system

Abstract

Provide a system. [Solution] The system includes: a means for receiving a chat log from a user; a means for analyzing the received chat log to extract the message content, sender, and sending date and time; a means for using a generative AI model to generate natural conversations based on the analysis results; and a means for suggesting the generated conversations to the user. A system including:

Description

[Technical Field]

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

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

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

[0004] Online communication has rapidly spread in modern society, but busy professionals and people looking for new friends or partners face the challenge of maintaining effective and natural conversations. Furthermore, efficient generation of conversations based on specific topics requires advanced natural language processing technology, but doing this manually requires a great deal of time and effort. Furthermore, the inability to provide conversation suggestions that take into account compatibility with the other person risks degrading the quality of communication. This invention aims to solve these problems. [Means for solving the problem]

[0005] This invention is a system that uses a generative AI model to receive chat logs from users, analyze the received chat logs to extract message content, sender, and sending date and time, and generate natural conversations based on the analysis results. Specifically, it includes means for suggesting the generated conversations to users and automatically matching compatible conversation partners based on the suggestions. It also includes means for receiving feedback from users and retraining and improving the generative AI model, thereby ensuring continuous service improvement. This allows users to maintain natural conversations without effort and enables effective communication with compatible partners.

[0006] "User" means an individual or organization that uses the system to upload chat logs and receive analysis results and generated conversations and matching suggestions.

[0007] A "chat log" is a record of interactions on social media or messaging applications, and is a data file containing information such as message content, sender, and date and time of sending.

[0008] "Receiving" is the process of taking the chat log file sent by the user into the system.

[0009] "Analysis" is the process of extracting important information from the received chat log, such as the message content, sender, and date and time of sending.

[0010] A "generative AI model" is an artificial intelligence model used to generate new, natural-sounding conversations based on past conversation data.

[0011] "Natural conversation" refers to messages that a generative AI model generates for a user in the form of a human conversation by analyzing past chat logs.

[0012] "Suggestion" refers to the act of introducing the generated conversation content and compatible conversation partners to the user.

[0013] A "compatible conversation partner" is another user who is determined to be capable of suitable communication based on the user's past conversation patterns and the generative AI model.

[0014] "Matching" is the process by which the system automatically selects suitable conversation partners and suggests them to the user.

[0015] "Feedback" is the act of a user providing the system with their opinions and evaluations of the generated conversations and suggestions.

[0016] "Retraining" is the process of improving the generative AI model based on the feedback received to generate more accurate conversations. [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 system of this invention receives chat logs from users, analyzes the received chat logs, generates natural conversations using a generative AI model, and suggests them to the user. It also includes a means for automatically matching compatible conversation partners and continuously improving the system based on user feedback.

[0039] Receive chat logs from users

[0040] The user selects the chat log they want to upload using their device. The selected chat log is sent to the server as an HTTP POST request by the device. The server waits for the incoming request and receives the chat log file. The received file is saved in a specified directory on the server.

[0041] Chat log analysis

[0042] The server reads the saved chat log file and analyzes the contents line by line. During the analysis, information such as the message content, sender, and date and time of sending is extracted. This extracted data is then filtered out and converted into a format that is easy for the generative AI model to handle.

[0043] Generating natural conversations using generative AI models

[0044] The server loads and initializes a pre-trained generative AI model. The converted dataset is input to the generative AI model, which generates new, natural-sounding conversations based on past conversation patterns. The generated conversations are output in text format.

[0045] Conversation suggestions and matching compatible conversation partners

[0046] The server proposes the generated conversations to the user. These proposals include conversations with other users who are deemed to be compatible. Compatible partners are selected by comparing the user's past conversation patterns with those of other users. The proposals are sent to the user's device, and the user confirms their contents.

[0047] Gathering feedback and improving the system

[0048] The user provides feedback on the proposed conversation content, which is then sent from the device to the server, which analyzes the received feedback and retrains and refines the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0049] Specific examples

[0050] For example, if User A uploads a chat log exported from the LINE app to the system, the server receives and analyzes this chat log. If there were many conversations about "movies" in past chats, the generative AI model will suggest to User A that they talk about movies they've recently seen. If another User B is also interested in movies, the server can suggest matching User A and User B. User A can provide feedback on the suggested conversation, allowing the system to provide even more accurate conversation suggestions.

[0051] In this way, users can carry on natural conversations without any hassle, and effective communication can be achieved.

[0052] The processing flow will be explained below.

[0053] Step 1:

[0054] The user uses the terminal interface to select the chat log file to be uploaded.

[0055] Step 2:

[0056] The device sends the selected chat log file to the server as an HTTP POST request.

[0057] Step 3:

[0058] The server receives the chat log file sent from the device and saves it in the specified directory.

[0059] Step 4:

[0060] The server reads the saved chat log file and parses the contents line by line.

[0061] Step 5:

[0062] The server extracts information such as message content, sender, and sent date and time from the analyzed chat log.

[0063] Step 6:

[0064] The server removes unnecessary information and converts the extracted data into a format suitable for the generative AI model.

[0065] Step 7:

[0066] The server loads and initializes the pre-trained generative AI model.

[0067] Step 8:

[0068] The server inputs the converted data set into a generative AI model to analyze past conversation patterns.

[0069] Step 9:

[0070] The server uses the generative AI model to generate new, natural-sounding conversations.

[0071] Step 10:

[0072] The server outputs the generated conversation in text format and suggests it to the user.

[0073] Step 11:

[0074] The server compares the user's past conversation patterns with those of other users and automatically matches them with compatible conversation partners.

[0075] Step 12:

[0076] The server transmits information about conversation partners who are compatible with the proposed conversation to the user's terminal.

[0077] Step 13:

[0078] The user checks the proposed conversation content on the terminal and provides feedback if necessary.

[0079] Step 14:

[0080] The terminal sends the user's feedback to the server.

[0081] Step 15:

[0082] The server analyzes the feedback received and retrains and refines the generative AI model.

[0083] Step 16:

[0084] The server applies the improved generative AI model to future conversation generation.

[0085] Example 1

[0086] 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."

[0087] Conventional chat systems have not been able to generate natural conversations between users or match compatible conversation partners, making it difficult to achieve effective communication. Furthermore, the accuracy of the system has not been improved sufficiently because continuous improvements based on user feedback have not been made.

[0088] 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.

[0089] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and transmission date and time, means for using a natural language processing model to generate natural conversation based on the analysis results, means for suggesting the generated conversation to the user, means for selecting a compatible conversation partner based on the user's past conversation patterns, and means for receiving user feedback on the suggested conversation content and analyzing the feedback to retrain and improve the natural language processing model. This allows users to enjoy natural conversation and communicate effectively with compatible conversation partners. Furthermore, since the system is continuously improved based on the feedback, the accuracy of conversation generation can continue to improve.

[0090] "User" means an individual or organization that uses the system to upload chat logs.

[0091] A "chat log" is a data file that records messages exchanged between users.

[0092] A "server" is a computer system that receives requests from users and analyzes, stores, and processes chat logs.

[0093] "Receiving means" refers to the process and technical elements for receiving chat logs from users.

[0094] The "analysis means" is a process for analyzing the contents of the received chat log and extracting the message contents, sender, and sending date and time.

[0095] A "natural language processing model" is a model that uses AI technology to understand and generate natural human language.

[0096] The "generation means" is a process for generating new conversations using a natural language processing model based on the analysis results.

[0097] The "suggestion means" refers to the process and technical elements for suggesting the generated conversation to the user.

[0098] The "compatibility selection means" is a process for analyzing the user's past conversation patterns and selecting a conversation partner with good compatibility.

[0099] "Feedback" refers to the evaluation or opinion that a user provides regarding the proposed conversation content.

[0100] "Analysis means (feedback)" is the process of analyzing user feedback and reflecting it in improving the accuracy of the system.

[0101] The "restudy and refinement" means a process for retraining and refining a natural language processing model based on received feedback.

[0102] The system according to the present invention receives chat logs from users, analyzes them, generates natural conversations, makes suggestions, and improves models based on the feedback. How this system is implemented will be described below in detail.

[0103] Receive chat logs from users

[0104] The user accesses the system from their device and selects a chat log file. For example, they can use a web browser or a dedicated mobile app to click the "Select File" button and select a chat history file from LINE or another service. The device then sends the selected chat log file to the server as an HTTP POST request. The server receives the HTTP request and saves the chat log file in the " / uploads / chatlogs / " directory.

[0105] Chat log analysis

[0106] The server reads the saved chat log file. It parses the file line by line and extracts information such as the message content, sender, and date and time of sending. Using regular expressions and pattern matching, it extracts the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello." The extracted information is then converted into a format that is easy for the generative AI model to handle. For example, it can be formatted as "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0107] Generating natural conversations using generative AI models

[0108] The server loads and initializes a pre-trained generative AI model (e.g., using TENSORFLOW (registered trademark) or PyTorch). It inputs a formatted dataset into the model and performs inference to generate new, natural-sounding conversations. The generated conversations are output in text format. For example, the model generates a new message such as, "What movie have you seen recently?"

[0109] Conversation suggestions and matching compatible conversation partners

[0110] The server then incorporates the generated conversation into a template and formats it in HTML or JSON to propose it to the user. Furthermore, the server reads the user's past conversation patterns from the server's database and analyzes their compatibility with other users. A matching algorithm is used to evaluate similarities and differences and select appropriate conversation partners. For example, it might suggest to user A, "Let's talk about movies you've seen recently," and match them with user B, who also likes movies. This suggestion is sent to the user's device as an HTTP response, and the user can view the content in their browser or app.

[0111] Gathering feedback and improving the system

[0112] The user provides feedback on the suggested conversation content. For example, they can enter an evaluation such as "helpful" or "not helpful" using a dedicated button. This feedback is sent to the server in JSON format. The server receives the feedback data, stores it in a database, and analyzes it to reflect in the retraining of the model. Retraining improves the accuracy of the generative AI model, and more appropriate conversation suggestions will be made the next time the system is used.

[0113] Prompt Sentence Examples

[0114] "How do I upload my LINE chat logs and receive new conversation suggestions?"

[0115] In this way, the system can continuously provide users with natural conversations and achieve effective communication with compatible partners. Furthermore, by incorporating user feedback, the system's accuracy can be continuously improved.

[0116] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0117] Step 1: Receive chat logs from users

[0118] A user accesses the system from his / her terminal and selects a chat log file.

[0119] Specifically, click the "Select File" button on your browser or dedicated app and select a chat history file such as LINE.

[0120] The terminal sends the selected chat log file to the server as an HTTP POST request.

[0121] The server receives the HTTP request and saves the chat log files in the " / uploads / chatlogs / " directory.

[0122] Input: Chat log file selected by the user

[0123] Output: Chat log file saved on the server

[0124] Step 2: Analyzing the chat logs

[0125] The server reads the saved chat log file.

[0126] Specifically, it analyzes the file line by line and extracts information such as the message content, sender, and date and time of sending.

[0127] Using regular expressions and pattern matching, extract the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello."

[0128] The extracted information is converted into a format that is easy for the generative AI model to handle, for example, "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0129] Input: Chat log file stored on the server

[0130] Output: Formatted chat data

[0131] Step 3: Generating natural conversation using a generative AI model

[0132] The server loads and initializes the pre-trained generative AI model.

[0133] Specifically, the generative AI model is loaded from disk into memory using libraries such as TensorFlow or PyTorch.

[0134] The formatted dataset is fed into the model for inference, which generates new, natural-sounding conversations. For example, the model generates new messages such as, "What movie have you seen recently?"

[0135] The generated conversation is output in text format.

[0136] Input: Formatted chat data

[0137] Output: The new conversational text generated by the generative AI model

[0138] Step 4: Conversation suggestions and matching with compatible conversation partners

[0139] The server proposes the generated conversations to the user.

[0140] Specifically, the generated conversation is incorporated into a template, configured in HTML or JSON format, and an HTTP response is prepared for sending to the user.

[0141] Furthermore, the server reads the user's past conversation patterns from a database and analyzes the similarities and differences between them and other users.

[0142] A matching algorithm is used to select an appropriate conversation partner. For example, if user A is asked to talk about a movie they recently saw, the system will match them with user B, who also likes movies.

[0143] The suggestions are sent to the user's device, and the user can view them in their browser or app.

[0144] Input: Generated conversation text, user's past conversation data

[0145] Output: Conversation suggestions for the user and selection of compatible conversation partners

[0146] Step 5: Gather feedback and refine the system

[0147] The user provides feedback on the proposed conversation content.

