based on the received search query and sentiment data.

[0498] For example, create a prompt like, "Emotions are excited. I'm asking about reviews of the 2021 film Inception. I need your input and support."

[0499] Input: Search query and sentiment data

[0500] Data processing: Prompt generation algorithm generates appropriate prompts

[0501] Output: Generated prompt

[0502] Step 5:

[0503] The server sends the generated prompt to the OpenAI API and requests a response.

[0504] Specifically, the server constructs an API request and sends it, including the prompt, to OpenAI's endpoint.

[0505] Input: Generated prompt

[0506] Output: The prompt sent to the OpenAI API

[0507] Step 6:

[0508] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[0509] For example, it might generate a response along the lines of, "Inception is a film directed by Christopher Nolan that has received very high praise overall..."

[0510] Input: Prompt sent

[0511] Data Computation: Response Generation with OpenAI's Natural Language Generation Model

[0512] Output: The generated response

[0513] Step 7:

[0514] The server sorts the returned responses, adjusts the tone and content to match the user's emotions, and then sends them back to the user's device.

[0515] For example, adjust the tone of your response to be more energetic based on the emotion "excited."

[0516] Input: The generated response

[0517] Data manipulation: Emotion engine adjusts tone and content of responses

[0518] Output: The adjusted response

[0519] Step 8:

[0520] The terminal receives the adjusted response and displays it to the user.

[0521] Specifically, the response content is displayed on the application's results screen.

[0522] Input: Adjusted response

[0523] Output: The response that is displayed to the user

[0524] Through these steps, users can quickly and efficiently search for trailers and reviews of previously distributed content, and obtain information that is appropriately tailored based on sentiment.

[0525] 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.

[0526] 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.

[0527] 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.

[0528] [Second embodiment]

[0529] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0530] 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.

[0531] 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).

[0532] 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.

[0533] 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.

[0534] 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).

[0535] 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.

[0536] 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.

[0537] 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.

[0538] 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.

[0539] 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.

[0540] 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."

[0541] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and responding to users immediately regarding their content. The system includes a terminal through which a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[0542] Search function embodiment

[0543] User enters and submits a search query

[0544] The user inputs a search query using the device interface. For example, they input the keyword "Business Announcement Contents for 2023" and press the search button. This operation sends the search query from the device to the server.

[0545] Server receives and processes search queries

[0546] The server receives a search query sent from the user's device, analyzes the received query, and generates and sends a prompt to a natural language generation model (e.g., OpenAI's API) based on the query's content. The natural language generation model generates a response related to the query and sends it back to the server.

[0547] Response generation and forwarding

[0548] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's device is based on the user's query, providing information quickly and accurately.

[0549] Detailed inquiry function embodiment

[0550] User enters and submits detailed request

[0551] To find out more about a specific business announcement, the user inputs its ID and sends a details request. For example, if the user wants to know the details of the business announcement with ID 1, the user inputs the ID and presses the details button. This operation sends a details request from the terminal to the server.

[0552] Server receives and processes detailed requests

[0553] The server receives a detailed request sent from the user's device. It analyzes the ID included in the request and retrieves the relevant sales notice information from the database. It generates and sends a prompt to the natural language generation model based on the relevant information. The natural language generation model generates a detailed response and sends it back to the server.

[0554] Response generation and forwarding

[0555] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's terminal is based on the detailed request, providing information quickly and accurately.

[0556] Specific examples

[0557] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[0558] {

[0559] "query": "2023 Business Announcement"

[0560] }

[0561] The server receives this query and generates and sends prompts to the natural language generation model, which then generates a response and returns it to the user.

[0562] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[0563] GET / detail / 1

[0564] The server receives this request, retrieves the relevant information from a database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user.

[0565] This system allows users to quickly and efficiently search for previously distributed business announcements and get immediate responses to their detailed information, significantly improving the efficiency of managing and using business announcements.

[0566] The processing flow will be explained below.

[0567] Search function processing steps

[0568] Step 1:

[0569] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[0570] Step 2:

[0571] The user presses the search button to send a search request including the query from the terminal to the server.

[0572] Step 3:

[0573] The server receives the search request, parses the query string, and generates a prompt to send to the OpenAI API.

[0574] Step 4:

[0575] The server sends the generated prompt to the OpenAI API and requests a response.

[0576] Step 5:

[0577] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[0578] Step 6:

[0579] The server compiles the response received from the OpenAI API and sends it back to the user.

[0580] Step 7:

[0581] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[0582] Processing steps for detailed inquiry function

[0583] Step 1:

[0584] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[0585] Step 2:

[0586] The user presses the details button, and a details request including the ID is sent from the terminal to the server.

[0587] Step 3:

[0588] The server receives the detailed request, extracts the ID included in the request, and searches the database based on that ID to retrieve the relevant business announcement.

[0589] Step 4:

[0590] The server uses the acquired business notice to generate a prompt to send to the OpenAI API.

[0591] Step 5:

[0592] The server sends the generated prompt to the OpenAI API and requests a response.

[0593] Step 6:

[0594] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[0595] Step 7:

[0596] The server compiles the response received from the OpenAI API and sends it back to the user.

[0597] Step 8:

[0598] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[0599] Example 1

[0600] 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."

[0601] Conventional information retrieval systems have difficulty generating fast and accurate responses to specific keywords or detailed requests. Furthermore, more flexible and efficient information processing methods are needed to respond to diverse user requests. In particular, there is a need for systems that can accurately search past information and respond to users immediately, but there are not enough systems available that can meet these demands.

[0602] 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.

[0603] In this invention, the server includes means for receiving input from a user, means for dynamically generating prompts for a generative model based on the input, and means for obtaining responses from the generative model using the dynamically generated prompts, thereby enabling the server to respond to a variety of user requests and provide responses quickly and accurately.

[0604] "Means for receiving input from a user" refers to the functionality or interface that allows a user to input information and for the system to receive it.

[0605] "Means for dynamically generating prompts for a generative model" refers to a function that automatically generates questions (prompts) that enable the generative model to return an appropriate response based on input information.

[0606] The "means for obtaining a response from a generative model" refers to a function for receiving a response returned from a generative model based on a generated prompt.

[0607] "Means for returning to the user" refers to a function within the system that sends information to the user's terminal in order to display the obtained response to the user.

[0608] The term "means for receiving a request including a specific identifier" refers to a function whereby a user provides an identifier for specifying specific information and the system receives the identifier.

[0609] "Means for obtaining corresponding content from information storage means" refers to the function of searching and retrieving information based on a specified identifier from a database or storage within the system.

[0610] The "means for dynamically generating a prompt based on specific conditions" refers to a function for dynamically generating an appropriate prompt in response to a user's input or request.

[0611] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and instantly responding to the user about their content. The system includes a terminal through which the user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[0612] Search function embodiment

[0613] User enters a search query

[0614] The user uses their own device to input the keywords they want to search for into the interface. For example, they can input "2023 business announcements" and press the search button to confirm the query input.

[0615] Sending queries by terminal

[0616] After pressing the search button, the terminal sends the entered query to the server. The data sent is as follows:

[0617] {

[0618] "query": "2023 Business Announcement"

[0619] }

[0620] Server receives and processes search queries

[0621] The server receives the search query sent from the device, analyzes the query, and dynamically generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[0622] "Please tell me what the business announcement for 2023 will be."

[0623] Prompting a Natural Language Generation Model

[0624] The server sends the generated prompts to a natural language generation model (e.g., a popular generative AI API), which generates a response based on the prompts and sends the result back to the server.

[0625] Receiving and formatting the response by the server

[0626] The server receives the response sent back by the generative model and formats it appropriately, which may include converting it to HTML or JSON format.

[0627] Terminal display of responses

[0628] The terminal receives the formatted response from the server and displays it to the user, allowing the user to quickly see the results of their query.

[0629] Detailed inquiry function embodiment

[0630] User enters detailed request

[0631] If the user wants to know the details of a specific business announcement, he or she inputs its ID and presses the Details button. For example, the user inputs "Details of the business announcement with ID 1."

[0632] Sending a details request by terminal

[0633] After pressing the Details button, the device will send the input request to the server, which will have the following format:

[0634] GET / detail / 1

[0635] Server receives and processes detailed requests

[0636] The server receives the detailed request sent from the device, analyzes the ID included in the request, and searches the database. Based on the information obtained from the database, it generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[0637] "Please tell me the details of the business announcement with ID 1."

[0638] Prompting a Natural Language Generation Model

[0639] The server sends the generated prompt to a natural language generation model to generate a detailed response, which is then sent back to the server.

[0640] Server response formatting and forwarding

[0641] The server receives the response from the generative model, formats it in an appropriate format, and sends it back to the device. Examples of returned data are as follows:

[0642] {

[0643] "response": "The details of the sales announcement with ID 1 are as follows..."

[0644] }

[0645] Terminal display of responses

[0646] The terminal displays the response received from the server on the screen, allowing the user to quickly check the correct information for the detailed request.

[0647] The present invention allows users to quickly and efficiently search past business announcements and get immediate responses to detailed information, greatly improving the efficiency of business announcement management and utilization.

[0648] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0649] Search function embodiment

[0650] Step 1:

[0651] User enters a search query

[0652] The user inputs the keywords they want to search for using the interface of their device. For example, they input "Business Notices for 2023" and press the search button. This operation confirms the query they entered.

[0653] Input: "Business Notice for 2023"

[0654] Output: The query is typed into the terminal

[0655] Step 2:

[0656] Sending queries by terminal

[0657] When the search button is pressed, the device sends the entered query to the server. Specifically, the query is packaged in JSON format and sent to the server as an HTTP request.

[0658] Input: "Business Notice for 2023"

[0659] Output: The query is sent to the server as an HTTP request.

[0660] Step 3:

[0661] Server receives search query

[0662] The server receives the query sent from the device, parses the received data in JSON format, and extracts the search keywords.

[0663] Input: Query from terminal

[0664] Output: Parsed search keywords

[0665] Step 4:

[0666] The server parses the query and generates prompts

[0667] The server analyzes the received query and dynamically generates a prompt based on its content, which becomes the question sent to the generative AI model.

[0668] How it works: Builds a prompt from search keywords.

[0669] Input: Parsed search keywords

[0670] Output: Generated prompt: "What is the content of your 2023 sales announcement?"

[0671] Step 5:

[0672] Prompting a Natural Language Generation Model

[0673] The server sends the generated prompt to a generative AI model (e.g., a general generative AI API).

[0674] What it does: Sends the prompt to the generative AI model as an HTTP request.

[0675] Input: Generated prompt

[0676] Output: A request to the NLG model

[0677] Step 6:

[0678] Receiving responses from generative AI models

[0679] The server receives a response from the generative AI model, which is a natural language answer generated based on the prompt.

[0680] Behavior: Receives the response as an HTTP response and parses it.

[0681] Input: The response returned by the generative AI model

[0682] Output: Parsed response content

[0683] Step 7:

[0684] Server response formatting

[0685] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display to the user.

[0686] Behavior: Formats the response content.

[0687] Input: Response from the generative AI model

[0688] Output: Formatted response data

[0689] Step 8:

[0690] Response from the server to the device

[0691] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[0692] Behavior: Sends response data as an HTTP response.

[0693] Input: Formatted response data

[0694] Output: Send response data to the terminal

[0695] Step 9:

[0696] Terminal display of responses

[0697] The terminal displays the response received from the server to the user, allowing the user to check the search results.

[0698] Behavior: Displays the response data in the interface.

[0699] Input: Response data from the server

[0700] Output: What is displayed to the user

[0701] Detailed inquiry function embodiment

[0702] Step 1:

[0703] User enters detailed request

[0704] If the user wants to know the details of a specific business announcement, he or she enters its ID and presses the Details button. For example, the user might enter "Details of the business announcement with ID 1." This operation confirms the request for details.

[0705] Input: "Details of business announcement with ID 1"

[0706] Output: Detailed request typed into terminal

[0707] Step 2:

[0708] Sending a details request by terminal

[0709] When the details button is pressed, the terminal transmits the input request to the server. Specifically, the request is transmitted to the server as an HTTP GET request.

[0710] Input: "Details of business announcement with ID 1"

[0711] Output: The detailed request is sent to the server as an HTTP GET request.

[0712] Step 3:

[0713] Server receives detailed request

[0714] The server receives a detailed request containing the ID sent by the device, which is then parsed to extract the ID.

[0715] Input: Detailed request from terminal

[0716] Output: Parsed ID

[0717] Step 4:

[0718] Server parses the request and performs a database lookup

[0719] The server searches the database based on the ID included in the detailed request and obtains the relevant business notice information.

[0720] What it does: Searches the database based on the ID and retrieves the relevant information.

[0721] Input: Parsed ID

[0722] Output: Business announcement information retrieved from the database

[0723] Step 5:

[0724] Server-generated prompts

[0725] Based on the information obtained, the server generates a prompt that is sent to the generative AI model, asking for more information.

[0726] How it works: Construct a prompt from the information obtained.

[0727] Input: Business announcement information retrieved from the database

[0728] Output: Generated prompt: "Please tell me the details of the sales announcement with ID 1."

[0729] Step 6:

[0730] Prompting a Natural Language Generation Model

[0731] The server sends the generated prompts to the generative AI model.

[0732] What it does: Sends the prompt to the generative AI model as an HTTP request.

[0733] Input: Generated prompt

[0734] Output: A request to the NLG model

[0735] Step 7:

[0736] Receiving responses from generative AI models

[0737] The server receives a response from the generative AI model, which is a detailed answer generated based on the prompt.

[0738] Behavior: Receives the response as an HTTP response and parses it.

[0739] Input: The response returned by the generative AI model

[0740] Output: Parsed response content

[0741] Step 8:

[0742] Server response formatting

[0743] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display on the device.

[0744] Behavior: Formats the response content.

[0745] Input: Response from the generative AI model

[0746] Output: Formatted response data

[0747] Step 9:

[0748] Response from the server to the device

[0749] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[0750] Behavior: Sends response data as an HTTP response.