[0148] Specifically, the user selects a rating such as "useful" or "not useful" and clicks a dedicated button.

[0149] The device sends the feedback data in JSON format to the server as a POST request.

[0150] The server stores the received feedback in a database, analyzes it, and reflects it in the retraining of the model.

[0151] Re-learning improves the accuracy of the generative AI model, allowing for more appropriate conversation suggestions the next time it is used.

[0152] Input: User feedback data

[0153] Output: Highly accurate conversation suggestions from an improved generative AI model

[0154] (Application example 1)

[0155] 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."

[0156] Modern communication tools make it difficult for users to find content and topics that interest them. Finding compatible conversation partners and generating natural conversations based on topics of interest is also challenging, especially for busy users. Furthermore, there is a lack of feedback mechanisms to improve the quality of generated content and conversations. This leaves users with insufficient support for effective communication.

[0157] 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.

[0158] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, senders, and sending dates and times, means for using a generative AI model to generate natural conversations and content based on the analysis results, and means for suggesting the generated conversations and content to users. This allows users to easily obtain natural conversations and content based on topics of interest and engage in appropriate communication.

[0159] A "chat log" is text data of the conversation content that a user sends and receives through a chat application.

[0160] "Analysis" is the process of extracting useful information from data and converting it into an understandable form.

[0161] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new conversations and content.

[0162] "Suggestion" is the act of presenting information or content that is deemed useful to the user.

[0163] "Content" refers to information or entertainment that may be of interest to users, such as movies, music, articles, etc.

[0164] A "compatible conversation partner" is another user with whom effective communication can be expected based on the user's interests and past conversation patterns.

[0165] "Feedback" refers to opinions and comments provided by users regarding a system or service.

[0166] "Retraining" is the process by which an AI model is retrained based on new data and feedback to improve its performance.

[0167] The system realizing this invention analyzes a user's chat log and uses a generative AI model to suggest natural conversations and content. A specific embodiment of this system will be described below.

[0168] Hardware and Software Use

[0169] Hardware:

[0170] Smartphone (OS: ANDROID (registered trademark), iOS)

[0171] Cloud server (e.g. AWS (registered trademark) EC2)

[0172] software:

[0173] Frontend: React Native (cross-platform mobile app development)

[0174] Backend: Node.js and Express (Building the API server)

[0175] Database: MongoDB (for storing chat logs)

[0176] AI model: GPT-4 (registered trademark) (using OpenAI (registered trademark) API)

[0177] System operation explanation

[0178] Receive chat logs:

[0179] The user selects the chat log they want to upload using their smartphone and sends it to the server via an HTTP POST request. The server receives it and saves it in the specified directory.

[0180] Chat log analysis:

[0181] The server reads the saved chat logs and parses them line by line, including the message content, sender, and date and time, and then removes unnecessary parts and converts them into a format that is easy for the generative AI model to handle.

[0182] Conversation and content generation using generative AI models:

[0183] The server loads and initializes a pre-trained generative AI model, inputs the converted dataset into the model, and generates new, natural-sounding conversations and content based on past conversation patterns, which are then output in text format.

[0184] Conversation suggestions and compatible conversation partner matching:

[0185] The server then proposes the generated conversations and content to the user. It compares the user's past conversation patterns with those of other users and also suggests compatible partners. The proposals are sent to the user's device, where the user can view them.

[0186] Gathering feedback and improving the system:

[0187] Users provide feedback on the suggestions, which is sent to the server and analyzed to retrain and refine the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0188] Specific examples

[0189] For example, if User A uploads past chat logs to the app, the server receives and analyzes these chat logs. If User A has had many conversations about "movies" in past chats, the generative AI model will generate content to "talk about movies you've recently seen" and suggest it to User A. At the same time, User B, who is also interested in movies, may also be matched.

[0190] Example prompt sentence:

[0191] User's past chat logs:

[0192] 1. "Movies have been interesting lately."

[0193] 2. "What kind of movie would you like?"

[0194] 3. "I like action. I like it when it's really epic."

[0195] Generates prompt:

[0196] "The user likes action movies. Tell the AI ​​model to talk about the top 10 latest action movies."

[0197] In this embodiment, the user can easily obtain natural conversations and content based on topics of interest, and can carry out appropriate communication.

[0198] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0199] Step 1:

[0200] A user selects a chat log using a smartphone and sends it to the server through an HTTP POST request. The input is the locally stored chat log file, and the output is the HTTP request sent to the server.

[0201] Step 2:

[0202] The server saves the received chat log in the specified directory. It writes the received data (chat log file) to the disk and gets the save path as output.

[0203] Step 3:

[0204] The server reads the saved chat log file and analyzes the contents line by line. This analysis extracts the message content, sender, and sending date and time. The input is the text data of the chat log, and the output is structured data (e.g., JSON format) as the analysis result.

[0205] Step 4:

[0206] The analyzed data is converted into a format that is easy for the generative AI model to handle. It is converted into data with a specific structure (e.g., context, entities, keywords, etc.) and output as a dataset to be input into the generative AI model.

[0207] Step 5:

[0208] The server loads a pre-trained generative AI model and inputs the converted dataset into the model. The generative AI model generates new, natural-sounding conversations and content based on past conversation patterns. The input is the initialization parameters of the generative AI model and the converted dataset, and the output is the text data of the generated conversation or content.

[0209] Step 6:

[0210] The generated conversation and content are sent to the user's device as a suggestion. The server returns the generated result as an HTTP response, and the user receives the suggestion on the application. The input is the generated text, and the output is the suggestion displayed on the user's device.

[0211] Step 7:

[0212] The user provides feedback on the proposed content. The feedback is sent from the user's terminal to the server as an HTTP POST request. The input is the user's feedback text, and the output is saved as feedback data to the server.

[0213] Step 8:

[0214] The server analyzes the received feedback and retrains and improves the generative AI model. It analyzes the feedback data to extract important information and uses this information to retrain the generative AI model. The input is the feedback data, and the output is an improved generative AI model.

[0215] The specific operations at each step are carried out in sequence by the cooperation of various components, such as the server, terminal, and user, and through processes such as data transmission and reception, analysis, generation, and feedback, effective conversations and content are provided to users.

[0216] 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.

[0217] The system of this invention receives chat logs from users, analyzes them, and extracts message content, sender, and sent date and time. It also uses a generative AI model to generate natural conversations and suggest them to the user. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions based on the user's emotions.

[0218] Receive chat logs from users

[0219] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[0220] Chat log analysis

[0221] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[0222] Generating natural conversations using generative AI models

[0223] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the server generates new, natural-sounding conversations. The generated conversations are output in text format.

[0224] Conversation suggestions and matching compatible conversation partners

[0225] The server adjusts the generated conversation based on the user's emotions and makes suggestions. The suggestions include conversations with other users who are deemed compatible. The emotion engine reflects appropriate conversation content based on the emotional information recognized by the server. The server then sends the suggestions to the user's device, where the user can confirm them.

[0226] Gathering feedback and improving the system

[0227] The user provides feedback on the proposed conversation content and sends it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0228] Specific examples

[0229] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes the chat log. The emotion engine also identifies the emotional state of the message content. For example, if past conversations have often focused on movies and the user appears to be enjoying themselves, the generative AI model will suggest, "Let's talk about an interesting movie you recently saw." If User B is also interested in movies and appears to be enjoying themselves, the server will suggest matching User A and User B, providing a conversation that is appropriate for both parties. By providing feedback on this suggestion, the system can reflect it in future conversation generation and matching, enabling it to provide even more accurate and natural conversations.

[0230] In this way, users can enjoy natural conversations without much effort and can achieve effective communication with people with whom they have good chemistry.

[0231] The processing flow will be explained below.

[0232] Step 1:

[0233] The user uses the terminal interface to select the chat log file to be uploaded.

[0234] Step 2:

[0235] The device sends the selected chat log file to the server as an HTTP POST request.

[0236] Step 3:

[0237] The server receives the chat log file sent from the device and saves it in the specified directory.

[0238] Step 4:

[0239] The server reads the saved chat log file and parses the contents line by line.

[0240] Step 5:

[0241] The server extracts message content, sender, and sent date and time from the analyzed chat log.

[0242] Step 6:

[0243] The server uses an emotion engine to analyze the emotional state of each message and add emotion information to the extracted data.

[0244] Step 7:

[0245] The server initializes the generative AI model and inputs the formatted dataset containing emotion information.

[0246] Step 8:

[0247] The server uses a generative AI model to generate new, natural conversations based on past conversation patterns and emotional information.

[0248] Step 9:

[0249] The server outputs the generated conversation in text format and adjusts it based on the user's emotions.

[0250] Step 10:

[0251] The server proposes tailored conversations to the user and compares them with other users' data to match compatible conversation partners.

[0252] Step 11:

[0253] The server identifies compatible conversation partners and sends suggestions to the user's device.

[0254] Step 12:

[0255] The user reviews the suggested conversation content on their device and adds feedback if necessary.

[0256] Step 13:

[0257] The terminal sends the user's feedback to the server.

[0258] Step 14:

[0259] The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine.

[0260] Step 15:

[0261] The server applies the improved generative AI model and emotion engine to future conversation generation.

[0262] Example 2

[0263] 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."

[0264] While conventional chat log analysis systems were able to extract message content, senders, and sending dates and times, they had problems in that they did not adequately generate natural conversations that took the user's emotional state into account or match compatible conversation partners. Furthermore, there was a lack of a way to effectively utilize user feedback on the generated conversation content to improve the system, making it difficult to consistently generate high-quality conversations.

[0265] 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.

[0266] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and sending date and time, means for identifying the user's emotional state using an emotion engine that estimates emotions, means for generating a format-converted dataset based on the analysis result and the identified emotional state, means for using a generative AI model that uses the generated dataset as input to generate natural conversations, means for adjusting and suggesting the generated conversations based on the user's emotions, and means for receiving feedback and improving the system. This makes it possible to generate natural conversations that reflect the user's emotions and match users with compatible conversation partners, and further allows the system to be continuously improved using user feedback.

[0267] A "chat log" is a data file that shows the history of messages that a user has sent and received in the past.

[0268] An "emotion engine" is software or algorithm that analyzes message content to infer a user's emotional state.

[0269] A "generative AI model" is an artificial intelligence model that generates natural conversations based on pre-trained data.

[0270] A "formatted dataset" is a collection of input data that has been converted into a format acceptable to a generative AI model based on the results of analysis and emotion estimation.

[0271] The "means for suggesting" is a method or function for presenting the generated conversation content to the user.

[0272] "Feedback" is the act of a user providing an evaluation or comment on the proposed conversation content.

[0273] "Retraining" is a new learning process that uses feedback to improve the performance of generative AI models and emotion engines.

[0274] "Matching means" refers to a method or function that automatically connects users who are determined to be compatible with each other.

[0275] This invention is a system that receives chat logs from users, analyzes the logs, and generates and suggests natural conversations. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback.

[0276] First, the user selects the chat log they want to upload using their device. The selected chat log is then sent from the device to the server. The server then saves the received chat log in a specified directory. This process uses standard HTTP POST requests and file I / O operations.

[0277] The server then reads the saved chat log file and analyzes it line by line. During the analysis process, it extracts the message content, sender, and date and time of sending. For example, this involves using regular expressions to extract this information from the text data. It then uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data includes the message content, sender, date and time, as well as emotional information.

[0278] Next, the server initializes a pre-trained generative AI model and inputs the converted dataset. Based on past conversation patterns and emotional information, the model generates new, natural-sounding conversations. For example, if the conversation revolves around "recent movies," the model might suggest, "Let's talk about some interesting movies you've seen recently."

[0279] The generated conversation is output in text format, and the server adjusts and suggests conversations based on the user's emotions. These suggestions include conversations with other users who are deemed compatible. Based on the emotional information recognized by the emotion engine, appropriate conversation content is generated and sent to the user's device.

[0280] Users can also provide feedback on the proposed conversation content and send it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0281] Specific examples

[0282] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes this chat log. The emotion engine also identifies the emotional state of the message content. For example, if the message "Hello" is evaluated as "positive," the generative AI model uses this information to generate a new message such as "How are you doing lately?"

[0283] Furthermore, if User B is also interested in movies and seems to enjoy them, the server will propose a match between User A and User B, providing a conversation suitable for both parties. Feedback on this proposal is provided, allowing the system to reflect this in future conversation generation and matching, enabling the system to provide even more accurate and natural conversations.