[0751] Input: Formatted response data

[0752] Output: Send response data to the terminal

[0753] Step 10:

[0754] Terminal display of responses

[0755] The terminal displays the response received from the server to the user, allowing the user to quickly check the information for the detailed request.

[0756] Behavior: Displays the response data in the interface.

[0757] Input: Response data from the server

[0758] Output: What is displayed to the user

[0759] (Application example 1)

[0760] 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."

[0761] On current online shopping sites, it can take a significant amount of time for users to quickly search and obtain information about past campaigns or detailed product information. Furthermore, there is a lack of systems that accurately provide the information users are looking for, resulting in a poor user experience. The present invention aims to solve these problems and provide a system that allows users to efficiently and quickly obtain the information they need.

[0762] 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.

[0763] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, means for providing a database for efficiently managing information searched in the past, and means for inputting a search query through an interface on a smartphone, thereby enabling users to quickly and accurately search and obtain information on past campaigns and detailed product information.

[0764] A "search query" is a keyword or phrase that a user enters to locate specific information.

[0765] A "natural language generation model" is an artificial intelligence technique that generates relevant sentences and responses based on input text.

[0766] A "response" is the information or answer provided by a system in response to a user's search query.

[0767] A "database" is a collection of data that stores previously searched information and detailed information in an organized manner so that it can be efficiently searched and retrieved later.

[0768] A "smartphone" is a type of mobile phone that has internet connectivity and a variety of applications, allowing users to perform various searches and obtain information.

[0769] An "interface" is a screen or operating means that a user uses to interact with a system.

[0770] A "detail request" is a request including a specific ID that a user enters when they want to know more about a specific piece of information.

[0771] The present invention is a system for quickly retrieving information and detailed information that a user has previously searched for. This system is operated via an interface installed on the user's smartphone. The main components of the system include the following:

[0772] System Configuration

[0773] Hardware elements

[0774] 1. Smartphone: A device where users enter search queries and detailed requests. It communicates with the server through an interface.

[0775] 2. Server: Receives queries and requests, retrieves information from the database, and generates and sends prompts to the natural language generation model.

[0776] Software Elements

[0777] 1. Interface: A smartphone application built using React Native that provides an interface for users to enter search queries and detailed requests.

[0778] 2. Server application: Built using Node.js and the Express framework, it handles user queries and requests, generates prompts, and sends them to the natural language generation model.

[0779] 3. Database Management System: MongoDB is used to store and manage past search queries and detailed information.

[0780] 4. Natural language generation models: Generative AI models provided as APIs (e.g., OpenAI's API) are used to generate relevant responses based on input queries.

[0781] Processing flow

[0782] Query Processing

[0783] A user uses the smartphone interface to enter a search query, for example, "Black Friday deals 2022," and presses the search button. This query is then sent from the smartphone to the server.

[0784] The server analyzes the received query and generates and sends the following prompt to the natural language generation model:

[0785] Search past sales announcements: Black Friday 2022 campaigns

[0786] The response returned by the generative AI model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain past campaign information.

[0787] Detailed Request Processing

[0788] If a user wants to know more information about a specific product or campaign, they enter its ID into the smartphone interface. For example, if they want to know the details of the sales announcement with ID 1, they enter the ID and press the details button. The details request is sent from the smartphone to the server.

[0789] The server receives the request, retrieves the relevant information from the database, and then generates and sends the following prompt to the natural language generation model:

[0790] Detailed information about business notice: [Content of information obtained]

[0791] The detailed response returned by the natural language generation model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain the required details.

[0792] This invention enables users to quickly and efficiently search and obtain information about past campaigns and detailed product information. The collaboration between the database and the natural language generation model improves the user experience and enables efficient information acquisition.

[0793] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0794] Step 1:

[0795] A user uses a smartphone interface to enter a search query and presses the submit button, where the input is a keyword such as "Black Friday deals 2022." As an output, the smartphone sends this query to a server.

[0796] Step 2:

[0797] The server receives the search query sent by the user. The input is the user's search query, and the server analyzes the received query as output. As a result of the analysis, it generates a prompt such as "Search past sales announcements: Black Friday campaigns 2022."

[0798] Step 3:

[0799] The server sends the generated prompt to the natural language generation model. The input is the generated prompt, and the output is a response from the natural language generation model. The specific operation is to send the prompt via an API request and obtain the response as an API response.

[0800] Step 4:

[0801] The server analyzes the response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it processes the response text to format it for the user.

[0802] Step 5:

[0803] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[0804] Step 6:

[0805] The user inputs a detailed request including a specific ID into the smartphone interface and presses the send button. The input is "Details of the business announcement with ID 1." This request is sent from the smartphone to the server as an output.

[0806] Step 7:

[0807] The server receives the detailed request sent by the user. The input is the detailed request from the user, and the received request is analyzed as the output. As a result of the analysis, the information is retrieved from the database using the corresponding ID.

[0808] Step 8:

[0809] The server generates a prompt based on the information retrieved from the database and sends it to the natural language generation model. The input is the information retrieved from the database, and the generated prompt is sent as output. Specifically, it creates a prompt called "Sales Notice Details: [Content of retrieved information]" and sends it via an API request.

[0810] Step 9:

[0811] The server receives the detailed response returned by the NLG model. The input is the generated prompt and the output is the response from the NLG model. The action is to get the response as an API response.

[0812] Step 10:

[0813] The server parses the detailed response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it performs a process to format the response text.

[0814] Step 11:

[0815] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[0816] Step 12:

[0817] Detailed information will be displayed on the user's smartphone, allowing the user to quickly and accurately obtain the information they need.

[0818] 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.

[0819] The present invention provides a system for quickly and efficiently searching for previously distributed sales announcements and instantly responding to users about their content. The system includes a terminal where a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model. The system also combines an emotion engine that recognizes the user's emotions and adjusts the response based on those emotions to provide more personalized responses.

[0820] Search function embodiment

[0821] User enters and submits a search query

[0822] The user inputs a search query using the device interface, for example, "sales announcements for 2023." The device then identifies the emotion from the user's input and generates emotion data using an emotion engine.

[0823] Server receives and processes search queries

[0824] The server receives search queries and emotion data sent from the user's device, analyzes the received queries and emotion data, and generates prompts to send to the OpenAI API, which include adjustments based on the user's emotion.

[0825] Response generation and forwarding

[0826] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[0827] Detailed inquiry function embodiment

[0828] User enters and submits detailed request

[0829] The user inputs the ID of a specific business announcement and sends a request for details. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[0830] Server receives and processes detailed requests

[0831] The server receives the detailed request and emotion data sent from the user's device, extracts the ID included in the request, and retrieves the corresponding sales announcement information from the database. Based on the retrieved information, it generates a prompt to send to the OpenAI API. This prompt also includes adjustments based on the user's emotion.

[0832] Response generation and forwarding

[0833] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[0834] Specific examples

[0835] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[0836] {

[0837] "query": "Business Notice 2023",

[0838] "emotion": "neutral"

[0839] }

[0840] The server receives the query and emotion data, generates prompts for the natural language generation model, and sends the resulting responses back to the user. The emotion engine is tuned to generate responses with a "neutral" tone.

[0841] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[0842] GET / detail / 1

[0843] {

[0844] "emotion": "curious"

[0845] }

[0846] The server receives the request and emotion data, retrieves the relevant information from the database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user. The emotion engine is tuned to generate responses with a "curious" tone.

[0847] The present invention enables a user to quickly and efficiently search for business announcements distributed in the past and to receive an immediate response regarding the detailed contents thereof, and furthermore, by providing a response adapted to the user's emotions, a more satisfying service can be realized.

[0848] The processing flow will be explained below.

[0849] Search function processing steps

[0850] Step 1:

[0851] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[0852] Step 2:

[0853] When a user presses the search button, a search request including a query is sent from the terminal to the server. At this time, the terminal identifies the emotion from the user's input and generates emotion data using an emotion engine.

[0854] Step 3:

[0855] The server receives the search request and sentiment data, parses the query string and sentiment data, and generates a prompt to send to the natural language generation model.

[0856] Step 4:

[0857] The server sends the generated prompt to a natural language generation model for a response, where the prompt includes a modulation that reflects the user's sentiment.

[0858] Step 5:

[0859] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[0860] Step 6:

[0861] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[0862] Step 7:

[0863] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[0864] Processing steps for detailed inquiry function

[0865] Step 1:

[0866] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[0867] Step 2:

[0868] When the user presses the details button, a detailed request including the ID is sent from the device to the server. At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[0869] Step 3:

[0870] The server receives the detailed request and emotion data. Specifically, it extracts the ID included in the request and searches the database based on that ID to retrieve the relevant sales announcement.

[0871] Step 4:

[0872] The server uses the acquired sales pitch to generate prompts for transmission to a natural language generation model, which also include adjustments based on the user's sentiment.

[0873] Step 5:

[0874] The server sends the generated prompt to a natural language generation model and requests a response.

[0875] Step 6:

[0876] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[0877] Step 7:

[0878] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[0879] Step 8:

[0880] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[0881] Example 2

[0882] 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."

[0883] Conventional systems have been unable to consider user emotions when quickly and accurately responding to users' search queries and detailed requests. This can result in mechanical, unpersonalized responses, potentially reducing user satisfaction. Furthermore, prompts are generated statically, making it difficult to provide intuitive responses. Therefore, a system that takes user emotions into account and generates more personalized responses quickly and efficiently is needed.

[0884] 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.

[0885] In this invention, the server includes a means for receiving a search query from a user, a means for generating a prompt sentence using the search query, and a means for obtaining a response from a generative AI model using the prompt sentence. This makes it possible to dynamically generate a prompt sentence in response to the user's search query and quickly and efficiently obtain a response from the generative AI model. Furthermore, by including a means for identifying the user's emotion and generating emotion data, and a means for adjusting the tone or content of the response based on the user's emotion data, it is possible to provide a personalized response based on the user's emotion.

[0886] "User" means a person who utilizes the system to enter a search query or detailed request.

[0887] A "search query" is text that a user enters into a system to search for information.

[0888] A "prompt" is a text instruction generated from a search query or detail request and sent to a generative AI model.

[0889] A "generative AI model" is an artificial intelligence that automatically generates a response based on an input prompt.

[0890] "Emotion data" is data indicating a user's emotion identified from the user's search query or detail request.

[0891] The "emotion engine" is the part of the system that identifies emotions from user input and generates emotion data.

[0892] A "response" is information that a generative AI model generates based on a prompt and sends back to the user.

[0893] A "detail request" is a user request for more information based on specific identifying information.

[0894] "Identification information" refers to information that uniquely identifies specific information within a database.

[0895] A "database" is a collection of information used to retrieve information in response to a user's request.

[0896] "Tone" refers to the emotional tone or mood of the generated response, which is adjusted based on the user's emotions.

[0897] The present invention is a system that quickly and efficiently generates responses based on search queries and detailed requests entered by users. The system includes a terminal where users enter queries, a server for receiving and processing queries, and a function for generating responses using a generative AI model. It also combines an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions to provide more personalized responses.

[0898] Processing search queries

[0899] User enters and submits a search query

[0900] The user inputs a search query using the device interface. For example, they input "Business announcements for 2023." The device then identifies emotions from the user's input and generates emotion data using an emotion engine. The emotion data and the search query are then sent to the server.

[0901] Server receives and processes search queries

[0902] The server receives the search query and emotion data sent from the user's device. It analyzes the received query and emotion data and generates a prompt sentence to send to the generative AI model. This prompt includes adjustments based on the user's emotion. For example, it might be in the form of "Please provide detailed information about the sales announcements for 2023 in a neutral tone." The server then sends this prompt sentence to the generative AI model.

[0903] Response generation and forwarding

[0904] The generative AI model generates a response based on the prompt received from the server and sends it back to the server, which then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user.

[0905] Detail Request Processing

[0906] User enters and submits detailed request

[0907] The user inputs and sends a detailed request including specific identification information. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies emotions from the user's input and generates emotion data using an emotion engine. This emotion data and the detailed request are then sent to the server.

[0908] Server receives and processes detailed requests

[0909] The server receives the detailed request and emotion data sent from the user's device. It extracts the identification information included in the request and retrieves the corresponding sales announcement information from the database. For example, it retrieves information such as "ID: 1, Title: New Product Announcement, Content: Details about New Products for 2023." Based on this, it generates a prompt to send to the generative AI model. This prompt also includes adjustments based on the user's emotion. The server then sends this prompt to the generative AI model.

[0910] Response generation and forwarding

[0911] The generative AI model generates a response based on the prompt received from the server and sends it back to the server. The server then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user. For example, a response such as "The sales announcement with ID 1 is about a new product announcement. Details of the new product to be introduced in 2023 are..." may be generated.

[0912] In this way, the present invention can quickly and efficiently generate personalized responses that take into account the user's emotions. Specific examples of prompt sentences are shown below.

[0913] Prompt Sentence Examples

[0914] 1. Prompt based on search query:

[0915] "Please provide detailed information in a neutral tone regarding the 2023 business announcement."

[0916] 2. Prompt sentence based on detailed request:

[0917] "Please explain the details of the business announcement with ID 1 in an interesting tone."

[0918] In this way, responses to user inputs are provided in appropriate tone and content.

[0919] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0920] Step 1: User enters and submits a search query

[0921] The user inputs a search query using the device interface. For example, the user inputs "sales announcements for 2023." The device then generates emotion data using an emotion engine to identify emotions from the user's input. The device then sends the search query together with this emotion data to the server.

[0922] Input: Search query ("Sales announcements for 2023"), emotion data ("neutral")

[0923] Output: Query data sent to the server

[0924] Specific operation: The user enters "Sales announcement for 2023" into the device and clicks the send button. The emotion engine analyzes this input and generates emotion data called "neutral." The device then sends the query data to the server.

[0925] Step 2: The server receives the search query and generates a prompt.