[0284] An example prompt is:

[0285] User A's emotional state: Happy

[0286] Past conversation topics: Movies, Travel, Music

[0287] Latest message: "What movie did you see last?"

[0288] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0289] Step 1:

[0290] The user selects the chat log they wish to upload on their device. This is done by using the device's file browser to find and select the specified chat log file. At this time, the system recognizes the selected file as the "selected chat log file."

[0291] Step 2:

[0292] The terminal sends the selected chat log file to the server. In this process, an HTTP POST request is used to send the file contents to the server. The input is the "selected chat log file" and the output is the "chat log file sent to the server."

[0293] Step 3:

[0294] The server saves the received chat log file in the specified directory. The file name is guaranteed to be unique using the user ID and timestamp. The input is the "chat log file sent to the server" and the output is the "saved chat log file."

[0295] Step 4:

[0296] The server reads and analyzes the saved chat log file line by line. Specifically, it uses regular expressions to extract the message content, sender, and sending date and time from each line. The input is the "saved chat log file," and the output is the "extracted message content, sender, and sending date and time."

[0297] Step 5:

[0298] The server uses an emotion engine to identify the user's emotional state from the analyzed message content. For example, it classifies the content into categories such as positive, negative, and neutral. The input is the extracted message content, and the output is the identified emotional state.

[0299] Step 6:

[0300] The server generates a format-converted dataset for the generative AI model based on the identified emotional state and analysis results. This is the process of formatting the data into a format that the generative AI model can understand. The input is the "analysis results and identified emotional state," and the output is the "format-converted dataset."

[0301] Step 7:

[0302] The server initializes the generative AI model and inputs the converted dataset, which generates natural-sounding conversations. The input is the converted dataset, and the output is the generated natural-sounding conversations.

[0303] Step 8:

[0304] The server adjusts the generated natural conversation based on the user's emotions and suggests it to the user. The input is the generated natural conversation and the identified emotional state, and the output is the suggested conversation content.

[0305] Step 9:

[0306] The user provides feedback on the proposed conversation content and sends it to the server via their device. The feedback can include ratings and specific comments. The input is the "proposed conversation content" and the output is "feedback from the user."

[0307] Step 10:

[0308] The server analyzes the received feedback and retrains and improves the generative AI model and emotion engine based on the content. The input is "feedback from the user" and the output is "an improved generative AI model and emotion engine."

[0309] These processing steps enable the system to generate natural conversations that reflect the user's emotions, match users with compatible conversation partners, and continuously improve the system.

[0310] (Application example 2)

[0311] 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."

[0312] Conventional chat analysis systems do not take into account users' emotional information when proposing advertisements, making it difficult to effectively present advertisements that match users' interests and needs. Furthermore, due to a lack of means to promote joint purchases and information exchange among users with similar interests, the relevance of advertisements tailored to users' preferences and the improvement of engagement are not sufficiently addressed.

[0313] 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.

[0314] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, sender, and sending date and time, means for using a generative AI model to generate natural conversations based on the analysis results, means for generating advertisements based on the content of the generated conversations and the user's emotional information and proposing them to the user, means for encouraging joint purchases and information exchanges with other users with similar interests based on the proposed advertisements, and means for receiving feedback on the advertisements from users and retraining and improving the generative AI model and the emotional engine. This makes it possible to propose advertisements that reflect the user's emotional state and interests, further increasing engagement among users with similar interests and maximizing the effectiveness of advertisements.

[0315] A "user" is a user who provides a chat log to the system and receives the analysis results and suggestions.

[0316] A "chat log" refers to messages and conversation history sent and received by a user, and is data that includes the content, sender, and date and time of sending.

[0317] "Emotion information" refers to the user's emotional state identified based on an analysis of the chat log, and includes positive, negative, neutral, and other emotions.

[0318] A "generative AI model" refers to a pre-trained artificial intelligence model, an algorithm for generating natural conversations and advertisements based on chat logs and emotional information.

[0319] "Advertisement" refers to marketing messages and product information that are generated based on the user's interests and emotional information and are proposed to the user.

[0320] "Suggestion" refers to the act of presenting generated conversations or advertisements to the user, which contain content that corresponds to the user's interests and emotions.

[0321] "Feedback" refers to user reactions and opinions on proposed conversations and advertisements, which are used to improve the system and retrain the generative AI model.

[0322] "Joint purchasing" refers to the act of purchasing products in collaboration with other users who share similar interests, and includes means by which the system encourages this.

[0323] "Information exchange" refers to the sharing of knowledge and opinions among other users with similar interests, and includes the means by which the system facilitates this.

[0324] To implement this invention, it is necessary to build a system that analyzes users' chat logs, generates optimal advertisements based on their emotional information, and proposes them to users. This system is composed of hardware and software such as a server, user terminals, a generative AI model, an emotion engine, and a feedback database.

[0325] The server receives and analyzes chat logs sent from users' devices. Specifically, it uses a natural language processing library (e.g., NLTK) and a sentiment analysis engine (e.g., Microsoft® Azure® Text Analytics) to extract message content, sender, and sending date and time from the chat logs. This also identifies the user's sentiment information.

[0326] The analyzed data is then fed into a generative AI model (e.g., OpenAI GPT-4), which then generates optimal ads based on past conversation patterns and emotional information. This generation process ensures that the ads include content that reflects the user's interests and emotions.

[0327] The generated advertisements are then presented to the user's device, where they can be viewed using a smartphone or smart glasses. The system also matches users with similar interests, facilitating joint purchases and information exchange.

[0328] User feedback on ads is sent to the server through a feedback collection system. The server analyzes this feedback and retrains and refines the generative AI model and emotion engine. This feedback process uses a database (e.g., MySQL®).

[0329] As a specific example, if a user uploads a chat log about "movies" to the system, the sentiment analysis engine will identify the user's positive emotions. The generative AI model will then refer to past conversation patterns about "movies" to generate ads for "interesting movies they've recently seen." These ads will be suggested to the user, encouraging them to share information and make joint purchases with other users who are interested in movies.

[0330] Here are some example prompts to input to a generative AI model:

[0331] "User conversation and sentiment: [{'content': 'I like movies', 'sender': 'user1', 'timestamp': '2023-10-24T14:53:00', 'sentiment': {'neg': 0.0, 'neu': 0.3, 'pos': 0.7, 'compound': 0.7}}]. Generate the best ad."

[0332] This allows for advertising suggestions that reflect a user's emotional state and interests, further increasing engagement between users with similar interests.

[0333] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0334] Step 1:

[0335] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[0336] Input: Chat log (data sent from the user's device)

[0337] Output: Saved chat log file (stored in a directory on the server)

[0338] Specific behavior:

[0339] 1. The user selects the chat log file through the terminal interface.

[0340] 2. The device sends the selected file to the server.

[0341] 3. The server saves the received file in the specified directory.

[0342] Step 2:

[0343] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[0344] Input: Saved chat log file

[0345] Output: Analysis results (message content, sender, sending date and time, emotional information)

[0346] Specific behavior:

[0347] 1. The server reads the chat log file.

[0348] 2. Use a natural language processing library (e.g., NLTK) to extract the message content, sender, and date and time.

[0349] 3. Identify the sentiment of each message using a sentiment analysis engine (e.g., Microsoft Azure Text Analytics).

[0350] 4. Create a dataset to input the generative AI model.

[0351] Step 3:

[0352] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the model generates new, natural-looking ads. The generated ads are output in text format.

[0353] Input: Analysis result dataset (including emotion information)

[0354] Output: Generated advertisement (text format)

[0355] Specific behavior:

[0356] 1. The server initializes a generative AI model (e.g., OpenAI GPT-4) and inputs the dataset.

[0357] 2. The generative AI model generates ads based on the input data.

[0358] 3. Create the ad copy generated as output.

[0359] Step 4:

[0360] The server then adjusts the generated advertisements based on the user's emotions and suggests them to the user. It also matches users with similar interests and shares the advertisements with them. The server then sends the suggestions to the user's device, where the user can confirm them.

[0361] Input: Generated Ad

[0362] Output: Tailored ads and matching information

[0363] Specific behavior:

[0364] 1. The server tailors the generated ads based on the user's emotions.

[0365] 2. Send the advertisement to the user's device through the interface.

[0366] 3. Using a matching algorithm to identify other users with similar interests.

[0367] 4. Share advertising information relevant to the user.

[0368] Step 5:

[0369] Users provide feedback on proposed ads and send it to the server from their devices. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of ad generation.

[0370] Input: User feedback

[0371] Output: Improved generative AI models and emotion engines

[0372] Specific behavior:

[0373] 1. The user submits feedback on the ad through the device interface.

[0374] 2. The server receives and analyzes the feedback.

[0375] 3. Based on the analysis results, the generative AI model and emotion engine are retrained.

[0376] 4. Reflect the learning results and improve the accuracy of the system.

[0377] 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.

[0378] 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.

[0379] 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.

[0380] [Second embodiment]

[0381] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0382] 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.

[0383] 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).

[0384] 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.

[0385] 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.

[0386] 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).

[0387] 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.

[0388] 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.

[0389] 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.

[0390] 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.

[0391] 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.

[0392] 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."

[0393] The system of this invention receives chat logs from users, analyzes the received chat logs, generates natural conversations using a generative AI model, and suggests them to the user. It also includes a means for automatically matching compatible conversation partners and continuously improving the system based on user feedback.

[0394] Receive chat logs from users

[0395] The user selects the chat log they want to upload using their device. The selected chat log is sent to the server as an HTTP POST request by the device. The server waits for the incoming request and receives the chat log file. The received file is saved in a specified directory on the server.

[0396] Chat log analysis

[0397] The server reads the saved chat log file and analyzes the contents line by line. During the analysis, information such as the message content, sender, and date and time of sending is extracted. This extracted data is then filtered out and converted into a format that is easy for the generative AI model to handle.

[0398] Generating natural conversations using generative AI models

[0399] The server loads and initializes a pre-trained generative AI model. The converted dataset is input to the generative AI model, which generates new, natural-sounding conversations based on past conversation patterns. The generated conversations are output in text format.

[0400] Conversation suggestions and matching compatible conversation partners

[0401] The server proposes the generated conversations to the user. These proposals include conversations with other users who are deemed to be compatible. Compatible partners are selected by comparing the user's past conversation patterns with those of other users. The proposals are sent to the user's device, and the user confirms their contents.

[0402] Gathering feedback and improving the system

[0403] The user provides feedback on the proposed conversation content, which is then sent from the device to the server, which analyzes the received feedback and retrains and refines the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0404] Specific examples

[0405] For example, if User A uploads a chat log exported from the LINE app to the system, the server receives and analyzes this chat log. If there were many conversations about "movies" in past chats, the generative AI model will suggest to User A that they talk about movies they've recently seen. If another User B is also interested in movies, the server can suggest matching User A and User B. User A can provide feedback on the suggested conversation, allowing the system to provide even more accurate conversation suggestions.

[0406] In this way, users can carry on natural conversations without any hassle, and effective communication can be achieved.

[0407] The processing flow will be explained below.

[0408] Step 1:

[0409] The user uses the terminal interface to select the chat log file to be uploaded.

[0410] Step 2:

[0411] The device sends the selected chat log file to the server as an HTTP POST request.

[0412] Step 3:

[0413] The server receives the chat log file sent from the device and saves it in the specified directory.

[0414] Step 4:

[0415] The server reads the saved chat log file and parses the contents line by line.

[0416] Step 5:

[0417] The server extracts information such as message content, sender, and sent date and time from the analyzed chat log.

[0418] Step 6:

[0419] The server removes unnecessary information and converts the extracted data into a format suitable for the generative AI model.

[0420] Step 7:

[0421] The server loads and initializes the pre-trained generative AI model.

[0422] Step 8:

[0423] The server inputs the converted data set into a generative AI model to analyze past conversation patterns.

[0424] Step 9:

[0425] The server uses the generative AI model to generate new, natural-sounding conversations.

[0426] Step 10:

[0427] The server outputs the generated conversation in text format and suggests it to the user.

[0428] Step 11:

[0429] The server compares the user's past conversation patterns with those of other users and automatically matches them with compatible conversation partners.

[0430] Step 12:

[0431] The server transmits information about conversation partners who are compatible with the proposed conversation to the user's terminal.

[0432] Step 13:

[0433] The user checks the proposed conversation content on the terminal and provides feedback if necessary.

[0434] Step 14:

[0435] The terminal sends the user's feedback to the server.

[0436] Step 15:

[0437] The server analyzes the feedback received and retrains and refines the generative AI model.

[0438] Step 16:

[0439] The server applies the improved generative AI model to future conversation generation.

[0440] Example 1

[0441] 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."

[0442] Conventional chat systems have not been able to generate natural conversations between users or match compatible conversation partners, making it difficult to achieve effective communication. Furthermore, the accuracy of the system has not been improved sufficiently because continuous improvements based on user feedback have not been made.

[0443] 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.

[0444] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and transmission date and time, means for using a natural language processing model to generate natural conversation based on the analysis results, means for suggesting the generated conversation to the user, means for selecting a compatible conversation partner based on the user's past conversation patterns, and means for receiving user feedback on the suggested conversation content and analyzing the feedback to retrain and improve the natural language processing model. This allows users to enjoy natural conversation and communicate effectively with compatible conversation partners. Furthermore, since the system is continuously improved based on the feedback, the accuracy of conversation generation can continue to improve.

[0445] "User" means an individual or organization that uses the system to upload chat logs.

[0446] A "chat log" is a data file that records messages exchanged between users.

[0447] A "server" is a computer system that receives requests from users and analyzes, stores, and processes chat logs.

[0448] "Receiving means" refers to the process and technical elements for receiving chat logs from users.

[0449] The "analysis means" is a process for analyzing the contents of the received chat log and extracting the message contents, sender, and sending date and time.

[0450] A "natural language processing model" is a model that uses AI technology to understand and generate natural human language.

[0451] The "generation means" is a process for generating new conversations using a natural language processing model based on the analysis results.

[0452] The "suggestion means" refers to the process and technical elements for suggesting the generated conversation to the user.

[0453] The "compatibility selection means" is a process for analyzing the user's past conversation patterns and selecting a conversation partner with good compatibility.

[0454] "Feedback" refers to the evaluation or opinion that a user provides regarding the proposed conversation content.

[0455] "Analysis means (feedback)" is the process of analyzing user feedback and reflecting it in improving the accuracy of the system.

[0456] The "restudy and refinement" means a process for retraining and refining a natural language processing model based on received feedback.

[0457] The system according to the present invention receives chat logs from users, analyzes them, generates natural conversations, makes suggestions, and improves models based on the feedback. How this system is implemented will be described below in detail.

[0458] Receive chat logs from users

[0459] The user accesses the system from their device and selects a chat log file. For example, they can use a web browser or a dedicated mobile app to click the "Select File" button and select a chat history file from LINE or another service. The device then sends the selected chat log file to the server as an HTTP POST request. The server receives the HTTP request and saves the chat log file in the " / uploads / chatlogs / " directory.

[0460] Chat log analysis

[0461] The server reads the saved chat log file. It parses the file line by line and extracts information such as the message content, sender, and date and time of sending. Using regular expressions and pattern matching, it extracts the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello." The extracted information is then converted into a format that is easy for the generative AI model to handle. For example, it can be formatted as "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0462] Generating natural conversations using generative AI models

[0463] The server loads and initializes a pre-trained generative AI model (e.g., using TensorFlow or PyTorch). It inputs a formatted dataset into the model and performs inference to generate new, natural-sounding conversations. The generated conversations are output in text format. For example, the model might generate a new message such as, "What movie have you seen recently?"

[0464] Conversation suggestions and matching compatible conversation partners

[0465] The server then incorporates the generated conversation into a template and formats it in HTML or JSON to propose it to the user. Furthermore, the server reads the user's past conversation patterns from the server's database and analyzes their compatibility with other users. A matching algorithm is used to evaluate similarities and differences and select appropriate conversation partners. For example, it might suggest to user A, "Let's talk about movies you've seen recently," and match them with user B, who also likes movies. This suggestion is sent to the user's device as an HTTP response, and the user can view the content in their browser or app.

[0466] Gathering feedback and improving the system

[0467] The user provides feedback on the suggested conversation content. For example, they can enter an evaluation such as "helpful" or "not helpful" using a dedicated button. This feedback is sent to the server in JSON format. The server receives the feedback data, stores it in a database, and analyzes it to reflect in the retraining of the model. Retraining improves the accuracy of the generative AI model, and more appropriate conversation suggestions will be made the next time the system is used.

[0468] Prompt Sentence Examples

[0469] "How do I upload my LINE chat logs and receive new conversation suggestions?"

[0470] In this way, the system can continuously provide users with natural conversations and achieve effective communication with compatible partners. Furthermore, by incorporating user feedback, the system's accuracy can be continuously improved.

[0471] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0472] Step 1: Receive chat logs from users

[0473] A user accesses the system from his / her terminal and selects a chat log file.

[0474] Specifically, click the "Select File" button on your browser or dedicated app and select a chat history file such as LINE.

[0475] The terminal sends the selected chat log file to the server as an HTTP POST request.

[0476] The server receives the HTTP request and saves the chat log files in the " / uploads / chatlogs / " directory.

[0477] Input: Chat log file selected by the user

[0478] Output: Chat log file saved on the server

[0479] Step 2: Analyzing the chat logs

[0480] The server reads the saved chat log file.

[0481] Specifically, it analyzes the file line by line and extracts information such as the message content, sender, and date and time of sending.

[0482] Using regular expressions and pattern matching, extract the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello."

[0483] The extracted information is converted into a format that is easy for the generative AI model to handle, for example, "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0484] Input: Chat log file stored on the server

[0485] Output: Formatted chat data

[0486] Step 3: Generating natural conversation using a generative AI model

[0487] The server loads and initializes the pre-trained generative AI model.

[0488] Specifically, the generative AI model is loaded from disk into memory using libraries such as TensorFlow or PyTorch.

[0489] The formatted dataset is fed into the model for inference, which generates new, natural-sounding conversations. For example, the model generates new messages such as, "What movie have you seen recently?"

[0490] The generated conversation is output in text format.

[0491] Input: Formatted chat data

[0492] Output: The new conversational text generated by the generative AI model

[0493] Step 4: Conversation suggestions and matching with compatible conversation partners

[0494] The server proposes the generated conversations to the user.

[0495] Specifically, the generated conversation is incorporated into a template, configured in HTML or JSON format, and an HTTP response is prepared for sending to the user.

[0496] Furthermore, the server reads the user's past conversation patterns from a database and analyzes the similarities and differences between them and other users.

[0497] A matching algorithm is used to select an appropriate conversation partner. For example, if user A is asked to talk about a movie they recently saw, the system will match them with user B, who also likes movies.

[0498] The suggestions are sent to the user's device, and the user can view them in their browser or app.

[0499] Input: Generated conversation text, user's past conversation data

[0500] Output: Conversation suggestions for the user and selection of compatible conversation partners

[0501] Step 5: Gather feedback and refine the system

[0502] The user provides feedback on the proposed conversation content.

[0503] Specifically, the user selects a rating such as "useful" or "not useful" and clicks a dedicated button.

[0504] The device sends the feedback data in JSON format to the server as a POST request.

[0505] The server stores the received feedback in a database, analyzes it, and reflects it in the retraining of the model.

[0506] Re-learning improves the accuracy of the generative AI model, allowing for more appropriate conversation suggestions the next time it is used.

[0507] Input: User feedback data

[0508] Output: Highly accurate conversation suggestions from an improved generative AI model

[0509] (Application example 1)

[0510] 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."

[0511] Modern communication tools make it difficult for users to find content and topics that interest them. Finding compatible conversation partners and generating natural conversations based on topics of interest is also challenging, especially for busy users. Furthermore, there is a lack of feedback mechanisms to improve the quality of generated content and conversations. This leaves users with insufficient support for effective communication.

[0512] 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.

[0513] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, senders, and sending dates and times, means for using a generative AI model to generate natural conversations and content based on the analysis results, and means for suggesting the generated conversations and content to users. This allows users to easily obtain natural conversations and content based on topics of interest and engage in appropriate communication.

[0514] A "chat log" is text data of the conversation content that a user sends and receives through a chat application.

[0515] "Analysis" is the process of extracting useful information from data and converting it into an understandable form.

[0516] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new conversations and content.

[0517] "Suggestion" is the act of presenting information or content that is deemed useful to the user.

[0518] "Content" refers to information or entertainment that may be of interest to users, such as movies, music, articles, etc.

[0519] A "compatible conversation partner" is another user with whom effective communication can be expected based on the user's interests and past conversation patterns.

[0520] "Feedback" refers to opinions and comments provided by users regarding a system or service.

[0521] "Retraining" is the process by which an AI model is retrained based on new data and feedback to improve its performance.

[0522] The system realizing this invention analyzes a user's chat log and uses a generative AI model to suggest natural conversations and content. A specific embodiment of this system will be described below.

[0523] Hardware and Software Use

[0524] Hardware:

[0525] Smartphone (OS: Android, iOS)

[0526] Cloud server (e.g. AWS EC2)

[0527] software:

[0528] Frontend: React Native (cross-platform mobile app development)

[0529] Backend: Node.js and Express (Building the API server)

[0530] Database: MongoDB (for storing chat logs)

[0531] AI model: GPT-4 (using OpenAI API)

[0532] System operation explanation

[0533] Receive chat logs:

[0534] The user selects the chat log they want to upload using their smartphone and sends it to the server via an HTTP POST request. The server receives it and saves it in the specified directory.

[0535] Chat log analysis:

[0536] The server reads the saved chat logs and parses them line by line, including the message content, sender, and date and time, and then removes unnecessary parts and converts them into a format that is easy for the generative AI model to handle.

[0537] Conversation and content generation using generative AI models:

[0538] The server loads and initializes a pre-trained generative AI model, inputs the converted dataset into the model, and generates new, natural-sounding conversations and content based on past conversation patterns, which are then output in text format.

[0539] Conversation suggestions and compatible conversation partner matching:

[0540] The server then proposes the generated conversations and content to the user. It compares the user's past conversation patterns with those of other users and also suggests compatible partners. The proposals are sent to the user's device, where the user can view them.

[0541] Gathering feedback and improving the system:

[0542] Users provide feedback on the suggestions, which is sent to the server and analyzed to retrain and refine the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0543] Specific examples

[0544] For example, if User A uploads past chat logs to the app, the server receives and analyzes these chat logs. If User A has had many conversations about "movies" in past chats, the generative AI model will generate content to "talk about movies you've recently seen" and suggest it to User A. At the same time, User B, who is also interested in movies, may also be matched.

[0545] Example prompt sentence:

[0546] User's past chat logs:

[0547] 1. "Movies have been interesting lately."

[0548] 2. "What kind of movie would you like?"

[0549] 3. "I like action. I like it when it's really epic."

[0550] Generates prompt:

[0551] "The user likes action movies. Tell the AI ​​model to talk about the top 10 latest action movies."

[0552] In this embodiment, the user can easily obtain natural conversations and content based on topics of interest, and can carry out appropriate communication.

[0553] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0554] Step 1:

[0555] A user selects a chat log using a smartphone and sends it to the server through an HTTP POST request. The input is the locally stored chat log file, and the output is the HTTP request sent to the server.

[0556] Step 2:

[0557] The server saves the received chat log in the specified directory. It writes the received data (chat log file) to the disk and gets the save path as output.

[0558] Step 3:

[0559] The server reads the saved chat log file and analyzes the contents line by line. This analysis extracts the message content, sender, and sending date and time. The input is the text data of the chat log, and the output is structured data (e.g., JSON format) as the analysis result.

[0560] Step 4:

[0561] The analyzed data is converted into a format that is easy for the generative AI model to handle. It is converted into data with a specific structure (e.g., context, entities, keywords, etc.) and output as a dataset to be input into the generative AI model.

[0562] Step 5:

[0563] The server loads a pre-trained generative AI model and inputs the converted dataset into the model. The generative AI model generates new, natural-sounding conversations and content based on past conversation patterns. The input is the initialization parameters of the generative AI model and the converted dataset, and the output is the text data of the generated conversation or content.

[0564] Step 6:

[0565] The generated conversation and content are sent to the user's device as a suggestion. The server returns the generated result as an HTTP response, and the user receives the suggestion on the application. The input is the generated text, and the output is the suggestion displayed on the user's device.

[0566] Step 7:

[0567] The user provides feedback on the proposed content. The feedback is sent from the user's terminal to the server as an HTTP POST request. The input is the user's feedback text, and the output is saved as feedback data to the server.