[0926] The server receives the search query and emotion data sent from the user's device. After receiving the query, it analyzes the query content and generates a prompt to send to the generative AI model. This prompt includes adjustments based on the user's emotion, such as "Please provide detailed information in a neutral tone about the sales announcements for 2023."

[0927] Input: Query data received by the server (search query, sentiment data)

[0928] Output: A prompt to send to the generative AI model

[0929] Specific operation: The server analyzes the received query data and generates a prompt, such as "Please provide detailed information about the sales announcements for 2023 in a neutral tone," and prepares to send the prompt to the generation AI model.

[0930] Step 3: The server sends a prompt to the generative AI model

[0931] The server sends the generated prompt to the generative AI model, which then generates a response based on the prompt.

[0932] Input: A prompt to send to the generative AI model

[0933] Output: The response sent back from the generative AI model

[0934] Specific operation: The server generates a prompt and sends it to the generative AI model, which receives it and generates a response.

[0935] Step 4: Response generation and organization and coordination by the server

[0936] The server receives the responses sent back by the generative AI model and organizes them, adjusting the tone and content based on the user's emotions, sometimes again using the emotion engine.

[0937] Input: The response returned by the generative AI model

[0938] Output: The adjusted response that is sent back to the user

[0939] What it does: The response sent back to the server is processed by a natural language analysis module, and the emotion engine is used to adjust the tone and content, so the response might be formatted as something like, "Here's what our 2023 sales announcement will include..."

[0940] Step 5: Server sends back response

[0941] The server sends the adjusted response back to the user's terminal.

[0942] Input: Adjusted response

[0943] Output: The response that is displayed to the user

[0944] Specific operation: The server sends the adjusted response to the user's terminal, and the user confirms the response through the terminal interface.

[0945] (Application example 2)

[0946] 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."

[0947] Conventionally, when users search for trailers or reviews of previously distributed content (movies or TV dramas), they lack a means to quickly and efficiently obtain relevant information. Furthermore, there is a lack of systems that can provide personalized responses based on the user's emotions, creating a demand for an improved user experience. In particular, it is difficult to generate responses with an appropriate tone based on the user's emotions.

[0948] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0949] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, and means for recognizing the user's emotion using emotion recognition means and adjusting the response based on the emotion, thereby enabling the user to quickly and efficiently search for trailers and reviews of previously distributed content (movies and dramas) and further provide personalized responses according to the user's emotion.

[0950] A "search query" refers to a word or phrase that a user enters to obtain specific information.

[0951] A "natural language generation model" refers to an algorithm or system that is trained on a huge amount of text data to generate natural-sounding sentences similar to those written by humans.

[0952] "Response" refers to the content generated and returned by the system in response to a user's search query.

[0953] "Emotion recognition means" refers to technology or systems that identify emotions from user input, facial expressions, voice, etc., and extract them as specific emotional data.

[0954] A "prompt" refers to an input sentence that instructs a natural language generation model what kind of response to generate.

[0955] A "detail request" refers to a request from a user that includes a specific identifier and asks for a more detailed response about a particular piece of information.

[0956] "Database" refers to a system or location for organizing, storing, and managing information.

[0957] "Identifier" refers to a string of symbols, numbers, or letters that uniquely identifies specific information.

[0958] This invention is a system that allows users to quickly and efficiently search for trailers and reviews of previously distributed content (movies and TV dramas) and receive instant responses regarding the details. It also includes various means for providing personalized responses according to the user's emotions. Specific methods for implementing this invention are described below.

[0959] System Configuration

[0960] Hardware

[0961] Device: Smartphone

[0962] Server: High-performance cloud server

[0963] Emotion Recognition Device: A module for recognizing emotions from voice and text data

[0964] software

[0965] Natural Language Generation model: OpenAI API (text-davinci-003)

[0966] Emotion Recognition Software: Emotion Recognition Model (e.g., Emotion AI SDK)

[0967] Program processing

[0968] Query Reception and Emotion Recognition

[0969] A user inputs a search query using a smartphone interface, for example, "Reviews of the 2021 movie 'Inception.'" The device receives the query and uses an emotion recognition device to identify the user's emotion from the input query.

[0970] Prompt Generation

[0971] The server generates a prompt to send to the natural language generation model based on the received search query and sentiment data. An example of a generated prompt would be, "I'm excited. I'm asking about a review of the 2021 film 'Inception.' I'd appreciate your information and support."

[0972] Generating and returning a response

[0973] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the prompt and sends it back to the server. The server then organizes the response, adjusts the tone and content to match the user's emotions, and sends it back to the user's smartphone.

[0974] Specific examples

[0975] For example, if a user searches for "2021 movie 'Inception' reviews," they'll get the following:

[0976] A user types "Review of the 2021 film Inception" into their device.

[0977] The terminal receives this query and obtains the emotion data "excited" from the emotion recognition device.

[0978] The server generates a prompt based on the query and emotion data and sends it to the OpenAI API in the form, "Emotion is excited. I'm asking about a review of the 2021 movie 'Inception'. Please provide information and support."

[0979] The OpenAI API generates a response and sends it back to the server.

[0980] The server sorts the responses, adjusts the tone appropriately based on the emotion of "excitement," and sends them back to the user's terminal.

[0981] The system allows users to quickly learn about previously distributed content and receive personalized, emotion-based responses.

[0982] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0983] Step 1:

[0984] A user enters a search query into a device.

[0985] Specifically, a user opens the application on their smartphone and types "reviews of the 2021 film 'Inception'" into the search bar.

[0986] Input: Search query (e.g., "2021 movie 'Inception' reviews")

[0987] Output: The search query entered by the user

[0988] Step 2:

[0989] The terminal receives the search query and recognizes the user's emotion using the emotion recognition means.

[0990] For example, the device analyzes the voice and word choice from the input text to identify the emotion "excitement."

[0991] Input: Search query from user

[0992] Data processing: Emotion recognition software analyzes text and extracts emotional data

[0993] Output: Search query and sentiment data (e.g., "excited")

[0994] Step 3:

[0995] The device sends the search query and emotion data to the server.

[0996] Specifically, the device generates a data packet to send the search query "2021 movie 'Inception' reviews" and the emotion "excited" to the server.

[0997] Input: Search query and sentiment data

[0998] Output: Data packet sent to the server

[0999] Step 4:

[1000] The server generates prompts based on the received search query and sentiment data.

[1001] For example, create a prompt like, "Emotions are excited. I'm asking about reviews of the 2021 film Inception. I need your input and support."

[1002] Input: Search query and sentiment data

[1003] Data processing: Prompt generation algorithm generates appropriate prompts

[1004] Output: Generated prompt

[1005] Step 5:

[1006] The server sends the generated prompt to the OpenAI API and requests a response.

[1007] Specifically, the server constructs an API request and sends it, including the prompt, to OpenAI's endpoint.

[1008] Input: Generated prompt

[1009] Output: The prompt sent to the OpenAI API

[1010] Step 6:

[1011] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1012] For example, it might generate a response along the lines of, "Inception is a film directed by Christopher Nolan that has received very high praise overall..."

[1013] Input: Prompt sent

[1014] Data Computation: Response Generation with OpenAI's Natural Language Generation Model

[1015] Output: The generated response

[1016] Step 7:

[1017] The server sorts the returned responses, adjusts the tone and content to match the user's emotions, and then sends them back to the user's device.

[1018] For example, adjust the tone of your response to be more energetic based on the emotion "excited."

[1019] Input: The generated response

[1020] Data manipulation: Emotion engine adjusts tone and content of responses

[1021] Output: The adjusted response

[1022] Step 8:

[1023] The terminal receives the adjusted response and displays it to the user.

[1024] Specifically, the response content is displayed on the application's results screen.

[1025] Input: Adjusted response

[1026] Output: The response that is displayed to the user

[1027] Through these steps, users can quickly and efficiently search for trailers and reviews of previously distributed content, and obtain information that is appropriately tailored based on sentiment.

[1028] 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.

[1029] 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.

[1030] 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.

[1031] [Third embodiment]

[1032] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1033] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[1034] 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).

[1035] 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.

[1036] 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.

[1037] 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).

[1038] 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.

[1039] 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.

[1040] 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.

[1041] 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.

[1042] 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.

[1043] 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."

[1044] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and responding to users immediately regarding their content. The system includes a terminal through which a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[1045] Search function embodiment

[1046] User enters and submits a search query

[1047] The user inputs a search query using the device interface. For example, they input the keyword "Business Announcement Contents for 2023" and press the search button. This operation sends the search query from the device to the server.

[1048] Server receives and processes search queries

[1049] The server receives a search query sent from the user's device, analyzes the received query, and generates and sends a prompt to a natural language generation model (e.g., OpenAI's API) based on the query's content. The natural language generation model generates a response related to the query and sends it back to the server.

[1050] Response generation and forwarding

[1051] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's device is based on the user's query, providing information quickly and accurately.

[1052] Detailed inquiry function embodiment

[1053] User enters and submits detailed request

[1054] To find out more about a specific business announcement, the user inputs its ID and sends a details request. For example, if the user wants to know the details of the business announcement with ID 1, the user inputs the ID and presses the details button. This operation sends a details request from the terminal to the server.

[1055] Server receives and processes detailed requests

[1056] The server receives a detailed request sent from the user's device. It analyzes the ID included in the request and retrieves the relevant sales notice information from the database. It generates and sends a prompt to the natural language generation model based on the relevant information. The natural language generation model generates a detailed response and sends it back to the server.

[1057] Response generation and forwarding

[1058] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's terminal is based on the detailed request, providing information quickly and accurately.

[1059] Specific examples

[1060] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[1061] {

[1062] "query": "2023 Business Announcement"

[1063] }

[1064] The server receives this query and generates and sends prompts to the natural language generation model, which then generates a response and returns it to the user.

[1065] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[1066] GET / detail / 1

[1067] The server receives this request, retrieves the relevant information from a database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user.

[1068] This system allows users to quickly and efficiently search for previously distributed business announcements and get immediate responses to their detailed information, significantly improving the efficiency of managing and using business announcements.

[1069] The processing flow will be explained below.

[1070] Search function processing steps

[1071] Step 1:

[1072] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[1073] Step 2:

[1074] The user presses the search button to send a search request including the query from the terminal to the server.

[1075] Step 3:

[1076] The server receives the search request, parses the query string, and generates a prompt to send to the OpenAI API.

[1077] Step 4:

[1078] The server sends the generated prompt to the OpenAI API and requests a response.

[1079] Step 5:

[1080] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1081] Step 6:

[1082] The server compiles the response received from the OpenAI API and sends it back to the user.

[1083] Step 7:

[1084] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[1085] Processing steps for detailed inquiry function

[1086] Step 1:

[1087] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[1088] Step 2:

[1089] The user presses the details button, and a details request including the ID is sent from the terminal to the server.

[1090] Step 3:

[1091] The server receives the detailed request, extracts the ID included in the request, and searches the database based on that ID to retrieve the relevant business announcement.

[1092] Step 4:

[1093] The server uses the acquired business notice to generate a prompt to send to the OpenAI API.

[1094] Step 5:

[1095] The server sends the generated prompt to the OpenAI API and requests a response.

[1096] Step 6:

[1097] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1098] Step 7:

[1099] The server compiles the response received from the OpenAI API and sends it back to the user.

[1100] Step 8:

[1101] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[1102] Example 1

[1103] 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."

[1104] Conventional information retrieval systems have difficulty generating fast and accurate responses to specific keywords or detailed requests. Furthermore, more flexible and efficient information processing methods are needed to respond to diverse user requests. In particular, there is a need for systems that can accurately search past information and respond to users immediately, but there are not enough systems available that can meet these demands.

[1105] 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.

[1106] In this invention, the server includes means for receiving input from a user, means for dynamically generating prompts for a generative model based on the input, and means for obtaining responses from the generative model using the dynamically generated prompts, thereby enabling the server to respond to a variety of user requests and provide responses quickly and accurately.

[1107] "Means for receiving input from a user" refers to the functionality or interface that allows a user to input information and for the system to receive it.

[1108] "Means for dynamically generating prompts for a generative model" refers to a function that automatically generates questions (prompts) that enable the generative model to return an appropriate response based on input information.

[1109] The "means for obtaining a response from a generative model" refers to a function for receiving a response returned from a generative model based on a generated prompt.

[1110] "Means for returning to the user" refers to a function within the system that sends information to the user's terminal in order to display the obtained response to the user.

[1111] The term "means for receiving a request including a specific identifier" refers to a function whereby a user provides an identifier for specifying specific information and the system receives the identifier.

[1112] "Means for obtaining corresponding content from information storage means" refers to the function of searching and retrieving information based on a specified identifier from a database or storage within the system.

[1113] The "means for dynamically generating a prompt based on specific conditions" refers to a function for dynamically generating an appropriate prompt in response to a user's input or request.

[1114] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and instantly responding to the user about their content. The system includes a terminal through which the user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[1115] Search function embodiment

[1116] User enters a search query

[1117] The user uses their own device to input the keywords they want to search for into the interface. For example, they can input "2023 business announcements" and press the search button to confirm the query input.

[1118] Sending queries by terminal

[1119] After pressing the search button, the terminal sends the entered query to the server. The data sent is as follows:

[1120] {

[1121] "query": "2023 Business Announcement"

[1122] }

[1123] Server receives and processes search queries

[1124] The server receives the search query sent from the device, analyzes the query, and dynamically generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[1125] "Please tell me what the business announcement for 2023 will be."

[1126] Prompting a Natural Language Generation Model

[1127] The server sends the generated prompts to a natural language generation model (e.g., a popular generative AI API), which generates a response based on the prompts and sends the result back to the server.

[1128] Receiving and formatting the response by the server

[1129] The server receives the response sent back by the generative model and formats it appropriately, which may include converting it to HTML or JSON format.

[1130] Terminal display of responses

[1131] The terminal receives the formatted response from the server and displays it to the user, allowing the user to quickly see the results of their query.