[0568] Step 8:

[0569] The server analyzes the received feedback and retrains and improves the generative AI model. It analyzes the feedback data to extract important information and uses this information to retrain the generative AI model. The input is the feedback data, and the output is an improved generative AI model.

[0570] The specific operations at each step are carried out in sequence by the cooperation of various components, such as the server, terminal, and user, and through processes such as data transmission and reception, analysis, generation, and feedback, effective conversations and content are provided to users.

[0571] 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.

[0572] The system of this invention receives chat logs from users, analyzes them, and extracts message content, sender, and sent date and time. It also uses a generative AI model to generate natural conversations and suggest them to the user. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions based on the user's emotions.

[0573] Receive chat logs from users

[0574] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[0575] Chat log analysis

[0576] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[0577] Generating natural conversations using generative AI models

[0578] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the server generates new, natural-sounding conversations. The generated conversations are output in text format.

[0579] Conversation suggestions and matching compatible conversation partners

[0580] The server adjusts the generated conversation based on the user's emotions and makes suggestions. The suggestions include conversations with other users who are deemed compatible. The emotion engine reflects appropriate conversation content based on the emotional information recognized by the server. The server then sends the suggestions to the user's device, where the user can confirm them.

[0581] Gathering feedback and improving the system

[0582] The user provides feedback on the proposed conversation content and sends it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0583] Specific examples

[0584] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes the chat log. The emotion engine also identifies the emotional state of the message content. For example, if past conversations have often focused on movies and the user appears to be enjoying themselves, the generative AI model will suggest, "Let's talk about an interesting movie you recently saw." If User B is also interested in movies and appears to be enjoying themselves, the server will suggest matching User A and User B, providing a conversation that is appropriate for both parties. By providing feedback on this suggestion, the system can reflect it in future conversation generation and matching, enabling it to provide even more accurate and natural conversations.

[0585] In this way, users can enjoy natural conversations without much effort and can achieve effective communication with people with whom they have good chemistry.

[0586] The processing flow will be explained below.

[0587] Step 1:

[0588] The user uses the terminal interface to select the chat log file to be uploaded.

[0589] Step 2:

[0590] The device sends the selected chat log file to the server as an HTTP POST request.

[0591] Step 3:

[0592] The server receives the chat log file sent from the device and saves it in the specified directory.

[0593] Step 4:

[0594] The server reads the saved chat log file and parses the contents line by line.

[0595] Step 5:

[0596] The server extracts message content, sender, and sent date and time from the analyzed chat log.

[0597] Step 6:

[0598] The server uses an emotion engine to analyze the emotional state of each message and add emotion information to the extracted data.

[0599] Step 7:

[0600] The server initializes the generative AI model and inputs the formatted dataset containing emotion information.

[0601] Step 8:

[0602] The server uses a generative AI model to generate new, natural conversations based on past conversation patterns and emotional information.

[0603] Step 9:

[0604] The server outputs the generated conversation in text format and adjusts it based on the user's emotions.

[0605] Step 10:

[0606] The server proposes tailored conversations to the user and compares them with other users' data to match compatible conversation partners.

[0607] Step 11:

[0608] The server identifies compatible conversation partners and sends suggestions to the user's device.

[0609] Step 12:

[0610] The user reviews the suggested conversation content on their device and adds feedback if necessary.

[0611] Step 13:

[0612] The terminal sends the user's feedback to the server.

[0613] Step 14:

[0614] The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine.

[0615] Step 15:

[0616] The server applies the improved generative AI model and emotion engine to future conversation generation.

[0617] Example 2

[0618] 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."

[0619] While conventional chat log analysis systems were able to extract message content, senders, and sending dates and times, they had problems in that they did not adequately generate natural conversations that took the user's emotional state into account or match compatible conversation partners. Furthermore, there was a lack of a way to effectively utilize user feedback on the generated conversation content to improve the system, making it difficult to consistently generate high-quality conversations.

[0620] 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.

[0621] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and sending date and time, means for identifying the user's emotional state using an emotion engine that estimates emotions, means for generating a format-converted dataset based on the analysis result and the identified emotional state, means for using a generative AI model that uses the generated dataset as input to generate natural conversations, means for adjusting and suggesting the generated conversations based on the user's emotions, and means for receiving feedback and improving the system. This makes it possible to generate natural conversations that reflect the user's emotions and match users with compatible conversation partners, and further allows the system to be continuously improved using user feedback.

[0622] A "chat log" is a data file that shows the history of messages that a user has sent and received in the past.

[0623] An "emotion engine" is software or algorithm that analyzes message content to infer a user's emotional state.

[0624] A "generative AI model" is an artificial intelligence model that generates natural conversations based on pre-trained data.

[0625] A "formatted dataset" is a collection of input data that has been converted into a format acceptable to a generative AI model based on the results of analysis and emotion estimation.

[0626] The "means for suggesting" is a method or function for presenting the generated conversation content to the user.

[0627] "Feedback" is the act of a user providing an evaluation or comment on the proposed conversation content.

[0628] "Retraining" is a new learning process that uses feedback to improve the performance of generative AI models and emotion engines.

[0629] "Matching means" refers to a method or function that automatically connects users who are determined to be compatible with each other.

[0630] This invention is a system that receives chat logs from users, analyzes the logs, and generates and suggests natural conversations. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback.

[0631] First, the user selects the chat log they want to upload using their device. The selected chat log is then sent from the device to the server. The server then saves the received chat log in a specified directory. This process uses standard HTTP POST requests and file I / O operations.

[0632] The server then reads the saved chat log file and analyzes it line by line. During the analysis process, it extracts the message content, sender, and date and time of sending. For example, this involves using regular expressions to extract this information from the text data. It then uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data includes the message content, sender, date and time, as well as emotional information.

[0633] Next, the server initializes a pre-trained generative AI model and inputs the converted dataset. Based on past conversation patterns and emotional information, the model generates new, natural-sounding conversations. For example, if the conversation revolves around "recent movies," the model might suggest, "Let's talk about some interesting movies you've seen recently."

[0634] The generated conversation is output in text format, and the server adjusts and suggests conversations based on the user's emotions. These suggestions include conversations with other users who are deemed compatible. Based on the emotional information recognized by the emotion engine, appropriate conversation content is generated and sent to the user's device.

[0635] Users can also provide feedback on the proposed conversation content and send it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0636] Specific examples

[0637] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes this chat log. The emotion engine also identifies the emotional state of the message content. For example, if the message "Hello" is evaluated as "positive," the generative AI model uses this information to generate a new message such as "How are you doing lately?"

[0638] Furthermore, if User B is also interested in movies and seems to enjoy them, the server will propose a match between User A and User B, providing a conversation suitable for both parties. Feedback on this proposal is provided, allowing the system to reflect this in future conversation generation and matching, enabling the system to provide even more accurate and natural conversations.

[0639] An example prompt is:

[0640] User A's emotional state: Happy

[0641] Past conversation topics: Movies, Travel, Music

[0642] Latest message: "What movie did you see last?"

[0643] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0644] Step 1:

[0645] The user selects the chat log they wish to upload on their device. This is done by using the device's file browser to find and select the specified chat log file. At this time, the system recognizes the selected file as the "selected chat log file."

[0646] Step 2:

[0647] The terminal sends the selected chat log file to the server. In this process, an HTTP POST request is used to send the file contents to the server. The input is the "selected chat log file" and the output is the "chat log file sent to the server."

[0648] Step 3:

[0649] The server saves the received chat log file in the specified directory. The file name is guaranteed to be unique using the user ID and timestamp. The input is the "chat log file sent to the server" and the output is the "saved chat log file."

[0650] Step 4:

[0651] The server reads and analyzes the saved chat log file line by line. Specifically, it uses regular expressions to extract the message content, sender, and sending date and time from each line. The input is the "saved chat log file," and the output is the "extracted message content, sender, and sending date and time."

[0652] Step 5:

[0653] The server uses an emotion engine to identify the user's emotional state from the analyzed message content. For example, it classifies the content into categories such as positive, negative, and neutral. The input is the extracted message content, and the output is the identified emotional state.

[0654] Step 6:

[0655] The server generates a format-converted dataset for the generative AI model based on the identified emotional state and analysis results. This is the process of formatting the data into a format that the generative AI model can understand. The input is the "analysis results and identified emotional state," and the output is the "format-converted dataset."

[0656] Step 7:

[0657] The server initializes the generative AI model and inputs the converted dataset, which generates natural-sounding conversations. The input is the converted dataset, and the output is the generated natural-sounding conversations.

[0658] Step 8:

[0659] The server adjusts the generated natural conversation based on the user's emotions and suggests it to the user. The input is the generated natural conversation and the identified emotional state, and the output is the suggested conversation content.

[0660] Step 9:

[0661] The user provides feedback on the proposed conversation content and sends it to the server via their device. The feedback can include ratings and specific comments. The input is the "proposed conversation content" and the output is "feedback from the user."

[0662] Step 10:

[0663] The server analyzes the received feedback and retrains and improves the generative AI model and emotion engine based on the content. The input is "feedback from the user" and the output is "an improved generative AI model and emotion engine."

[0664] These processing steps enable the system to generate natural conversations that reflect the user's emotions, match users with compatible conversation partners, and continuously improve the system.

[0665] (Application example 2)

[0666] 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."

[0667] Conventional chat analysis systems do not take into account users' emotional information when proposing advertisements, making it difficult to effectively present advertisements that match users' interests and needs. Furthermore, due to a lack of means to promote joint purchases and information exchange among users with similar interests, the relevance of advertisements tailored to users' preferences and the improvement of engagement are not sufficiently addressed.

[0668] 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.

[0669] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, sender, and sending date and time, means for using a generative AI model to generate natural conversations based on the analysis results, means for generating advertisements based on the content of the generated conversations and the user's emotional information and proposing them to the user, means for encouraging joint purchases and information exchanges with other users with similar interests based on the proposed advertisements, and means for receiving feedback on the advertisements from users and retraining and improving the generative AI model and the emotional engine. This makes it possible to propose advertisements that reflect the user's emotional state and interests, further increasing engagement among users with similar interests and maximizing the effectiveness of advertisements.

[0670] A "user" is a user who provides a chat log to the system and receives the analysis results and suggestions.

[0671] A "chat log" refers to messages and conversation history sent and received by a user, and is data that includes the content, sender, and date and time of sending.

[0672] "Emotion information" refers to the user's emotional state identified based on an analysis of the chat log, and includes positive, negative, neutral, and other emotions.

[0673] A "generative AI model" refers to a pre-trained artificial intelligence model, an algorithm for generating natural conversations and advertisements based on chat logs and emotional information.

[0674] "Advertisement" refers to marketing messages and product information that are generated based on the user's interests and emotional information and are proposed to the user.

[0675] "Suggestion" refers to the act of presenting generated conversations or advertisements to the user, which contain content that corresponds to the user's interests and emotions.

[0676] "Feedback" refers to user reactions and opinions on proposed conversations and advertisements, which are used to improve the system and retrain the generative AI model.

[0677] "Joint purchasing" refers to the act of purchasing products in collaboration with other users who share similar interests, and includes means by which the system encourages this.

[0678] "Information exchange" refers to the sharing of knowledge and opinions among other users with similar interests, and includes the means by which the system facilitates this.

[0679] To implement this invention, it is necessary to build a system that analyzes users' chat logs, generates optimal advertisements based on their emotional information, and proposes them to users. This system is composed of hardware and software such as a server, user terminals, a generative AI model, an emotion engine, and a feedback database.

[0680] The server receives and analyzes chat logs sent from users' devices. Specifically, it uses a natural language processing library (e.g., NLTK) and a sentiment analysis engine (e.g., Microsoft Azure Text Analytics) to extract message content, sender, and date and time from the chat logs. This also identifies the user's sentiment information.

[0681] The analyzed data is then fed into a generative AI model (e.g., OpenAI GPT-4), which then generates optimal ads based on past conversation patterns and emotional information. This generation process ensures that the ads include content that reflects the user's interests and emotions.

[0682] The generated advertisements are then presented to the user's device, where they can be viewed using a smartphone or smart glasses. The system also matches users with similar interests, facilitating joint purchases and information exchange.

[0683] User feedback on ads is sent to the server through a feedback collection system. The server analyzes this feedback and retrains and refines the generative AI model and emotion engine. A database (e.g., MySQL) is used for this feedback process.

[0684] As a specific example, if a user uploads a chat log about "movies" to the system, the sentiment analysis engine will identify the user's positive emotions. The generative AI model will then refer to past conversation patterns about "movies" to generate ads for "interesting movies they've recently seen." These ads will be suggested to the user, encouraging them to share information and make joint purchases with other users who are interested in movies.

[0685] Here are some example prompts to input to a generative AI model:

[0686] "User conversation and sentiment: [{'content': 'I like movies', 'sender': 'user1', 'timestamp': '2023-10-24T14:53:00', 'sentiment': {'neg': 0.0, 'neu': 0.3, 'pos': 0.7, 'compound': 0.7}}]. Generate the best ad."

[0687] This allows for advertising suggestions that reflect a user's emotional state and interests, further increasing engagement between users with similar interests.

[0688] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0689] Step 1:

[0690] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[0691] Input: Chat log (data sent from the user's device)

[0692] Output: Saved chat log file (stored in a directory on the server)

[0693] Specific behavior:

[0694] 1. The user selects the chat log file through the terminal interface.

[0695] 2. The device sends the selected file to the server.

[0696] 3. The server saves the received file in the specified directory.

[0697] Step 2:

[0698] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[0699] Input: Saved chat log file

[0700] Output: Analysis results (message content, sender, sending date and time, emotional information)

[0701] Specific behavior:

[0702] 1. The server reads the chat log file.

[0703] 2. Use a natural language processing library (e.g., NLTK) to extract the message content, sender, and date and time.

[0704] 3. Identify the sentiment of each message using a sentiment analysis engine (e.g., Microsoft Azure Text Analytics).

[0705] 4. Create a dataset to input the generative AI model.

[0706] Step 3:

[0707] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the model generates new, natural-looking ads. The generated ads are output in text format.

[0708] Input: Analysis result dataset (including emotion information)

[0709] Output: Generated advertisement (text format)

[0710] Specific behavior:

[0711] 1. The server initializes a generative AI model (e.g., OpenAI GPT-4) and inputs the dataset.

[0712] 2. The generative AI model generates ads based on the input data.

[0713] 3. Create the ad copy generated as output.

[0714] Step 4:

[0715] The server then adjusts the generated advertisements based on the user's emotions and suggests them to the user. It also matches users with similar interests and shares the advertisements with them. The server then sends the suggestions to the user's device, where the user can confirm them.

[0716] Input: Generated Ad

[0717] Output: Tailored ads and matching information

[0718] Specific behavior:

[0719] 1. The server tailors the generated ads based on the user's emotions.

[0720] 2. Send the advertisement to the user's device through the interface.

[0721] 3. Using a matching algorithm to identify other users with similar interests.

[0722] 4. Share advertising information relevant to the user.

[0723] Step 5:

[0724] Users provide feedback on proposed ads and send it to the server from their devices. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of ad generation.

[0725] Input: User feedback

[0726] Output: Improved generative AI models and emotion engines

[0727] Specific behavior:

[0728] 1. The user submits feedback on the ad through the device interface.

[0729] 2. The server receives and analyzes the feedback.

[0730] 3. Based on the analysis results, the generative AI model and emotion engine are retrained.

[0731] 4. Reflect the learning results and improve the accuracy of the system.

[0732] 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.

[0733] 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.

[0734] 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.

[0735] [Third embodiment]

[0736] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0737] 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.

[0738] 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).

[0739] 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.

[0740] 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.

[0741] 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).

[0742] 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.

[0743] 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.

[0744] 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.

[0745] 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.

[0746] 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.

[0747] 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."

[0748] The system of this invention receives chat logs from users, analyzes the received chat logs, generates natural conversations using a generative AI model, and suggests them to the user. It also includes a means for automatically matching compatible conversation partners and continuously improving the system based on user feedback.

[0749] Receive chat logs from users

[0750] The user selects the chat log they want to upload using their device. The selected chat log is sent to the server as an HTTP POST request by the device. The server waits for the incoming request and receives the chat log file. The received file is saved in a specified directory on the server.

[0751] Chat log analysis

[0752] The server reads the saved chat log file and analyzes the contents line by line. During the analysis, information such as the message content, sender, and date and time of sending is extracted. This extracted data is then filtered out and converted into a format that is easy for the generative AI model to handle.

[0753] Generating natural conversations using generative AI models

[0754] The server loads and initializes a pre-trained generative AI model. The converted dataset is input to the generative AI model, which generates new, natural-sounding conversations based on past conversation patterns. The generated conversations are output in text format.

[0755] Conversation suggestions and matching compatible conversation partners

[0756] The server proposes the generated conversations to the user. These proposals include conversations with other users who are deemed to be compatible. Compatible partners are selected by comparing the user's past conversation patterns with those of other users. The proposals are sent to the user's device, and the user confirms their contents.

[0757] Gathering feedback and improving the system

[0758] The user provides feedback on the proposed conversation content, which is then sent from the device to the server, which analyzes the received feedback and retrains and refines the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0759] Specific examples

[0760] For example, if User A uploads a chat log exported from the LINE app to the system, the server receives and analyzes this chat log. If there were many conversations about "movies" in past chats, the generative AI model will suggest to User A that they talk about movies they've recently seen. If another User B is also interested in movies, the server can suggest matching User A and User B. User A can provide feedback on the suggested conversation, allowing the system to provide even more accurate conversation suggestions.

[0761] In this way, users can carry on natural conversations without any hassle, and effective communication can be achieved.

[0762] The processing flow will be explained below.

[0763] Step 1:

[0764] The user uses the terminal interface to select the chat log file to be uploaded.

[0765] Step 2:

[0766] The device sends the selected chat log file to the server as an HTTP POST request.

[0767] Step 3:

[0768] The server receives the chat log file sent from the device and saves it in the specified directory.

[0769] Step 4:

[0770] The server reads the saved chat log file and parses the contents line by line.

[0771] Step 5:

[0772] The server extracts information such as message content, sender, and sent date and time from the analyzed chat log.

[0773] Step 6:

[0774] The server removes unnecessary information and converts the extracted data into a format suitable for the generative AI model.

[0775] Step 7:

[0776] The server loads and initializes the pre-trained generative AI model.

[0777] Step 8:

[0778] The server inputs the converted data set into a generative AI model to analyze past conversation patterns.

[0779] Step 9:

[0780] The server uses the generative AI model to generate new, natural-sounding conversations.

[0781] Step 10:

[0782] The server outputs the generated conversation in text format and suggests it to the user.

[0783] Step 11:

[0784] The server compares the user's past conversation patterns with those of other users and automatically matches them with compatible conversation partners.

[0785] Step 12:

[0786] The server transmits information about conversation partners who are compatible with the proposed conversation to the user's terminal.

[0787] Step 13:

[0788] The user checks the proposed conversation content on the terminal and provides feedback if necessary.

[0789] Step 14:

[0790] The terminal sends the user's feedback to the server.

[0791] Step 15:

[0792] The server analyzes the feedback received and retrains and refines the generative AI model.

[0793] Step 16:

[0794] The server applies the improved generative AI model to future conversation generation.

[0795] Example 1

[0796] 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."

[0797] Conventional chat systems have not been able to generate natural conversations between users or match compatible conversation partners, making it difficult to achieve effective communication. Furthermore, the accuracy of the system has not been improved sufficiently because continuous improvements based on user feedback have not been made.

[0798] 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.

[0799] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and transmission date and time, means for using a natural language processing model to generate natural conversation based on the analysis results, means for suggesting the generated conversation to the user, means for selecting a compatible conversation partner based on the user's past conversation patterns, and means for receiving user feedback on the suggested conversation content and analyzing the feedback to retrain and improve the natural language processing model. This allows users to enjoy natural conversation and communicate effectively with compatible conversation partners. Furthermore, since the system is continuously improved based on the feedback, the accuracy of conversation generation can continue to improve.

[0800] "User" means an individual or organization that uses the system to upload chat logs.

[0801] A "chat log" is a data file that records messages exchanged between users.

[0802] A "server" is a computer system that receives requests from users and analyzes, stores, and processes chat logs.

[0803] "Receiving means" refers to the process and technical elements for receiving chat logs from users.

[0804] The "analysis means" is a process for analyzing the contents of the received chat log and extracting the message contents, sender, and sending date and time.

[0805] A "natural language processing model" is a model that uses AI technology to understand and generate natural human language.

[0806] The "generation means" is a process for generating new conversations using a natural language processing model based on the analysis results.

[0807] The "suggestion means" refers to the process and technical elements for suggesting the generated conversation to the user.

[0808] The "compatibility selection means" is a process for analyzing the user's past conversation patterns and selecting a conversation partner with good compatibility.

[0809] "Feedback" refers to the evaluation or opinion that a user provides regarding the proposed conversation content.

[0810] "Analysis means (feedback)" is the process of analyzing user feedback and reflecting it in improving the accuracy of the system.

[0811] The "restudy and refinement" means a process for retraining and refining a natural language processing model based on received feedback.

[0812] The system according to the present invention receives chat logs from users, analyzes them, generates natural conversations, makes suggestions, and improves models based on the feedback. How this system is implemented will be described below in detail.

[0813] Receive chat logs from users

[0814] The user accesses the system from their device and selects a chat log file. For example, they can use a web browser or a dedicated mobile app to click the "Select File" button and select a chat history file from LINE or another service. The device then sends the selected chat log file to the server as an HTTP POST request. The server receives the HTTP request and saves the chat log file in the " / uploads / chatlogs / " directory.

[0815] Chat log analysis

[0816] The server reads the saved chat log file. It parses the file line by line and extracts information such as the message content, sender, and date and time of sending. Using regular expressions and pattern matching, it extracts the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello." The extracted information is then converted into a format that is easy for the generative AI model to handle. For example, it can be formatted as "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0817] Generating natural conversations using generative AI models

[0818] The server loads and initializes a pre-trained generative AI model (e.g., using TensorFlow or PyTorch). It inputs a formatted dataset into the model and performs inference to generate new, natural-sounding conversations. The generated conversations are output in text format. For example, the model might generate a new message such as, "What movie have you seen recently?"

[0819] Conversation suggestions and matching compatible conversation partners

[0820] The server then incorporates the generated conversation into a template and formats it in HTML or JSON to propose it to the user. Furthermore, the server reads the user's past conversation patterns from the server's database and analyzes their compatibility with other users. A matching algorithm is used to evaluate similarities and differences and select appropriate conversation partners. For example, it might suggest to user A, "Let's talk about movies you've seen recently," and match them with user B, who also likes movies. This suggestion is sent to the user's device as an HTTP response, and the user can view the content in their browser or app.

[0821] Gathering feedback and improving the system

[0822] The user provides feedback on the suggested conversation content. For example, they can enter an evaluation such as "helpful" or "not helpful" using a dedicated button. This feedback is sent to the server in JSON format. The server receives the feedback data, stores it in a database, and analyzes it to reflect in the retraining of the model. Retraining improves the accuracy of the generative AI model, and more appropriate conversation suggestions will be made the next time the system is used.

[0823] Prompt Sentence Examples

[0824] "How do I upload my LINE chat logs and receive new conversation suggestions?"

[0825] In this way, the system can continuously provide users with natural conversations and achieve effective communication with compatible partners. Furthermore, by incorporating user feedback, the system's accuracy can be continuously improved.

[0826] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0827] Step 1: Receive chat logs from users

[0828] A user accesses the system from his / her terminal and selects a chat log file.

[0829] Specifically, click the "Select File" button on your browser or dedicated app and select a chat history file such as LINE.

[0830] The terminal sends the selected chat log file to the server as an HTTP POST request.

[0831] The server receives the HTTP request and saves the chat log files in the " / uploads / chatlogs / " directory.

[0832] Input: Chat log file selected by the user

[0833] Output: Chat log file saved on the server

[0834] Step 2: Analyzing the chat logs

[0835] The server reads the saved chat log file.

[0836] Specifically, it analyzes the file line by line and extracts information such as the message content, sender, and date and time of sending.

[0837] Using regular expressions and pattern matching, extract the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello."

[0838] The extracted information is converted into a format that is easy for the generative AI model to handle, for example, "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[0839] Input: Chat log file stored on the server

[0840] Output: Formatted chat data

[0841] Step 3: Generating natural conversation using a generative AI model

[0842] The server loads and initializes the pre-trained generative AI model.

[0843] Specifically, the generative AI model is loaded from disk into memory using libraries such as TensorFlow or PyTorch.

[0844] The formatted dataset is fed into the model for inference, which generates new, natural-sounding conversations. For example, the model generates new messages such as, "What movie have you seen recently?"