[1132] Detailed inquiry function embodiment

[1133] User enters detailed request

[1134] If the user wants to know the details of a specific business announcement, he or she inputs its ID and presses the Details button. For example, the user inputs "Details of the business announcement with ID 1."

[1135] Sending a details request by terminal

[1136] After pressing the Details button, the device will send the input request to the server, which will have the following format:

[1137] GET / detail / 1

[1138] Server receives and processes detailed requests

[1139] The server receives the detailed request sent from the device, analyzes the ID included in the request, and searches the database. Based on the information obtained from the database, it generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[1140] "Please tell me the details of the business announcement with ID 1."

[1141] Prompting a Natural Language Generation Model

[1142] The server sends the generated prompt to a natural language generation model to generate a detailed response, which is then sent back to the server.

[1143] Server response formatting and forwarding

[1144] The server receives the response from the generative model, formats it in an appropriate format, and sends it back to the device. Examples of returned data are as follows:

[1145] {

[1146] "response": "The details of the sales announcement with ID 1 are as follows..."

[1147] }

[1148] Terminal display of responses

[1149] The terminal displays the response received from the server on the screen, allowing the user to quickly check the correct information for the detailed request.

[1150] The present invention allows users to quickly and efficiently search past business announcements and get immediate responses to detailed information, greatly improving the efficiency of business announcement management and utilization.

[1151] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1152] Search function embodiment

[1153] Step 1:

[1154] User enters a search query

[1155] The user inputs the keywords they want to search for using the interface of their device. For example, they input "Business Notices for 2023" and press the search button. This operation confirms the query they entered.

[1156] Input: "Business Notice for 2023"

[1157] Output: The query is typed into the terminal

[1158] Step 2:

[1159] Sending queries by terminal

[1160] When the search button is pressed, the device sends the entered query to the server. Specifically, the query is packaged in JSON format and sent to the server as an HTTP request.

[1161] Input: "Business Notice for 2023"

[1162] Output: The query is sent to the server as an HTTP request.

[1163] Step 3:

[1164] Server receives search query

[1165] The server receives the query sent from the device, parses the received data in JSON format, and extracts the search keywords.

[1166] Input: Query from terminal

[1167] Output: Parsed search keywords

[1168] Step 4:

[1169] The server parses the query and generates prompts

[1170] The server analyzes the received query and dynamically generates a prompt based on its content, which becomes the question sent to the generative AI model.

[1171] How it works: Builds a prompt from search keywords.

[1172] Input: Parsed search keywords

[1173] Output: Generated prompt: "What is the content of your 2023 sales announcement?"

[1174] Step 5:

[1175] Prompting a Natural Language Generation Model

[1176] The server sends the generated prompt to a generative AI model (e.g., a general generative AI API).

[1177] What it does: Sends the prompt to the generative AI model as an HTTP request.

[1178] Input: Generated prompt

[1179] Output: A request to the NLG model

[1180] Step 6:

[1181] Receiving responses from generative AI models

[1182] The server receives a response from the generative AI model, which is a natural language answer generated based on the prompt.

[1183] Behavior: Receives the response as an HTTP response and parses it.

[1184] Input: The response returned by the generative AI model

[1185] Output: Parsed response content

[1186] Step 7:

[1187] Server response formatting

[1188] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display to the user.

[1189] Behavior: Formats the response content.

[1190] Input: Response from the generative AI model

[1191] Output: Formatted response data

[1192] Step 8:

[1193] Response from the server to the device

[1194] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[1195] Behavior: Sends response data as an HTTP response.

[1196] Input: Formatted response data

[1197] Output: Send response data to the terminal

[1198] Step 9:

[1199] Terminal display of responses

[1200] The terminal displays the response received from the server to the user, allowing the user to check the search results.

[1201] Behavior: Displays the response data in the interface.

[1202] Input: Response data from the server

[1203] Output: What is displayed to the user

[1204] Detailed inquiry function embodiment

[1205] Step 1:

[1206] User enters detailed request

[1207] If the user wants to know the details of a specific business announcement, he or she enters its ID and presses the Details button. For example, the user might enter "Details of the business announcement with ID 1." This operation confirms the request for details.

[1208] Input: "Details of business announcement with ID 1"

[1209] Output: Detailed request typed into terminal

[1210] Step 2:

[1211] Sending a details request by terminal

[1212] When the details button is pressed, the terminal transmits the input request to the server. Specifically, the request is transmitted to the server as an HTTP GET request.

[1213] Input: "Details of business announcement with ID 1"

[1214] Output: The detailed request is sent to the server as an HTTP GET request.

[1215] Step 3:

[1216] Server receives detailed request

[1217] The server receives a detailed request containing the ID sent by the device, which is then parsed to extract the ID.

[1218] Input: Detailed request from terminal

[1219] Output: Parsed ID

[1220] Step 4:

[1221] Server parses the request and performs a database lookup

[1222] The server searches the database based on the ID included in the detailed request and obtains the relevant business notice information.

[1223] What it does: Searches the database based on the ID and retrieves the relevant information.

[1224] Input: Parsed ID

[1225] Output: Business announcement information retrieved from the database

[1226] Step 5:

[1227] Server-generated prompts

[1228] Based on the information obtained, the server generates a prompt that is sent to the generative AI model, asking for more information.

[1229] How it works: Construct a prompt from the information obtained.

[1230] Input: Business announcement information retrieved from the database

[1231] Output: Generated prompt: "Please tell me the details of the sales announcement with ID 1."

[1232] Step 6:

[1233] Prompting a Natural Language Generation Model

[1234] The server sends the generated prompts to the generative AI model.

[1235] What it does: Sends the prompt to the generative AI model as an HTTP request.

[1236] Input: Generated prompt

[1237] Output: A request to the NLG model

[1238] Step 7:

[1239] Receiving responses from generative AI models

[1240] The server receives a response from the generative AI model, which is a detailed answer generated based on the prompt.

[1241] Behavior: Receives the response as an HTTP response and parses it.

[1242] Input: The response returned by the generative AI model

[1243] Output: Parsed response content

[1244] Step 8:

[1245] Server response formatting

[1246] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display on the device.

[1247] Behavior: Formats the response content.

[1248] Input: Response from the generative AI model

[1249] Output: Formatted response data

[1250] Step 9:

[1251] Response from the server to the device

[1252] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[1253] Behavior: Sends response data as an HTTP response.

[1254] Input: Formatted response data

[1255] Output: Send response data to the terminal

[1256] Step 10:

[1257] Terminal display of responses

[1258] The terminal displays the response received from the server to the user, allowing the user to quickly check the information for the detailed request.

[1259] Behavior: Displays the response data in the interface.

[1260] Input: Response data from the server

[1261] Output: What is displayed to the user

[1262] (Application example 1)

[1263] 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."

[1264] On current online shopping sites, it can take a significant amount of time for users to quickly search and obtain information about past campaigns or detailed product information. Furthermore, there is a lack of systems that accurately provide the information users are looking for, resulting in a poor user experience. The present invention aims to solve these problems and provide a system that allows users to efficiently and quickly obtain the information they need.

[1265] 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.

[1266] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, means for providing a database for efficiently managing information searched in the past, and means for inputting a search query through an interface on a smartphone, thereby enabling users to quickly and accurately search and obtain information on past campaigns and detailed product information.

[1267] A "search query" is a keyword or phrase that a user enters to locate specific information.

[1268] A "natural language generation model" is an artificial intelligence technique that generates relevant sentences and responses based on input text.

[1269] A "response" is the information or answer provided by a system in response to a user's search query.

[1270] A "database" is a collection of data that stores previously searched information and detailed information in an organized manner so that it can be efficiently searched and retrieved later.

[1271] A "smartphone" is a type of mobile phone that has internet connectivity and a variety of applications, allowing users to perform various searches and obtain information.

[1272] An "interface" is a screen or operating means that a user uses to interact with a system.

[1273] A "detail request" is a request including a specific ID that a user enters when they want to know more about a specific piece of information.

[1274] The present invention is a system for quickly retrieving information and detailed information that a user has previously searched for. This system is operated via an interface installed on the user's smartphone. The main components of the system include the following:

[1275] System Configuration

[1276] Hardware elements

[1277] 1. Smartphone: A device where users enter search queries and detailed requests. It communicates with the server through an interface.

[1278] 2. Server: Receives queries and requests, retrieves information from the database, and generates and sends prompts to the natural language generation model.

[1279] Software Elements

[1280] 1. Interface: A smartphone application built using React Native that provides an interface for users to enter search queries and detailed requests.

[1281] 2. Server application: Built using Node.js and the Express framework, it handles user queries and requests, generates prompts, and sends them to the natural language generation model.

[1282] 3. Database Management System: MongoDB is used to store and manage past search queries and detailed information.

[1283] 4. Natural language generation models: Generative AI models provided as APIs (e.g., OpenAI's API) are used to generate relevant responses based on input queries.

[1284] Processing flow

[1285] Query Processing

[1286] A user uses the smartphone interface to enter a search query, for example, "Black Friday deals 2022," and presses the search button. This query is then sent from the smartphone to the server.

[1287] The server analyzes the received query and generates and sends the following prompt to the natural language generation model:

[1288] Search past sales announcements: Black Friday 2022 campaigns

[1289] The response returned by the generative AI model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain past campaign information.

[1290] Detailed Request Processing

[1291] If a user wants to know more information about a specific product or campaign, they enter its ID into the smartphone interface. For example, if they want to know the details of the sales announcement with ID 1, they enter the ID and press the details button. The details request is sent from the smartphone to the server.

[1292] The server receives the request, retrieves the relevant information from the database, and then generates and sends the following prompt to the natural language generation model:

[1293] Detailed information about business notice: [Content of information obtained]

[1294] The detailed response returned by the natural language generation model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain the required details.

[1295] This invention enables users to quickly and efficiently search and obtain information about past campaigns and detailed product information. The collaboration between the database and the natural language generation model improves the user experience and enables efficient information acquisition.

[1296] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1297] Step 1:

[1298] A user uses a smartphone interface to enter a search query and presses the submit button, where the input is a keyword such as "Black Friday deals 2022." As an output, the smartphone sends this query to a server.

[1299] Step 2:

[1300] The server receives the search query sent by the user. The input is the user's search query, and the server analyzes the received query as output. As a result of the analysis, it generates a prompt such as "Search past sales announcements: Black Friday campaigns 2022."

[1301] Step 3:

[1302] The server sends the generated prompt to the natural language generation model. The input is the generated prompt, and the output is a response from the natural language generation model. The specific operation is to send the prompt via an API request and obtain the response as an API response.

[1303] Step 4:

[1304] The server analyzes the response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it processes the response text to format it for the user.

[1305] Step 5:

[1306] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[1307] Step 6:

[1308] The user inputs a detailed request including a specific ID into the smartphone interface and presses the send button. The input is "Details of the business announcement with ID 1." This request is sent from the smartphone to the server as an output.

[1309] Step 7:

[1310] The server receives the detailed request sent by the user. The input is the detailed request from the user, and the received request is analyzed as the output. As a result of the analysis, the information is retrieved from the database using the corresponding ID.

[1311] Step 8:

[1312] The server generates a prompt based on the information retrieved from the database and sends it to the natural language generation model. The input is the information retrieved from the database, and the generated prompt is sent as output. Specifically, it creates a prompt called "Sales Notice Details: [Content of retrieved information]" and sends it via an API request.

[1313] Step 9:

[1314] The server receives the detailed response returned by the NLG model. The input is the generated prompt and the output is the response from the NLG model. The action is to get the response as an API response.

[1315] Step 10:

[1316] The server parses the detailed response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it performs a process to format the response text.

[1317] Step 11:

[1318] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[1319] Step 12:

[1320] Detailed information will be displayed on the user's smartphone, allowing the user to quickly and accurately obtain the information they need.

[1321] 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.

[1322] The present invention provides a system for quickly and efficiently searching for previously distributed sales announcements and instantly responding to users about their content. The system includes a terminal where a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model. The system also combines an emotion engine that recognizes the user's emotions and adjusts the response based on those emotions to provide more personalized responses.

[1323] Search function embodiment

[1324] User enters and submits a search query

[1325] The user inputs a search query using the device interface, for example, "sales announcements for 2023." The device then identifies the emotion from the user's input and generates emotion data using an emotion engine.

[1326] Server receives and processes search queries

[1327] The server receives search queries and emotion data sent from the user's device, analyzes the received queries and emotion data, and generates prompts to send to the OpenAI API, which include adjustments based on the user's emotion.

[1328] Response generation and forwarding

[1329] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[1330] Detailed inquiry function embodiment

[1331] User enters and submits detailed request

[1332] The user inputs the ID of a specific business announcement and sends a request for details. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[1333] Server receives and processes detailed requests

[1334] The server receives the detailed request and emotion data sent from the user's device, extracts the ID included in the request, and retrieves the corresponding sales announcement information from the database. Based on the retrieved information, it generates a prompt to send to the OpenAI API. This prompt also includes adjustments based on the user's emotion.

[1335] Response generation and forwarding

[1336] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[1337] Specific examples

[1338] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[1339] {

[1340] "query": "Business Notice 2023",

[1341] "emotion": "neutral"

[1342] }

[1343] The server receives the query and emotion data, generates prompts for the natural language generation model, and sends the resulting responses back to the user. The emotion engine is tuned to generate responses with a "neutral" tone.

[1344] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[1345] GET / detail / 1

[1346] {

[1347] "emotion": "curious"

[1348] }

[1349] The server receives the request and emotion data, retrieves the relevant information from the database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user. The emotion engine is tuned to generate responses with a "curious" tone.

[1350] The present invention enables a user to quickly and efficiently search for business announcements distributed in the past and to receive an immediate response regarding the detailed contents thereof, and furthermore, by providing a response adapted to the user's emotions, a more satisfying service can be realized.

[1351] The processing flow will be explained below.

[1352] Search function processing steps

[1353] Step 1:

[1354] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[1355] Step 2:

[1356] When a user presses the search button, a search request including a query is sent from the terminal to the server. At this time, the terminal identifies the emotion from the user's input and generates emotion data using an emotion engine.