[0845] The generated conversation is output in text format.

[0846] Input: Formatted chat data

[0847] Output: The new conversational text generated by the generative AI model

[0848] Step 4: Conversation suggestions and matching with compatible conversation partners

[0849] The server proposes the generated conversations to the user.

[0850] Specifically, the generated conversation is incorporated into a template, configured in HTML or JSON format, and an HTTP response is prepared for sending to the user.

[0851] Furthermore, the server reads the user's past conversation patterns from a database and analyzes the similarities and differences between them and other users.

[0852] A matching algorithm is used to select an appropriate conversation partner. For example, if user A is asked to talk about a movie they recently saw, the system will match them with user B, who also likes movies.

[0853] The suggestions are sent to the user's device, and the user can view them in their browser or app.

[0854] Input: Generated conversation text, user's past conversation data

[0855] Output: Conversation suggestions for the user and selection of compatible conversation partners

[0856] Step 5: Gather feedback and refine the system

[0857] The user provides feedback on the proposed conversation content.

[0858] Specifically, the user selects a rating such as "useful" or "not useful" and clicks a dedicated button.

[0859] The device sends the feedback data in JSON format to the server as a POST request.

[0860] The server stores the received feedback in a database, analyzes it, and reflects it in the retraining of the model.

[0861] Re-learning improves the accuracy of the generative AI model, allowing for more appropriate conversation suggestions the next time it is used.

[0862] Input: User feedback data

[0863] Output: Highly accurate conversation suggestions from an improved generative AI model

[0864] (Application example 1)

[0865] 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."

[0866] Modern communication tools make it difficult for users to find content and topics that interest them. Finding compatible conversation partners and generating natural conversations based on topics of interest is also challenging, especially for busy users. Furthermore, there is a lack of feedback mechanisms to improve the quality of generated content and conversations. This leaves users with insufficient support for effective communication.

[0867] 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.

[0868] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, senders, and sending dates and times, means for using a generative AI model to generate natural conversations and content based on the analysis results, and means for suggesting the generated conversations and content to users. This allows users to easily obtain natural conversations and content based on topics of interest and engage in appropriate communication.

[0869] A "chat log" is text data of the conversation content that a user sends and receives through a chat application.

[0870] "Analysis" is the process of extracting useful information from data and converting it into an understandable form.

[0871] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new conversations and content.

[0872] "Suggestion" is the act of presenting information or content that is deemed useful to the user.

[0873] "Content" refers to information or entertainment that may be of interest to users, such as movies, music, articles, etc.

[0874] A "compatible conversation partner" is another user with whom effective communication can be expected based on the user's interests and past conversation patterns.

[0875] "Feedback" refers to opinions and comments provided by users regarding a system or service.

[0876] "Retraining" is the process by which an AI model is retrained based on new data and feedback to improve its performance.

[0877] The system realizing this invention analyzes a user's chat log and uses a generative AI model to suggest natural conversations and content. A specific embodiment of this system will be described below.

[0878] Hardware and Software Use

[0879] Hardware:

[0880] Smartphone (OS: Android, iOS)

[0881] Cloud server (e.g. AWS EC2)

[0882] software:

[0883] Frontend: React Native (cross-platform mobile app development)

[0884] Backend: Node.js and Express (Building the API server)

[0885] Database: MongoDB (for storing chat logs)

[0886] AI model: GPT-4 (using OpenAI API)

[0887] System operation explanation

[0888] Receive chat logs:

[0889] The user selects the chat log they want to upload using their smartphone and sends it to the server via an HTTP POST request. The server receives it and saves it in the specified directory.

[0890] Chat log analysis:

[0891] The server reads the saved chat logs and parses them line by line, including the message content, sender, and date and time, and then removes unnecessary parts and converts them into a format that is easy for the generative AI model to handle.

[0892] Conversation and content generation using generative AI models:

[0893] The server loads and initializes a pre-trained generative AI model, inputs the converted dataset into the model, and generates new, natural-sounding conversations and content based on past conversation patterns, which are then output in text format.

[0894] Conversation suggestions and compatible conversation partner matching:

[0895] The server then proposes the generated conversations and content to the user. It compares the user's past conversation patterns with those of other users and also suggests compatible partners. The proposals are sent to the user's device, where the user can view them.

[0896] Gathering feedback and improving the system:

[0897] Users provide feedback on the suggestions, which is sent to the server and analyzed to retrain and refine the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[0898] Specific examples

[0899] For example, if User A uploads past chat logs to the app, the server receives and analyzes these chat logs. If User A has had many conversations about "movies" in past chats, the generative AI model will generate content to "talk about movies you've recently seen" and suggest it to User A. At the same time, User B, who is also interested in movies, may also be matched.

[0900] Example prompt sentence:

[0901] User's past chat logs:

[0902] 1. "Movies have been interesting lately."

[0903] 2. "What kind of movie would you like?"

[0904] 3. "I like action. I like it when it's really epic."

[0905] Generates prompt:

[0906] "The user likes action movies. Tell the AI ​​model to talk about the top 10 latest action movies."

[0907] In this embodiment, the user can easily obtain natural conversations and content based on topics of interest, and can carry out appropriate communication.

[0908] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0909] Step 1:

[0910] A user selects a chat log using a smartphone and sends it to the server through an HTTP POST request. The input is the locally stored chat log file, and the output is the HTTP request sent to the server.

[0911] Step 2:

[0912] The server saves the received chat log in the specified directory. It writes the received data (chat log file) to the disk and gets the save path as output.

[0913] Step 3:

[0914] The server reads the saved chat log file and analyzes the contents line by line. This analysis extracts the message content, sender, and sending date and time. The input is the text data of the chat log, and the output is structured data (e.g., JSON format) as the analysis result.

[0915] Step 4:

[0916] The analyzed data is converted into a format that is easy for the generative AI model to handle. It is converted into data with a specific structure (e.g., context, entities, keywords, etc.) and output as a dataset to be input into the generative AI model.

[0917] Step 5:

[0918] The server loads a pre-trained generative AI model and inputs the converted dataset into the model. The generative AI model generates new, natural-sounding conversations and content based on past conversation patterns. The input is the initialization parameters of the generative AI model and the converted dataset, and the output is the text data of the generated conversation or content.

[0919] Step 6:

[0920] The generated conversation and content are sent to the user's device as a suggestion. The server returns the generated result as an HTTP response, and the user receives the suggestion on the application. The input is the generated text, and the output is the suggestion displayed on the user's device.

[0921] Step 7:

[0922] The user provides feedback on the proposed content. The feedback is sent from the user's terminal to the server as an HTTP POST request. The input is the user's feedback text, and the output is saved as feedback data to the server.

[0923] Step 8:

[0924] The server analyzes the received feedback and retrains and improves the generative AI model. It analyzes the feedback data to extract important information and uses this information to retrain the generative AI model. The input is the feedback data, and the output is an improved generative AI model.

[0925] The specific operations at each step are carried out in sequence by the cooperation of various components, such as the server, terminal, and user, and through processes such as data transmission and reception, analysis, generation, and feedback, effective conversations and content are provided to users.

[0926] 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.

[0927] The system of this invention receives chat logs from users, analyzes them, and extracts message content, sender, and sent date and time. It also uses a generative AI model to generate natural conversations and suggest them to the user. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions based on the user's emotions.

[0928] Receive chat logs from users

[0929] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[0930] Chat log analysis

[0931] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[0932] Generating natural conversations using generative AI models

[0933] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the server generates new, natural-sounding conversations. The generated conversations are output in text format.

[0934] Conversation suggestions and matching compatible conversation partners

[0935] The server adjusts the generated conversation based on the user's emotions and makes suggestions. The suggestions include conversations with other users who are deemed compatible. The emotion engine reflects appropriate conversation content based on the emotional information recognized by the server. The server then sends the suggestions to the user's device, where the user can confirm them.

[0936] Gathering feedback and improving the system

[0937] The user provides feedback on the proposed conversation content and sends it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0938] Specific examples

[0939] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes the chat log. The emotion engine also identifies the emotional state of the message content. For example, if past conversations have often focused on movies and the user appears to be enjoying themselves, the generative AI model will suggest, "Let's talk about an interesting movie you recently saw." If User B is also interested in movies and appears to be enjoying themselves, the server will suggest matching User A and User B, providing a conversation that is appropriate for both parties. By providing feedback on this suggestion, the system can reflect it in future conversation generation and matching, enabling it to provide even more accurate and natural conversations.

[0940] In this way, users can enjoy natural conversations without much effort and can achieve effective communication with people with whom they have good chemistry.

[0941] The processing flow will be explained below.

[0942] Step 1:

[0943] The user uses the terminal interface to select the chat log file to be uploaded.

[0944] Step 2:

[0945] The device sends the selected chat log file to the server as an HTTP POST request.

[0946] Step 3:

[0947] The server receives the chat log file sent from the device and saves it in the specified directory.

[0948] Step 4:

[0949] The server reads the saved chat log file and parses the contents line by line.

[0950] Step 5:

[0951] The server extracts message content, sender, and sent date and time from the analyzed chat log.

[0952] Step 6:

[0953] The server uses an emotion engine to analyze the emotional state of each message and add emotion information to the extracted data.

[0954] Step 7:

[0955] The server initializes the generative AI model and inputs the formatted dataset containing emotion information.

[0956] Step 8:

[0957] The server uses a generative AI model to generate new, natural conversations based on past conversation patterns and emotional information.

[0958] Step 9:

[0959] The server outputs the generated conversation in text format and adjusts it based on the user's emotions.

[0960] Step 10:

[0961] The server proposes tailored conversations to the user and compares them with other users' data to match compatible conversation partners.

[0962] Step 11:

[0963] The server identifies compatible conversation partners and sends suggestions to the user's device.

[0964] Step 12:

[0965] The user reviews the suggested conversation content on their device and adds feedback if necessary.

[0966] Step 13:

[0967] The terminal sends the user's feedback to the server.

[0968] Step 14:

[0969] The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine.

[0970] Step 15:

[0971] The server applies the improved generative AI model and emotion engine to future conversation generation.

[0972] Example 2

[0973] 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."

[0974] While conventional chat log analysis systems were able to extract message content, senders, and sending dates and times, they had problems in that they did not adequately generate natural conversations that took the user's emotional state into account or match compatible conversation partners. Furthermore, there was a lack of a way to effectively utilize user feedback on the generated conversation content to improve the system, making it difficult to consistently generate high-quality conversations.

[0975] 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.

[0976] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and sending date and time, means for identifying the user's emotional state using an emotion engine that estimates emotions, means for generating a format-converted dataset based on the analysis result and the identified emotional state, means for using a generative AI model that uses the generated dataset as input to generate natural conversations, means for adjusting and suggesting the generated conversations based on the user's emotions, and means for receiving feedback and improving the system. This makes it possible to generate natural conversations that reflect the user's emotions and match users with compatible conversation partners, and further allows the system to be continuously improved using user feedback.

[0977] A "chat log" is a data file that shows the history of messages that a user has sent and received in the past.

[0978] An "emotion engine" is software or algorithm that analyzes message content to infer a user's emotional state.

[0979] A "generative AI model" is an artificial intelligence model that generates natural conversations based on pre-trained data.

[0980] A "formatted dataset" is a collection of input data that has been converted into a format acceptable to a generative AI model based on the results of analysis and emotion estimation.

[0981] The "means for suggesting" is a method or function for presenting the generated conversation content to the user.

[0982] "Feedback" is the act of a user providing an evaluation or comment on the proposed conversation content.

[0983] "Retraining" is a new learning process that uses feedback to improve the performance of generative AI models and emotion engines.

[0984] "Matching means" refers to a method or function that automatically connects users who are determined to be compatible with each other.

[0985] This invention is a system that receives chat logs from users, analyzes the logs, and generates and suggests natural conversations. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback.

[0986] First, the user selects the chat log they want to upload using their device. The selected chat log is then sent from the device to the server. The server then saves the received chat log in a specified directory. This process uses standard HTTP POST requests and file I / O operations.

[0987] The server then reads the saved chat log file and analyzes it line by line. During the analysis process, it extracts the message content, sender, and date and time of sending. For example, this involves using regular expressions to extract this information from the text data. It then uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data includes the message content, sender, date and time, as well as emotional information.

[0988] Next, the server initializes a pre-trained generative AI model and inputs the converted dataset. Based on past conversation patterns and emotional information, the model generates new, natural-sounding conversations. For example, if the conversation revolves around "recent movies," the model might suggest, "Let's talk about some interesting movies you've seen recently."