[1357] Step 3:

[1358] The server receives the search request and sentiment data, parses the query string and sentiment data, and generates a prompt to send to the natural language generation model.

[1359] Step 4:

[1360] The server sends the generated prompt to a natural language generation model for a response, where the prompt includes a modulation that reflects the user's sentiment.

[1361] Step 5:

[1362] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[1363] Step 6:

[1364] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[1365] Step 7:

[1366] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[1367] Processing steps for detailed inquiry function

[1368] Step 1:

[1369] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[1370] Step 2:

[1371] When the user presses the details button, a detailed request including the ID is sent from the device to the server. At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[1372] Step 3:

[1373] The server receives the detailed request and emotion data. Specifically, it extracts the ID included in the request and searches the database based on that ID to retrieve the relevant sales announcement.

[1374] Step 4:

[1375] The server uses the acquired sales pitch to generate prompts for transmission to a natural language generation model, which also include adjustments based on the user's sentiment.

[1376] Step 5:

[1377] The server sends the generated prompt to a natural language generation model and requests a response.

[1378] Step 6:

[1379] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[1380] Step 7:

[1381] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[1382] Step 8:

[1383] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[1384] Example 2

[1385] 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."

[1386] Conventional systems have been unable to consider user emotions when quickly and accurately responding to users' search queries and detailed requests. This can result in mechanical, unpersonalized responses, potentially reducing user satisfaction. Furthermore, prompts are generated statically, making it difficult to provide intuitive responses. Therefore, a system that takes user emotions into account and generates more personalized responses quickly and efficiently is needed.

[1387] 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.

[1388] In this invention, the server includes a means for receiving a search query from a user, a means for generating a prompt sentence using the search query, and a means for obtaining a response from a generative AI model using the prompt sentence. This makes it possible to dynamically generate a prompt sentence in response to the user's search query and quickly and efficiently obtain a response from the generative AI model. Furthermore, by including a means for identifying the user's emotion and generating emotion data, and a means for adjusting the tone or content of the response based on the user's emotion data, it is possible to provide a personalized response based on the user's emotion.

[1389] "User" means a person who utilizes the system to enter a search query or detailed request.

[1390] A "search query" is text that a user enters into a system to search for information.

[1391] A "prompt" is a text instruction generated from a search query or detail request and sent to a generative AI model.

[1392] A "generative AI model" is an artificial intelligence that automatically generates a response based on an input prompt.

[1393] "Emotion data" is data indicating a user's emotion identified from the user's search query or detail request.

[1394] The "emotion engine" is the part of the system that identifies emotions from user input and generates emotion data.

[1395] A "response" is information that a generative AI model generates based on a prompt and sends back to the user.

[1396] A "detail request" is a user request for more information based on specific identifying information.

[1397] "Identification information" refers to information that uniquely identifies specific information within a database.

[1398] A "database" is a collection of information used to retrieve information in response to a user's request.

[1399] "Tone" refers to the emotional tone or mood of the generated response, which is adjusted based on the user's emotions.

[1400] The present invention is a system that quickly and efficiently generates responses based on search queries and detailed requests entered by users. The system includes a terminal where users enter queries, a server for receiving and processing queries, and a function for generating responses using a generative AI model. It also combines an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions to provide more personalized responses.

[1401] Processing search queries

[1402] User enters and submits a search query

[1403] The user inputs a search query using the device interface. For example, they input "Business announcements for 2023." The device then identifies emotions from the user's input and generates emotion data using an emotion engine. The emotion data and the search query are then sent to the server.

[1404] Server receives and processes search queries

[1405] The server receives the search query and emotion data sent from the user's device. It analyzes the received query and emotion data and generates a prompt sentence to send to the generative AI model. This prompt includes adjustments based on the user's emotion. For example, it might be in the form of "Please provide detailed information about the sales announcements for 2023 in a neutral tone." The server then sends this prompt sentence to the generative AI model.

[1406] Response generation and forwarding

[1407] The generative AI model generates a response based on the prompt received from the server and sends it back to the server, which then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user.

[1408] Detail Request Processing

[1409] User enters and submits detailed request

[1410] The user inputs and sends a detailed request including specific identification information. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies emotions from the user's input and generates emotion data using an emotion engine. This emotion data and the detailed request are then sent to the server.

[1411] Server receives and processes detailed requests

[1412] The server receives the detailed request and emotion data sent from the user's device. It extracts the identification information included in the request and retrieves the corresponding sales announcement information from the database. For example, it retrieves information such as "ID: 1, Title: New Product Announcement, Content: Details about New Products for 2023." Based on this, it generates a prompt to send to the generative AI model. This prompt also includes adjustments based on the user's emotion. The server then sends this prompt to the generative AI model.

[1413] Response generation and forwarding

[1414] The generative AI model generates a response based on the prompt received from the server and sends it back to the server. The server then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user. For example, a response such as "The sales announcement with ID 1 is about a new product announcement. Details of the new product to be introduced in 2023 are..." may be generated.

[1415] In this way, the present invention can quickly and efficiently generate personalized responses that take into account the user's emotions. Specific examples of prompt sentences are shown below.

[1416] Prompt Sentence Examples

[1417] 1. Prompt based on search query:

[1418] "Please provide detailed information in a neutral tone regarding the 2023 business announcement."

[1419] 2. Prompt sentence based on detailed request:

[1420] "Please explain the details of the business announcement with ID 1 in an interesting tone."

[1421] In this way, responses to user inputs are provided in appropriate tone and content.

[1422] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1423] Step 1: User enters and submits a search query

[1424] The user inputs a search query using the device interface. For example, the user inputs "sales announcements for 2023." The device then generates emotion data using an emotion engine to identify emotions from the user's input. The device then sends the search query together with this emotion data to the server.

[1425] Input: Search query ("Sales announcements for 2023"), emotion data ("neutral")

[1426] Output: Query data sent to the server

[1427] Specific operation: The user enters "Sales announcement for 2023" into the device and clicks the send button. The emotion engine analyzes this input and generates emotion data called "neutral." The device then sends the query data to the server.

[1428] Step 2: The server receives the search query and generates a prompt.

[1429] The server receives the search query and emotion data sent from the user's device. After receiving the query, it analyzes the query content and generates a prompt to send to the generative AI model. This prompt includes adjustments based on the user's emotion, such as "Please provide detailed information in a neutral tone about the sales announcements for 2023."

[1430] Input: Query data received by the server (search query, sentiment data)

[1431] Output: A prompt to send to the generative AI model

[1432] Specific operation: The server analyzes the received query data and generates a prompt, such as "Please provide detailed information about the sales announcements for 2023 in a neutral tone," and prepares to send the prompt to the generation AI model.

[1433] Step 3: The server sends a prompt to the generative AI model

[1434] The server sends the generated prompt to the generative AI model, which then generates a response based on the prompt.

[1435] Input: A prompt to send to the generative AI model

[1436] Output: The response sent back from the generative AI model

[1437] Specific operation: The server generates a prompt and sends it to the generative AI model, which receives it and generates a response.

[1438] Step 4: Response generation and organization and coordination by the server

[1439] The server receives the responses sent back by the generative AI model and organizes them, adjusting the tone and content based on the user's emotions, sometimes again using the emotion engine.

[1440] Input: The response returned by the generative AI model

[1441] Output: The adjusted response that is sent back to the user

[1442] What it does: The response sent back to the server is processed by a natural language analysis module, and the emotion engine is used to adjust the tone and content, so the response might be formatted as something like, "Here's what our 2023 sales announcement will include..."

[1443] Step 5: Server sends back response

[1444] The server sends the adjusted response back to the user's terminal.

[1445] Input: Adjusted response

[1446] Output: The response that is displayed to the user

[1447] Specific operation: The server sends the adjusted response to the user's terminal, and the user confirms the response through the terminal interface.

[1448] (Application example 2)

[1449] 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."

[1450] Conventionally, when users search for trailers or reviews of previously distributed content (movies or TV dramas), they lack a means to quickly and efficiently obtain relevant information. Furthermore, there is a lack of systems that can provide personalized responses based on the user's emotions, creating a demand for an improved user experience. In particular, it is difficult to generate responses with an appropriate tone based on the user's emotions.

[1451] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1452] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, and means for recognizing the user's emotion using emotion recognition means and adjusting the response based on the emotion, thereby enabling the user to quickly and efficiently search for trailers and reviews of previously distributed content (movies and dramas) and further provide personalized responses according to the user's emotion.

[1453] A "search query" refers to a word or phrase that a user enters to obtain specific information.

[1454] A "natural language generation model" refers to an algorithm or system that is trained on a huge amount of text data to generate natural-sounding sentences similar to those written by humans.

[1455] "Response" refers to the content generated and returned by the system in response to a user's search query.

[1456] "Emotion recognition means" refers to technology or systems that identify emotions from user input, facial expressions, voice, etc., and extract them as specific emotional data.

[1457] A "prompt" refers to an input sentence that instructs a natural language generation model what kind of response to generate.

[1458] A "detail request" refers to a request from a user that includes a specific identifier and asks for a more detailed response about a particular piece of information.

[1459] "Database" refers to a system or location for organizing, storing, and managing information.

[1460] "Identifier" refers to a string of symbols, numbers, or letters that uniquely identifies specific information.

[1461] This invention is a system that allows users to quickly and efficiently search for trailers and reviews of previously distributed content (movies and TV dramas) and receive instant responses regarding the details. It also includes various means for providing personalized responses according to the user's emotions. Specific methods for implementing this invention are described below.

[1462] System Configuration

[1463] Hardware

[1464] Device: Smartphone

[1465] Server: High-performance cloud server

[1466] Emotion Recognition Device: A module for recognizing emotions from voice and text data

[1467] software

[1468] Natural Language Generation model: OpenAI API (text-davinci-003)

[1469] Emotion Recognition Software: Emotion Recognition Model (e.g., Emotion AI SDK)

[1470] Program processing

[1471] Query Reception and Emotion Recognition

[1472] A user inputs a search query using a smartphone interface, for example, "Reviews of the 2021 movie 'Inception.'" The device receives the query and uses an emotion recognition device to identify the user's emotion from the input query.

[1473] Prompt Generation

[1474] The server generates a prompt to send to the natural language generation model based on the received search query and sentiment data. An example of a generated prompt would be, "I'm excited. I'm asking about a review of the 2021 film 'Inception.' I'd appreciate your information and support."

[1475] Generating and returning a response

[1476] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the prompt and sends it back to the server. The server then organizes the response, adjusts the tone and content to match the user's emotions, and sends it back to the user's smartphone.

[1477] Specific examples

[1478] For example, if a user searches for "2021 movie 'Inception' reviews," they'll get the following:

[1479] A user types "Review of the 2021 film Inception" into their device.

[1480] The terminal receives this query and obtains the emotion data "excited" from the emotion recognition device.

[1481] The server generates a prompt based on the query and emotion data and sends it to the OpenAI API in the form, "Emotion is excited. I'm asking about a review of the 2021 movie 'Inception'. Please provide information and support."

[1482] The OpenAI API generates a response and sends it back to the server.

[1483] The server sorts the responses, adjusts the tone appropriately based on the emotion of "excitement," and sends them back to the user's terminal.

[1484] The system allows users to quickly learn about previously distributed content and receive personalized, emotion-based responses.

[1485] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1486] Step 1:

[1487] A user enters a search query into a device.

[1488] Specifically, a user opens the application on their smartphone and types "reviews of the 2021 film 'Inception'" into the search bar.

[1489] Input: Search query (e.g., "2021 movie 'Inception' reviews")

[1490] Output: The search query entered by the user

[1491] Step 2:

[1492] The terminal receives the search query and recognizes the user's emotion using the emotion recognition means.

[1493] For example, the device analyzes the voice and word choice from the input text to identify the emotion "excitement."

[1494] Input: Search query from user

[1495] Data processing: Emotion recognition software analyzes text and extracts emotional data

[1496] Output: Search query and sentiment data (e.g., "excited")

[1497] Step 3:

[1498] The device sends the search query and emotion data to the server.

[1499] Specifically, the device generates a data packet to send the search query "2021 movie 'Inception' reviews" and the emotion "excited" to the server.

[1500] Input: Search query and sentiment data

[1501] Output: Data packet sent to the server

[1502] Step 4:

[1503] The server generates prompts based on the received search query and sentiment data.

[1504] For example, create a prompt like, "Emotions are excited. I'm asking about reviews of the 2021 film Inception. I need your input and support."

[1505] Input: Search query and sentiment data

[1506] Data processing: Prompt generation algorithm generates appropriate prompts

[1507] Output: Generated prompt

[1508] Step 5:

[1509] The server sends the generated prompt to the OpenAI API and requests a response.

[1510] Specifically, the server constructs an API request and sends it, including the prompt, to OpenAI's endpoint.

[1511] Input: Generated prompt

[1512] Output: The prompt sent to the OpenAI API

[1513] Step 6:

[1514] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1515] For example, it might generate a response along the lines of, "Inception is a film directed by Christopher Nolan that has received very high praise overall..."

[1516] Input: Prompt sent

[1517] Data Computation: Response Generation with OpenAI's Natural Language Generation Model

[1518] Output: The generated response

[1519] Step 7:

[1520] The server sorts the returned responses, adjusts the tone and content to match the user's emotions, and then sends them back to the user's device.

[1521] For example, adjust the tone of your response to be more energetic based on the emotion "excited."

[1522] Input: The generated response

[1523] Data manipulation: Emotion engine adjusts tone and content of responses

[1524] Output: The adjusted response

[1525] Step 8:

[1526] The terminal receives the adjusted response and displays it to the user.

[1527] Specifically, the response content is displayed on the application's results screen.

[1528] Input: Adjusted response

[1529] Output: The response that is displayed to the user

[1530] Through these steps, users can quickly and efficiently search for trailers and reviews of previously distributed content, and obtain information that is appropriately tailored based on sentiment.

[1531] 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.

[1532] 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.

[1533] 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.

[1534] [Fourth embodiment]

[1535] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1536] 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.

[1537] 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).

[1538] 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.

[1539] 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.

[1540] 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).

[1541] 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.

[1542] 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.

[1543] 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.

[1544] 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.

[1545] 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.

[1546] 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.

[1547] 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."

[1548] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and responding to users immediately regarding their content. The system includes a terminal through which a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[1549] Search function embodiment

[1550] User enters and submits a search query

[1551] The user inputs a search query using the device interface. For example, they input the keyword "Business Announcement Contents for 2023" and press the search button. This operation sends the search query from the device to the server.

[1552] Server receives and processes search queries

[1553] The server receives a search query sent from the user's device, analyzes the received query, and generates and sends a prompt to a natural language generation model (e.g., OpenAI's API) based on the query's content. The natural language generation model generates a response related to the query and sends it back to the server.

[1554] Response generation and forwarding

[1555] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's device is based on the user's query, providing information quickly and accurately.

[1556] Detailed inquiry function embodiment

[1557] User enters and submits detailed request

[1558] To find out more about a specific business announcement, the user inputs its ID and sends a details request. For example, if the user wants to know the details of the business announcement with ID 1, the user inputs the ID and presses the details button. This operation sends a details request from the terminal to the server.

[1559] Server receives and processes detailed requests

[1560] The server receives a detailed request sent from the user's device. It analyzes the ID included in the request and retrieves the relevant sales notice information from the database. It generates and sends a prompt to the natural language generation model based on the relevant information. The natural language generation model generates a detailed response and sends it back to the server.

[1561] Response generation and forwarding

[1562] The server receives the response from the natural language generation model and returns it to the user in an appropriate format. The response displayed on the user's terminal is based on the detailed request, providing information quickly and accurately.

[1563] Specific examples

[1564] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[1565] {

[1566] "query": "2023 Business Announcement"

[1567] }

[1568] The server receives this query and generates and sends prompts to the natural language generation model, which then generates a response and returns it to the user.

[1569] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[1570] GET / detail / 1

[1571] The server receives this request, retrieves the relevant information from a database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user.

[1572] This system allows users to quickly and efficiently search for previously distributed business announcements and get immediate responses to their detailed information, significantly improving the efficiency of managing and using business announcements.

[1573] The processing flow will be explained below.

[1574] Search function processing steps

[1575] Step 1:

[1576] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[1577] Step 2:

[1578] The user presses the search button to send a search request including the query from the terminal to the server.

[1579] Step 3:

[1580] The server receives the search request, parses the query string, and generates a prompt to send to the OpenAI API.

[1581] Step 4:

[1582] The server sends the generated prompt to the OpenAI API and requests a response.

[1583] Step 5:

[1584] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1585] Step 6:

[1586] The server compiles the response received from the OpenAI API and sends it back to the user.

[1587] Step 7:

[1588] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[1589] Processing steps for detailed inquiry function

[1590] Step 1:

[1591] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[1592] Step 2:

[1593] The user presses the details button, and a details request including the ID is sent from the terminal to the server.

[1594] Step 3:

[1595] The server receives the detailed request, extracts the ID included in the request, and searches the database based on that ID to retrieve the relevant business announcement.

[1596] Step 4:

[1597] The server uses the acquired business notice to generate a prompt to send to the OpenAI API.

[1598] Step 5:

[1599] The server sends the generated prompt to the OpenAI API and requests a response.

[1600] Step 6:

[1601] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[1602] Step 7:

[1603] The server compiles the response received from the OpenAI API and sends it back to the user.

[1604] Step 8:

[1605] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[1606] Example 1

[1607] 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."

[1608] Conventional information retrieval systems have difficulty generating fast and accurate responses to specific keywords or detailed requests. Furthermore, more flexible and efficient information processing methods are needed to respond to diverse user requests. In particular, there is a need for systems that can accurately search past information and respond to users immediately, but there are not enough systems available that can meet these demands.

[1609] 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.

[1610] In this invention, the server includes means for receiving input from a user, means for dynamically generating prompts for a generative model based on the input, and means for obtaining responses from the generative model using the dynamically generated prompts, thereby enabling the server to respond to a variety of user requests and provide responses quickly and accurately.

[1611] "Means for receiving input from a user" refers to the functionality or interface that allows a user to input information and for the system to receive it.

[1612] "Means for dynamically generating prompts for a generative model" refers to a function that automatically generates questions (prompts) that enable the generative model to return an appropriate response based on input information.

[1613] The "means for obtaining a response from a generative model" refers to a function for receiving a response returned from a generative model based on a generated prompt.

[1614] "Means for returning to the user" refers to a function within the system that sends information to the user's terminal in order to display the obtained response to the user.

[1615] The term "means for receiving a request including a specific identifier" refers to a function whereby a user provides an identifier for specifying specific information and the system receives the identifier.

[1616] "Means for obtaining corresponding content from information storage means" refers to the function of searching and retrieving information based on a specified identifier from a database or storage within the system.

[1617] The "means for dynamically generating a prompt based on specific conditions" refers to a function for dynamically generating an appropriate prompt in response to a user's input or request.

[1618] The present invention provides a system for quickly and efficiently searching for previously distributed business announcements and instantly responding to the user about their content. The system includes a terminal through which the user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model.

[1619] Search function embodiment

[1620] User enters a search query

[1621] The user uses their own device to input the keywords they want to search for into the interface. For example, they can input "2023 business announcements" and press the search button to confirm the query input.

[1622] Sending queries by terminal

[1623] After pressing the search button, the terminal sends the entered query to the server. The data sent is as follows:

[1624] {

[1625] "query": "2023 Business Announcement"

[1626] }

[1627] Server receives and processes search queries

[1628] The server receives the search query sent from the device, analyzes the query, and dynamically generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[1629] "Please tell me what the business announcement for 2023 will be."

[1630] Prompting a Natural Language Generation Model

[1631] The server sends the generated prompts to a natural language generation model (e.g., a popular generative AI API), which generates a response based on the prompts and sends the result back to the server.

[1632] Receiving and formatting the response by the server

[1633] The server receives the response sent back by the generative model and formats it appropriately, which may include converting it to HTML or JSON format.

[1634] Terminal display of responses

[1635] The terminal receives the formatted response from the server and displays it to the user, allowing the user to quickly see the results of their query.

[1636] Detailed inquiry function embodiment

[1637] User enters detailed request

[1638] If the user wants to know the details of a specific business announcement, he or she inputs its ID and presses the Details button. For example, the user inputs "Details of the business announcement with ID 1."

[1639] Sending a details request by terminal

[1640] After pressing the Details button, the device will send the input request to the server, which will have the following format:

[1641] GET / detail / 1

[1642] Server receives and processes detailed requests

[1643] The server receives the detailed request sent from the device, analyzes the ID included in the request, and searches the database. Based on the information obtained from the database, it generates a prompt for the generative model. In this case, the generated prompt is in the following text format:

[1644] "Please tell me the details of the business announcement with ID 1."

[1645] Prompting a Natural Language Generation Model

[1646] The server sends the generated prompt to a natural language generation model to generate a detailed response, which is then sent back to the server.

[1647] Server response formatting and forwarding

[1648] The server receives the response from the generative model, formats it in an appropriate format, and sends it back to the device. Examples of returned data are as follows:

[1649] {

[1650] "response": "The details of the sales announcement with ID 1 are as follows..."

[1651] }

[1652] Terminal display of responses

[1653] The terminal displays the response received from the server on the screen, allowing the user to quickly check the correct information for the detailed request.

[1654] The present invention allows users to quickly and efficiently search past business announcements and get immediate responses to detailed information, greatly improving the efficiency of business announcement management and utilization.

[1655] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1656] Search function embodiment

[1657] Step 1:

[1658] User enters a search query

[1659] The user inputs the keywords they want to search for using the interface of their device. For example, they input "Business Notices for 2023" and press the search button. This operation confirms the query they entered.

[1660] Input: "Business Notice for 2023"

[1661] Output: The query is typed into the terminal

[1662] Step 2:

[1663] Sending queries by terminal

[1664] When the search button is pressed, the device sends the entered query to the server. Specifically, the query is packaged in JSON format and sent to the server as an HTTP request.

[1665] Input: "Business Notice for 2023"

[1666] Output: The query is sent to the server as an HTTP request.

[1667] Step 3:

[1668] Server receives search query

[1669] The server receives the query sent from the device, parses the received data in JSON format, and extracts the search keywords.

[1670] Input: Query from terminal

[1671] Output: Parsed search keywords

[1672] Step 4:

[1673] The server parses the query and generates prompts

[1674] The server analyzes the received query and dynamically generates a prompt based on its content, which becomes the question sent to the generative AI model.

[1675] How it works: Builds a prompt from search keywords.

[1676] Input: Parsed search keywords

[1677] Output: Generated prompt: "What is the content of your 2023 sales announcement?"

[1678] Step 5:

[1679] Prompting a Natural Language Generation Model

[1680] The server sends the generated prompt to a generative AI model (e.g., a general generative AI API).

[1681] What it does: Sends the prompt to the generative AI model as an HTTP request.

[1682] Input: Generated prompt

[1683] Output: A request to the NLG model

[1684] Step 6:

[1685] Receiving responses from generative AI models

[1686] The server receives a response from the generative AI model, which is a natural language answer generated based on the prompt.

[1687] Behavior: Receives the response as an HTTP response and parses it.

[1688] Input: The response returned by the generative AI model

[1689] Output: Parsed response content

[1690] Step 7:

[1691] Server response formatting

[1692] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display to the user.

[1693] Behavior: Formats the response content.

[1694] Input: Response from the generative AI model

[1695] Output: Formatted response data

[1696] Step 8:

[1697] Response from the server to the device

[1698] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[1699] Behavior: Sends response data as an HTTP response.

[1700] Input: Formatted response data

[1701] Output: Send response data to the terminal

[1702] Step 9:

[1703] Terminal display of responses

[1704] The terminal displays the response received from the server to the user, allowing the user to check the search results.

[1705] Behavior: Displays the response data in the interface.

[1706] Input: Response data from the server

[1707] Output: What is displayed to the user

[1708] Detailed inquiry function embodiment

[1709] Step 1:

[1710] User enters detailed request

[1711] If the user wants to know the details of a specific business announcement, he or she enters its ID and presses the Details button. For example, the user might enter "Details of the business announcement with ID 1." This operation confirms the request for details.

[1712] Input: "Details of business announcement with ID 1"

[1713] Output: Detailed request typed into terminal

[1714] Step 2:

[1715] Sending a details request by terminal

[1716] When the details button is pressed, the terminal transmits the input request to the server. Specifically, the request is transmitted to the server as an HTTP GET request.

[1717] Input: "Details of business announcement with ID 1"

[1718] Output: The detailed request is sent to the server as an HTTP GET request.

[1719] Step 3:

[1720] Server receives detailed request

[1721] The server receives a detailed request containing the ID sent by the device, which is then parsed to extract the ID.

[1722] Input: Detailed request from terminal

[1723] Output: Parsed ID

[1724] Step 4:

[1725] Server parses the request and performs a database lookup

[1726] The server searches the database based on the ID included in the detailed request and obtains the relevant business notice information.

[1727] What it does: Searches the database based on the ID and retrieves the relevant information.

[1728] Input: Parsed ID

[1729] Output: Business announcement information retrieved from the database

[1730] Step 5:

[1731] Server-generated prompts

[1732] Based on the information obtained, the server generates a prompt that is sent to the generative AI model, asking for more information.

[1733] How it works: Construct a prompt from the information obtained.

[1734] Input: Business announcement information retrieved from the database

[1735] Output: Generated prompt: "Please tell me the details of the sales announcement with ID 1."

[1736] Step 6:

[1737] Prompting a Natural Language Generation Model

[1738] The server sends the generated prompts to the generative AI model.

[1739] What it does: Sends the prompt to the generative AI model as an HTTP request.

[1740] Input: Generated prompt

[1741] Output: A request to the NLG model

[1742] Step 7:

[1743] Receiving responses from generative AI models

[1744] The server receives a response from the generative AI model, which is a detailed answer generated based on the prompt.

[1745] Behavior: Receives the response as an HTTP response and parses it.

[1746] Input: The response returned by the generative AI model

[1747] Output: Parsed response content

[1748] Step 8:

[1749] Server response formatting

[1750] The server formats the response from the generative AI model into an appropriate format, such as HTML or JSON, for easy display on the device.

[1751] Behavior: Formats the response content.

[1752] Input: Response from the generative AI model

[1753] Output: Formatted response data

[1754] Step 9:

[1755] Response from the server to the device

[1756] The server returns the formatted response data to the terminal, which is sent as an HTTP response.

[1757] Behavior: Sends response data as an HTTP response.

[1758] Input: Formatted response data

[1759] Output: Send response data to the terminal

[1760] Step 10:

[1761] Terminal display of responses

[1762] The terminal displays the response received from the server to the user, allowing the user to quickly check the information for the detailed request.

[1763] Behavior: Displays the response data in the interface.

[1764] Input: Response data from the server

[1765] Output: What is displayed to the user

[1766] (Application example 1)

[1767] 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."

[1768] On current online shopping sites, it can take a significant amount of time for users to quickly search and obtain information about past campaigns or detailed product information. Furthermore, there is a lack of systems that accurately provide the information users are looking for, resulting in a poor user experience. The present invention aims to solve these problems and provide a system that allows users to efficiently and quickly obtain the information they need.

[1769] 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.

[1770] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, means for providing a database for efficiently managing information searched in the past, and means for inputting a search query through an interface on a smartphone, thereby enabling users to quickly and accurately search and obtain information on past campaigns and detailed product information.

[1771] A "search query" is a keyword or phrase that a user enters to locate specific information.

[1772] A "natural language generation model" is an artificial intelligence technique that generates relevant sentences and responses based on input text.

[1773] A "response" is the information or answer provided by a system in response to a user's search query.

[1774] A "database" is a collection of data that stores previously searched information and detailed information in an organized manner so that it can be efficiently searched and retrieved later.

[1775] A "smartphone" is a type of mobile phone that has internet connectivity and a variety of applications, allowing users to perform various searches and obtain information.

[1776] An "interface" is a screen or operating means that a user uses to interact with a system.

[1777] A "detail request" is a request including a specific ID that a user enters when they want to know more about a specific piece of information.

[1778] The present invention is a system for quickly retrieving information and detailed information that a user has previously searched for. This system is operated via an interface installed on the user's smartphone. The main components of the system include the following:

[1779] System Configuration

[1780] Hardware elements

[1781] 1. Smartphone: A device where users enter search queries and detailed requests. It communicates with the server through an interface.

[1782] 2. Server: Receives queries and requests, retrieves information from the database, and generates and sends prompts to the natural language generation model.

[1783] Software Elements

[1784] 1. Interface: A smartphone application built using React Native that provides an interface for users to enter search queries and detailed requests.

[1785] 2. Server application: Built using Node.js and the Express framework, it handles user queries and requests, generates prompts, and sends them to the natural language generation model.

[1786] 3. Database Management System: MongoDB is used to store and manage past search queries and detailed information.

[1787] 4. Natural language generation models: Generative AI models provided as APIs (e.g., OpenAI's API) are used to generate relevant responses based on input queries.

[1788] Processing flow

[1789] Query Processing

[1790] A user uses the smartphone interface to enter a search query, for example, "Black Friday deals 2022," and presses the search button. This query is then sent from the smartphone to the server.

[1791] The server analyzes the received query and generates and sends the following prompt to the natural language generation model:

[1792] Search past sales announcements: Black Friday 2022 campaigns

[1793] The response returned by the generative AI model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain past campaign information.

[1794] Detailed Request Processing

[1795] If a user wants to know more information about a specific product or campaign, they enter its ID into the smartphone interface. For example, if they want to know the details of the sales announcement with ID 1, they enter the ID and press the details button. The details request is sent from the smartphone to the server.

[1796] The server receives the request, retrieves the relevant information from the database, and then generates and sends the following prompt to the natural language generation model:

[1797] Detailed information about business notice: [Content of information obtained]

[1798] The detailed response returned by the natural language generation model is received by the server and displayed to the user through the smartphone interface, allowing the user to quickly and accurately obtain the required details.

[1799] This invention enables users to quickly and efficiently search and obtain information about past campaigns and detailed product information. The collaboration between the database and the natural language generation model improves the user experience and enables efficient information acquisition.

[1800] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1801] Step 1:

[1802] A user uses a smartphone interface to enter a search query and presses the submit button, where the input is a keyword such as "Black Friday deals 2022." As an output, the smartphone sends this query to a server.

[1803] Step 2:

[1804] The server receives the search query sent by the user. The input is the user's search query, and the server analyzes the received query as output. As a result of the analysis, it generates a prompt such as "Search past sales announcements: Black Friday campaigns 2022."

[1805] Step 3:

[1806] The server sends the generated prompt to the natural language generation model. The input is the generated prompt, and the output is a response from the natural language generation model. The specific operation is to send the prompt via an API request and obtain the response as an API response.

[1807] Step 4:

[1808] The server analyzes the response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it processes the response text to format it for the user.

[1809] Step 5:

[1810] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[1811] Step 6:

[1812] The user inputs a detailed request including a specific ID into the smartphone interface and presses the send button. The input is "Details of the business announcement with ID 1." This request is sent from the smartphone to the server as an output.

[1813] Step 7:

[1814] The server receives the detailed request sent by the user. The input is the detailed request from the user, and the received request is analyzed as the output. As a result of the analysis, the information is retrieved from the database using the corresponding ID.

[1815] Step 8:

[1816] The server generates a prompt based on the information retrieved from the database and sends it to the natural language generation model. The input is the information retrieved from the database, and the generated prompt is sent as output. Specifically, it creates a prompt called "Sales Notice Details: [Content of retrieved information]" and sends it via an API request.

[1817] Step 9:

[1818] The server receives the detailed response returned by the NLG model. The input is the generated prompt and the output is the response from the NLG model. The action is to get the response as an API response.

[1819] Step 10:

[1820] The server parses the detailed response it receives and converts it into a format for sending back to the user. The input is the response from the natural language generation model, and the output is well-formed data that is sent back to the user. Specifically, it performs a process to format the response text.

[1821] Step 11:

[1822] The server sends the formatted response to the user's smartphone. The input is the formatted response data, and the output is the response displayed on the user's smartphone. The specific operation is to send the response data via an HTTP response.

[1823] Step 12:

[1824] Detailed information will be displayed on the user's smartphone, allowing the user to quickly and accurately obtain the information they need.

[1825] 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.

[1826] The present invention provides a system for quickly and efficiently searching for previously distributed sales announcements and instantly responding to users about their content. The system includes a terminal where a user inputs a query, a server for receiving and processing the query, and a function for generating a response using a natural language generation model. The system also combines an emotion engine that recognizes the user's emotions and adjusts the response based on those emotions to provide more personalized responses.

[1827] Search function embodiment

[1828] User enters and submits a search query

[1829] The user inputs a search query using the device interface, for example, "sales announcements for 2023." The device then identifies the emotion from the user's input and generates emotion data using an emotion engine.

[1830] Server receives and processes search queries

[1831] The server receives search queries and emotion data sent from the user's device, analyzes the received queries and emotion data, and generates prompts to send to the OpenAI API, which include adjustments based on the user's emotion.

[1832] Response generation and forwarding

[1833] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[1834] Detailed inquiry function embodiment

[1835] User enters and submits detailed request

[1836] The user inputs the ID of a specific business announcement and sends a request for details. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[1837] Server receives and processes detailed requests

[1838] The server receives the detailed request and emotion data sent from the user's device, extracts the ID included in the request, and retrieves the corresponding sales announcement information from the database. Based on the retrieved information, it generates a prompt to send to the OpenAI API. This prompt also includes adjustments based on the user's emotion.

[1839] Response generation and forwarding

[1840] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the received prompt and sends it back to the server. The server then organizes the response received from the OpenAI API, adjusting the tone and content to match the user's emotions, and sends it back to the user.

[1841] Specific examples

[1842] For example, if a user searches for "Business Announcements 2023," the search query sent from the device might look like this:

[1843] {

[1844] "query": "Business Notice 2023",

[1845] "emotion": "neutral"

[1846] }

[1847] The server receives the query and emotion data, generates prompts for the natural language generation model, and sends the resulting responses back to the user. The emotion engine is tuned to generate responses with a "neutral" tone.

[1848] Also, if a user inquires about "details of the business announcement with ID 1," the detailed request sent from the terminal will look like this:

[1849] GET / detail / 1

[1850] {

[1851] "emotion": "curious"

[1852] }

[1853] The server receives the request and emotion data, retrieves the relevant information from the database, and then generates and sends a prompt to the natural language generation model, which then returns a detailed response to the user. The emotion engine is tuned to generate responses with a "curious" tone.

[1854] The present invention enables a user to quickly and efficiently search for business announcements distributed in the past and to receive an immediate response regarding the detailed contents thereof, and furthermore, by providing a response adapted to the user's emotions, a more satisfying service can be realized.

[1855] The processing flow will be explained below.

[1856] Search function processing steps

[1857] Step 1:

[1858] A user uses a device interface to enter a search query, for example, "2023 business announcements."

[1859] Step 2:

[1860] When a user presses the search button, a search request including a query is sent from the terminal to the server. At this time, the terminal identifies the emotion from the user's input and generates emotion data using an emotion engine.

[1861] Step 3:

[1862] The server receives the search request and sentiment data, parses the query string and sentiment data, and generates a prompt to send to the natural language generation model.

[1863] Step 4:

[1864] The server sends the generated prompt to a natural language generation model for a response, where the prompt includes a modulation that reflects the user's sentiment.

[1865] Step 5:

[1866] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[1867] Step 6:

[1868] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[1869] Step 7:

[1870] The user sees the response displayed on the device, allowing the user to obtain information based on the search query.

[1871] Processing steps for detailed inquiry function

[1872] Step 1:

[1873] The user inputs the ID of a specific business notice through the terminal interface, for example, inputting "details of business notice with ID 1."

[1874] Step 2:

[1875] When the user presses the details button, a detailed request including the ID is sent from the device to the server. At this time, the device identifies the emotion from the user's input and generates emotion data using the emotion engine.

[1876] Step 3:

[1877] The server receives the detailed request and emotion data. Specifically, it extracts the ID included in the request and searches the database based on that ID to retrieve the relevant sales announcement.

[1878] Step 4:

[1879] The server uses the acquired sales pitch to generate prompts for transmission to a natural language generation model, which also include adjustments based on the user's sentiment.

[1880] Step 5:

[1881] The server sends the generated prompt to a natural language generation model and requests a response.

[1882] Step 6:

[1883] The natural language generation model generates a response based on the received prompt and sends the response back to the server.

[1884] Step 7:

[1885] The server organizes the responses it receives from the natural language generation model, adjusts the tone and content to match the user's emotions, and then sends them back to the user.

[1886] Step 8:

[1887] The user checks the response displayed on the terminal, allowing the user to obtain information based on the detailed request.

[1888] Example 2

[1889] 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."

[1890] Conventional systems have been unable to consider user emotions when quickly and accurately responding to users' search queries and detailed requests. This can result in mechanical, unpersonalized responses, potentially reducing user satisfaction. Furthermore, prompts are generated statically, making it difficult to provide intuitive responses. Therefore, a system that takes user emotions into account and generates more personalized responses quickly and efficiently is needed.

[1891] 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.

[1892] In this invention, the server includes a means for receiving a search query from a user, a means for generating a prompt sentence using the search query, and a means for obtaining a response from a generative AI model using the prompt sentence. This makes it possible to dynamically generate a prompt sentence in response to the user's search query and quickly and efficiently obtain a response from the generative AI model. Furthermore, by including a means for identifying the user's emotion and generating emotion data, and a means for adjusting the tone or content of the response based on the user's emotion data, it is possible to provide a personalized response based on the user's emotion.

[1893] "User" means a person who utilizes the system to enter a search query or detailed request.

[1894] A "search query" is text that a user enters into a system to search for information.

[1895] A "prompt" is a text instruction generated from a search query or detail request and sent to a generative AI model.

[1896] A "generative AI model" is an artificial intelligence that automatically generates a response based on an input prompt.

[1897] "Emotion data" is data indicating a user's emotion identified from the user's search query or detail request.

[1898] The "emotion engine" is the part of the system that identifies emotions from user input and generates emotion data.

[1899] A "response" is information that a generative AI model generates based on a prompt and sends back to the user.

[1900] A "detail request" is a user request for more information based on specific identifying information.

[1901] "Identification information" refers to information that uniquely identifies specific information within a database.

[1902] A "database" is a collection of information used to retrieve information in response to a user's request.

[1903] "Tone" refers to the emotional tone or mood of the generated response, which is adjusted based on the user's emotions.

[1904] The present invention is a system that quickly and efficiently generates responses based on search queries and detailed requests entered by users. The system includes a terminal where users enter queries, a server for receiving and processing queries, and a function for generating responses using a generative AI model. It also combines an emotion engine that recognizes the user's emotions and adjusts responses based on those emotions to provide more personalized responses.

[1905] Processing search queries

[1906] User enters and submits a search query

[1907] The user inputs a search query using the device interface. For example, they input "Business announcements for 2023." The device then identifies emotions from the user's input and generates emotion data using an emotion engine. The emotion data and the search query are then sent to the server.

[1908] Server receives and processes search queries

[1909] The server receives the search query and emotion data sent from the user's device. It analyzes the received query and emotion data and generates a prompt sentence to send to the generative AI model. This prompt includes adjustments based on the user's emotion. For example, it might be in the form of "Please provide detailed information about the sales announcements for 2023 in a neutral tone." The server then sends this prompt sentence to the generative AI model.

[1910] Response generation and forwarding

[1911] The generative AI model generates a response based on the prompt received from the server and sends it back to the server, which then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user.

[1912] Detail Request Processing

[1913] User enters and submits detailed request

[1914] The user inputs and sends a detailed request including specific identification information. For example, the user inputs "Details of the business announcement with ID 1." At this time, the device identifies emotions from the user's input and generates emotion data using an emotion engine. This emotion data and the detailed request are then sent to the server.

[1915] Server receives and processes detailed requests

[1916] The server receives the detailed request and emotion data sent from the user's device. It extracts the identification information included in the request and retrieves the corresponding sales announcement information from the database. For example, it retrieves information such as "ID: 1, Title: New Product Announcement, Content: Details about New Products for 2023." Based on this, it generates a prompt to send to the generative AI model. This prompt also includes adjustments based on the user's emotion. The server then sends this prompt to the generative AI model.

[1917] Response generation and forwarding

[1918] The generative AI model generates a response based on the prompt received from the server and sends it back to the server. The server then organizes the response received from the generative AI model, adjusts the tone and content to match the user's emotions, and sends it back to the user. For example, a response such as "The sales announcement with ID 1 is about a new product announcement. Details of the new product to be introduced in 2023 are..." may be generated.

[1919] In this way, the present invention can quickly and efficiently generate personalized responses that take into account the user's emotions. Specific examples of prompt sentences are shown below.

[1920] Prompt Sentence Examples

[1921] 1. Prompt based on search query:

[1922] "Please provide detailed information in a neutral tone regarding the 2023 business announcement."

[1923] 2. Prompt sentence based on detailed request:

[1924] "Please explain the details of the business announcement with ID 1 in an interesting tone."

[1925] In this way, responses to user inputs are provided in appropriate tone and content.

[1926] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1927] Step 1: User enters and submits a search query

[1928] The user inputs a search query using the device interface. For example, the user inputs "sales announcements for 2023." The device then generates emotion data using an emotion engine to identify emotions from the user's input. The device then sends the search query together with this emotion data to the server.

[1929] Input: Search query ("Sales announcements for 2023"), emotion data ("neutral")

[1930] Output: Query data sent to the server

[1931] Specific operation: The user enters "Sales announcement for 2023" into the device and clicks the send button. The emotion engine analyzes this input and generates emotion data called "neutral." The device then sends the query data to the server.

[1932] Step 2: The server receives the search query and generates a prompt.

[1933] The server receives the search query and emotion data sent from the user's device. After receiving the query, it analyzes the query content and generates a prompt to send to the generative AI model. This prompt includes adjustments based on the user's emotion, such as "Please provide detailed information in a neutral tone about the sales announcements for 2023."

[1934] Input: Query data received by the server (search query, sentiment data)

[1935] Output: A prompt to send to the generative AI model

[1936] Specific operation: The server analyzes the received query data and generates a prompt, such as "Please provide detailed information about the sales announcements for 2023 in a neutral tone," and prepares to send the prompt to the generation AI model.

[1937] Step 3: The server sends a prompt to the generative AI model

[1938] The server sends the generated prompt to the generative AI model, which then generates a response based on the prompt.

[1939] Input: A prompt to send to the generative AI model

[1940] Output: The response sent back from the generative AI model

[1941] Specific operation: The server generates a prompt and sends it to the generative AI model, which receives it and generates a response.

[1942] Step 4: Response generation and organization and coordination by the server

[1943] The server receives the responses sent back by the generative AI model and organizes them, adjusting the tone and content based on the user's emotions, sometimes again using the emotion engine.

[1944] Input: The response returned by the generative AI model

[1945] Output: The adjusted response that is sent back to the user

[1946] What it does: The response sent back to the server is processed by a natural language analysis module, and the emotion engine is used to adjust the tone and content, so the response might be formatted as something like, "Here's what our 2023 sales announcement will include..."

[1947] Step 5: Server sends back response

[1948] The server sends the adjusted response back to the user's terminal.

[1949] Input: Adjusted response

[1950] Output: The response that is displayed to the user

[1951] Specific operation: The server sends the adjusted response to the user's terminal, and the user confirms the response through the terminal interface.

[1952] (Application example 2)

[1953] 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."

[1954] Conventionally, when users search for trailers or reviews of previously distributed content (movies or TV dramas), they lack a means to quickly and efficiently obtain relevant information. Furthermore, there is a lack of systems that can provide personalized responses based on the user's emotions, creating a demand for an improved user experience. In particular, it is difficult to generate responses with an appropriate tone based on the user's emotions.

[1955] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1956] In this invention, the server includes means for receiving a search query from a user, means for obtaining a response from a natural language generation model using the search query, means for returning the obtained response to the user, and means for recognizing the user's emotion using emotion recognition means and adjusting the response based on the emotion, thereby enabling the user to quickly and efficiently search for trailers and reviews of previously distributed content (movies and dramas) and further provide personalized responses according to the user's emotion.

[1957] A "search query" refers to a word or phrase that a user enters to obtain specific information.

[1958] A "natural language generation model" refers to an algorithm or system that is trained on a huge amount of text data to generate natural-sounding sentences similar to those written by humans.

[1959] "Response" refers to the content generated and returned by the system in response to a user's search query.

[1960] "Emotion recognition means" refers to technology or systems that identify emotions from user input, facial expressions, voice, etc., and extract them as specific emotional data.

[1961] A "prompt" refers to an input sentence that instructs a natural language generation model what kind of response to generate.

[1962] A "detail request" refers to a request from a user that includes a specific identifier and asks for a more detailed response about a particular piece of information.

[1963] "Database" refers to a system or location for organizing, storing, and managing information.

[1964] "Identifier" refers to a string of symbols, numbers, or letters that uniquely identifies specific information.

[1965] This invention is a system that allows users to quickly and efficiently search for trailers and reviews of previously distributed content (movies and TV dramas) and receive instant responses regarding the details. It also includes various means for providing personalized responses according to the user's emotions. Specific methods for implementing this invention are described below.

[1966] System Configuration

[1967] Hardware

[1968] Device: Smartphone

[1969] Server: High-performance cloud server

[1970] Emotion Recognition Device: A module for recognizing emotions from voice and text data

[1971] software

[1972] Natural Language Generation model: OpenAI API (text-davinci-003)

[1973] Emotion Recognition Software: Emotion Recognition Model (e.g., Emotion AI SDK)

[1974] Program processing

[1975] Query Reception and Emotion Recognition

[1976] A user inputs a search query using a smartphone interface, for example, "Reviews of the 2021 movie 'Inception.'" The device receives the query and uses an emotion recognition device to identify the user's emotion from the input query.

[1977] Prompt Generation

[1978] The server generates a prompt to send to the natural language generation model based on the received search query and sentiment data. An example of a generated prompt would be, "I'm excited. I'm asking about a review of the 2021 film 'Inception.' I'd appreciate your information and support."

[1979] Generating and returning a response

[1980] The server sends the generated prompt to the OpenAI API and requests a response. The OpenAI API generates a response based on the prompt and sends it back to the server. The server then organizes the response, adjusts the tone and content to match the user's emotions, and sends it back to the user's smartphone.

[1981] Specific examples

[1982] For example, if a user searches for "2021 movie 'Inception' reviews," they'll get the following:

[1983] A user types "Review of the 2021 film Inception" into their device.

[1984] The terminal receives this query and obtains the emotion data "excited" from the emotion recognition device.

[1985] The server generates a prompt based on the query and emotion data and sends it to the OpenAI API in the form, "Emotion is excited. I'm asking about a review of the 2021 movie 'Inception'. Please provide information and support."

[1986] The OpenAI API generates a response and sends it back to the server.

[1987] The server sorts the responses, adjusts the tone appropriately based on the emotion of "excitement," and sends them back to the user's terminal.

[1988] The system allows users to quickly learn about previously distributed content and receive personalized, emotion-based responses.

[1989] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1990] Step 1:

[1991] A user enters a search query into a device.

[1992] Specifically, a user opens the application on their smartphone and types "reviews of the 2021 film 'Inception'" into the search bar.

[1993] Input: Search query (e.g., "2021 movie 'Inception' reviews")

[1994] Output: The search query entered by the user

[1995] Step 2:

[1996] The terminal receives the search query and recognizes the user's emotion using the emotion recognition means.

[1997] For example, the device analyzes the voice and word choice from the input text to identify the emotion "excitement."

[1998] Input: Search query from user

[1999] Data processing: Emotion recognition software analyzes text and extracts emotional data

[2000] Output: Search query and sentiment data (e.g., "excited")

[2001] Step 3:

[2002] The device sends the search query and emotion data to the server.

[2003] Specifically, the device generates a data packet to send the search query "2021 movie 'Inception' reviews" and the emotion "excited" to the server.

[2004] Input: Search query and sentiment data

[2005] Output: Data packet sent to the server

[2006] Step 4:

[2007] The server generates prompts based on the received search query and sentiment data.

[2008] For example, create a prompt like, "Emotions are excited. I'm asking about reviews of the 2021 film Inception. I need your input and support."

[2009] Input: Search query and sentiment data

[2010] Data processing: Prompt generation algorithm generates appropriate prompts

[2011] Output: Generated prompt

[2012] Step 5:

[2013] The server sends the generated prompt to the OpenAI API and requests a response.

[2014] Specifically, the server constructs an API request and sends it, including the prompt, to OpenAI's endpoint.

[2015] Input: Generated prompt

[2016] Output: The prompt sent to the OpenAI API

[2017] Step 6:

[2018] The OpenAI API generates a response based on the prompt it receives and sends the response back to the server.

[2019] For example, it might generate a response along the lines of, "Inception is a film directed by Christopher Nolan that has received very high praise overall..."

[2020] Input: Prompt sent

[2021] Data Computation: Response Generation with OpenAI's Natural Language Generation Model

[2022] Output: The generated response

[2023] Step 7:

[2024] The server sorts the returned responses, adjusts the tone and content to match the user's emotions, and then sends them back to the user's device.

[2025] For example, adjust the tone of your response to be more energetic based on the emotion "excited."

[2026] Input: The generated response

[2027] Data manipulation: Emotion engine adjusts tone and content of responses

[2028] Output: The adjusted response

[2029] Step 8:

[2030] The terminal receives the adjusted response and displays it to the user.

[2031] Specifically, the response content is displayed on the application's results screen.

[2032] Input: Adjusted response

[2033] Output: The response that is displayed to the user

[2034] Through these steps, users can quickly and efficiently search for trailers and reviews of previously distributed content, and obtain information that is appropriately tailored based on sentiment.

[2035] 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.

[2036] 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.

[2037] 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.

[2038] 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.

[2039] 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.

[2040] 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.

[2041] 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).

[2042] 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.

[2043] 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."

[2044] 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.

[2045] 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).

[2046] 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.

[2047] 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.

[2048] 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.

[2049] 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.

[2050] 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.

[2051] 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.

[2052] 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.

[2053] 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.

[2054] 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.

[2055] 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.

[2056] The following is further disclosed regarding the above embodiment.

[2057] (Claim 1)

[2058] means for receiving a search query from a user;

[2059] means for obtaining a response from a natural language generation model using the search query;

[2060] means for returning the obtained response to a user;

[2061] A system including:

[2062] (Claim 2)

[2063] means for receiving a detailed request from a user, the request including the specific ID;

[2064] A means for retrieving the relevant content from a database based on the ID;

[2065] means for obtaining a response from a natural language generation model using the obtained content;

[2066] means for returning the obtained response to a user;

[2067] 10. The system of claim 1, comprising:

[2068] (Claim 3)

[2069] 10. The system of claim 1, further comprising: means for dynamically generating a prompt to send to a natural language generation model when the search query and detail request are used to obtain a response from the natural language generation model.

[2070] "Example 1"

[2071] (Claim 1)

[2072] means for receiving input from a user;

[2073] means for dynamically generating prompts for a generative model based on the input;

[2074] means for obtaining a response from a generative model using the dynamically generated prompt;

[2075] means for returning the obtained response to a user;

[2076] A system including:

[2077] (Claim 2)

[2078] means for receiving a request from a user, the request including a specific identifier;

[2079] means for acquiring corresponding content from an information storage means based on the identifier;

[2080] means for generating a prompt for a generative model based on the acquired content;

[2081] means for obtaining a response from a generative model using the generated prompt;

[2082] means for returning the obtained response to a user;

[2083] 10. The system of claim 1, comprising:

[2084] (Claim 3)

[2085] 10. The system of claim 1, wherein the means for generating a prompt dynamically generates a prompt based on a specified condition.

[2086] "Application Example 1"

[2087] (Claim 1)

[2088] means for receiving a search query from a user;

[2089] means for obtaining a response from a natural language generation model using the search query;

[2090] means for returning the obtained response to a user;

[2091] means for providing a database for efficiently managing previously retrieved information;

[2092] a means for inputting a search query through an interface on a smartphone;

[2093] A system including:

[2094] (Claim 2)

[2095] means for receiving a detailed request from a user, the request including the specific ID;

[2096] A means for retrieving the relevant content from a database based on the ID;

[2097] means for obtaining a response from a natural language generation model using the obtained content;

[2098] means for returning the obtained response to a user;

[2099] a means for inputting detailed requests through an interface on a smartphone;

[2100] 10. The system of claim 1, comprising:

[2101] (Claim 3)

[2102] 10. The system of claim 1, further comprising: means for dynamically generating a prompt to send to a natural language generation model when the search query and detail request are used to obtain a response from the natural language generation model.

[2103] "Example 2: Combining Emotion Engines"

[2104] (Claim 1)

[2105] means for receiving a search query from a user;

[2106] means for generating a prompt sentence using the search query;

[2107] means for obtaining a response from a generative AI model using the prompt sentence;

[2108] means for returning the obtained response to a user;

[2109] means for identifying a user's emotion and generating emotion data;

[2110] means for adjusting the tone or content of the response based on the user's emotional data;

[2111] A system including:

[2112] (Claim 2)

[2113] means for receiving a request for details from a user, the request including specific identifying information;

[2114] means for acquiring corresponding content from a database based on the identification information;

[2115] means for generating a prompt sentence using the acquired content;

[2116] means for obtaining a response from a generative AI model using the prompt sentence;

[2117] means for returning the obtained response to a user;

[2118] 10. The system of claim 1, comprising:

[2119] (Claim 3)

[2120] 2. The system of claim 1, further comprising: means for dynamically generating a prompt sentence to send to the generative AI model when the search query and detail request are used to obtain a response from the generative AI model.

[2121] "Application example 2 when combining emotion engines"

[2122] (Claim 1)

[2123] means for receiving a search query from a user;

[2124] means for obtaining a response from a natural language generation model using the search query;

[2125] means for returning the obtained response to a user;

[2126] means for recognizing a user's emotion using emotion recognition means and adjusting a response based on said emotion;

[2127] A system including:

[2128] (Claim 2)

[2129] means for receiving a details request from a user, the details request including a specific identifier;

[2130] means for retrieving the relevant content from a database based on the identifier;

[2131] means for obtaining a response from a natural language generation model using the obtained content;

[2132] means for returning the obtained response to a user;

[2133] means for recognizing a user's emotion using emotion recognition means and adjusting a response based on said emotion;

[2134] 10. The system of claim 1, comprising:

[2135] (Claim 3)

[2136] 10. The system of claim 1, further comprising: means for dynamically generating a prompt to send to a natural language generation model when the search query and detail request are used to obtain a response from the natural language generation model. [Explanation of symbols]

[2137] 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 search query from a user; means for obtaining a response from a natural language generation model using the search query; means for returning the obtained response to a user; A system including:

2. means for receiving a detailed request from a user, the request including the specific ID; A means for retrieving the relevant content from a database based on the ID; means for obtaining a response from a natural language generation model using the obtained content; means for returning the obtained response to a user; The system of claim 1 , comprising:

3. The system of claim 1 , further comprising: means for dynamically generating prompts to send to a natural language generation model when the search query and detail request are used to obtain a response from the natural language generation model.

Images