[0989] The generated conversation is output in text format, and the server adjusts and suggests conversations based on the user's emotions. These suggestions include conversations with other users who are deemed compatible. Based on the emotional information recognized by the emotion engine, appropriate conversation content is generated and sent to the user's device.

[0990] Users can also provide feedback on the proposed conversation content and send it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[0991] Specific examples

[0992] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes this chat log. The emotion engine also identifies the emotional state of the message content. For example, if the message "Hello" is evaluated as "positive," the generative AI model uses this information to generate a new message such as "How are you doing lately?"

[0993] Furthermore, if User B is also interested in movies and seems to enjoy them, the server will propose a match between User A and User B, providing a conversation suitable for both parties. Feedback on this proposal is provided, allowing the system to reflect this in future conversation generation and matching, enabling the system to provide even more accurate and natural conversations.

[0994] An example prompt is:

[0995] User A's emotional state: Happy

[0996] Past conversation topics: Movies, Travel, Music

[0997] Latest message: "What movie did you see last?"

[0998] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0999] Step 1:

[1000] The user selects the chat log they wish to upload on their device. This is done by using the device's file browser to find and select the specified chat log file. At this time, the system recognizes the selected file as the "selected chat log file."

[1001] Step 2:

[1002] The terminal sends the selected chat log file to the server. In this process, an HTTP POST request is used to send the file contents to the server. The input is the "selected chat log file" and the output is the "chat log file sent to the server."

[1003] Step 3:

[1004] The server saves the received chat log file in the specified directory. The file name is guaranteed to be unique using the user ID and timestamp. The input is the "chat log file sent to the server" and the output is the "saved chat log file."

[1005] Step 4:

[1006] The server reads and analyzes the saved chat log file line by line. Specifically, it uses regular expressions to extract the message content, sender, and sending date and time from each line. The input is the "saved chat log file," and the output is the "extracted message content, sender, and sending date and time."

[1007] Step 5:

[1008] The server uses an emotion engine to identify the user's emotional state from the analyzed message content. For example, it classifies the content into categories such as positive, negative, and neutral. The input is the extracted message content, and the output is the identified emotional state.

[1009] Step 6:

[1010] The server generates a format-converted dataset for the generative AI model based on the identified emotional state and analysis results. This is the process of formatting the data into a format that the generative AI model can understand. The input is the "analysis results and identified emotional state," and the output is the "format-converted dataset."

[1011] Step 7:

[1012] The server initializes the generative AI model and inputs the converted dataset, which generates natural-sounding conversations. The input is the converted dataset, and the output is the generated natural-sounding conversations.

[1013] Step 8:

[1014] The server adjusts the generated natural conversation based on the user's emotions and suggests it to the user. The input is the generated natural conversation and the identified emotional state, and the output is the suggested conversation content.

[1015] Step 9:

[1016] The user provides feedback on the proposed conversation content and sends it to the server via their device. The feedback can include ratings and specific comments. The input is the "proposed conversation content" and the output is "feedback from the user."

[1017] Step 10:

[1018] The server analyzes the received feedback and retrains and improves the generative AI model and emotion engine based on the content. The input is "feedback from the user" and the output is "an improved generative AI model and emotion engine."

[1019] These processing steps enable the system to generate natural conversations that reflect the user's emotions, match users with compatible conversation partners, and continuously improve the system.

[1020] (Application example 2)

[1021] 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."

[1022] Conventional chat analysis systems do not take into account users' emotional information when proposing advertisements, making it difficult to effectively present advertisements that match users' interests and needs. Furthermore, due to a lack of means to promote joint purchases and information exchange among users with similar interests, the relevance of advertisements tailored to users' preferences and the improvement of engagement are not sufficiently addressed.

[1023] 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.

[1024] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, sender, and sending date and time, means for using a generative AI model to generate natural conversations based on the analysis results, means for generating advertisements based on the content of the generated conversations and the user's emotional information and proposing them to the user, means for encouraging joint purchases and information exchanges with other users with similar interests based on the proposed advertisements, and means for receiving feedback on the advertisements from users and retraining and improving the generative AI model and the emotional engine. This makes it possible to propose advertisements that reflect the user's emotional state and interests, further increasing engagement among users with similar interests and maximizing the effectiveness of advertisements.

[1025] A "user" is a user who provides a chat log to the system and receives the analysis results and suggestions.

[1026] A "chat log" refers to messages and conversation history sent and received by a user, and is data that includes the content, sender, and date and time of sending.

[1027] "Emotion information" refers to the user's emotional state identified based on an analysis of the chat log, and includes positive, negative, neutral, and other emotions.

[1028] A "generative AI model" refers to a pre-trained artificial intelligence model, an algorithm for generating natural conversations and advertisements based on chat logs and emotional information.

[1029] "Advertisement" refers to marketing messages and product information that are generated based on the user's interests and emotional information and are proposed to the user.

[1030] "Suggestion" refers to the act of presenting generated conversations or advertisements to the user, which contain content that corresponds to the user's interests and emotions.

[1031] "Feedback" refers to user reactions and opinions on proposed conversations and advertisements, which are used to improve the system and retrain the generative AI model.

[1032] "Joint purchasing" refers to the act of purchasing products in collaboration with other users who share similar interests, and includes means by which the system encourages this.

[1033] "Information exchange" refers to the sharing of knowledge and opinions among other users with similar interests, and includes the means by which the system facilitates this.

[1034] To implement this invention, it is necessary to build a system that analyzes users' chat logs, generates optimal advertisements based on their emotional information, and proposes them to users. This system is composed of hardware and software such as a server, user terminals, a generative AI model, an emotion engine, and a feedback database.

[1035] The server receives and analyzes chat logs sent from users' devices. Specifically, it uses a natural language processing library (e.g., NLTK) and a sentiment analysis engine (e.g., Microsoft Azure Text Analytics) to extract message content, sender, and date and time from the chat logs. This also identifies the user's sentiment information.

[1036] The analyzed data is then fed into a generative AI model (e.g., OpenAI GPT-4), which then generates optimal ads based on past conversation patterns and emotional information. This generation process ensures that the ads include content that reflects the user's interests and emotions.

[1037] The generated advertisements are then presented to the user's device, where they can be viewed using a smartphone or smart glasses. The system also matches users with similar interests, facilitating joint purchases and information exchange.

[1038] User feedback on ads is sent to the server through a feedback collection system. The server analyzes this feedback and retrains and refines the generative AI model and emotion engine. A database (e.g., MySQL) is used for this feedback process.

[1039] As a specific example, if a user uploads a chat log about "movies" to the system, the sentiment analysis engine will identify the user's positive emotions. The generative AI model will then refer to past conversation patterns about "movies" to generate ads for "interesting movies they've recently seen." These ads will be suggested to the user, encouraging them to share information and make joint purchases with other users who are interested in movies.

[1040] Here are some example prompts to input to a generative AI model:

[1041] "User conversation and sentiment: [{'content': 'I like movies', 'sender': 'user1', 'timestamp': '2023-10-24T14:53:00', 'sentiment': {'neg': 0.0, 'neu': 0.3, 'pos': 0.7, 'compound': 0.7}}]. Generate the best ad."

[1042] This allows for advertising suggestions that reflect a user's emotional state and interests, further increasing engagement between users with similar interests.

[1043] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1044] Step 1:

[1045] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[1046] Input: Chat log (data sent from the user's device)

[1047] Output: Saved chat log file (stored in a directory on the server)

[1048] Specific behavior:

[1049] 1. The user selects the chat log file through the terminal interface.

[1050] 2. The device sends the selected file to the server.

[1051] 3. The server saves the received file in the specified directory.

[1052] Step 2:

[1053] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[1054] Input: Saved chat log file

[1055] Output: Analysis results (message content, sender, sending date and time, emotional information)

[1056] Specific behavior:

[1057] 1. The server reads the chat log file.

[1058] 2. Use a natural language processing library (e.g., NLTK) to extract the message content, sender, and date and time.

[1059] 3. Identify the sentiment of each message using a sentiment analysis engine (e.g., Microsoft Azure Text Analytics).

[1060] 4. Create a dataset to input the generative AI model.

[1061] Step 3:

[1062] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the model generates new, natural-looking ads. The generated ads are output in text format.

[1063] Input: Analysis result dataset (including emotion information)

[1064] Output: Generated advertisement (text format)

[1065] Specific behavior:

[1066] 1. The server initializes a generative AI model (e.g., OpenAI GPT-4) and inputs the dataset.

[1067] 2. The generative AI model generates ads based on the input data.

[1068] 3. Create the ad copy generated as output.

[1069] Step 4:

[1070] The server then adjusts the generated advertisements based on the user's emotions and suggests them to the user. It also matches users with similar interests and shares the advertisements with them. The server then sends the suggestions to the user's device, where the user can confirm them.

[1071] Input: Generated Ad

[1072] Output: Tailored ads and matching information

[1073] Specific behavior:

[1074] 1. The server tailors the generated ads based on the user's emotions.

[1075] 2. Send the advertisement to the user's device through the interface.

[1076] 3. Using a matching algorithm to identify other users with similar interests.

[1077] 4. Share advertising information relevant to the user.

[1078] Step 5:

[1079] Users provide feedback on proposed ads and send it to the server from their devices. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of ad generation.

[1080] Input: User feedback

[1081] Output: Improved generative AI models and emotion engines

[1082] Specific behavior:

[1083] 1. The user submits feedback on the ad through the device interface.

[1084] 2. The server receives and analyzes the feedback.

[1085] 3. Based on the analysis results, the generative AI model and emotion engine are retrained.

[1086] 4. Reflect the learning results and improve the accuracy of the system.

[1087] 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.

[1088] 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.

[1089] 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.

[1090] [Fourth embodiment]

[1091] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1092] 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.

[1093] 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).

[1094] 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.

[1095] 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.

[1096] 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).

[1097] 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.

[1098] 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.

[1099] 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.

[1100] 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.

[1101] 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.

[1102] 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.

[1103] 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."

[1104] The system of this invention receives chat logs from users, analyzes the received chat logs, generates natural conversations using a generative AI model, and suggests them to the user. It also includes a means for automatically matching compatible conversation partners and continuously improving the system based on user feedback.

[1105] Receive chat logs from users

[1106] The user selects the chat log they want to upload using their device. The selected chat log is sent to the server as an HTTP POST request by the device. The server waits for the incoming request and receives the chat log file. The received file is saved in a specified directory on the server.

[1107] Chat log analysis

[1108] The server reads the saved chat log file and analyzes the contents line by line. During the analysis, information such as the message content, sender, and date and time of sending is extracted. This extracted data is then filtered out and converted into a format that is easy for the generative AI model to handle.

[1109] Generating natural conversations using generative AI models

[1110] The server loads and initializes a pre-trained generative AI model. The converted dataset is input to the generative AI model, which generates new, natural-sounding conversations based on past conversation patterns. The generated conversations are output in text format.

[1111] Conversation suggestions and matching compatible conversation partners

[1112] The server proposes the generated conversations to the user. These proposals include conversations with other users who are deemed to be compatible. Compatible partners are selected by comparing the user's past conversation patterns with those of other users. The proposals are sent to the user's device, and the user confirms their contents.

[1113] Gathering feedback and improving the system

[1114] The user provides feedback on the proposed conversation content, which is then sent from the device to the server, which analyzes the received feedback and retrains and refines the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[1115] Specific examples

[1116] For example, if User A uploads a chat log exported from the LINE app to the system, the server receives and analyzes this chat log. If there were many conversations about "movies" in past chats, the generative AI model will suggest to User A that they talk about movies they've recently seen. If another User B is also interested in movies, the server can suggest matching User A and User B. User A can provide feedback on the suggested conversation, allowing the system to provide even more accurate conversation suggestions.

[1117] In this way, users can carry on natural conversations without any hassle, and effective communication can be achieved.

[1118] The processing flow will be explained below.

[1119] Step 1:

[1120] The user uses the terminal interface to select the chat log file to be uploaded.

[1121] Step 2:

[1122] The device sends the selected chat log file to the server as an HTTP POST request.

[1123] Step 3:

[1124] The server receives the chat log file sent from the device and saves it in the specified directory.

[1125] Step 4:

[1126] The server reads the saved chat log file and parses the contents line by line.

[1127] Step 5:

[1128] The server extracts information such as message content, sender, and sent date and time from the analyzed chat log.

[1129] Step 6:

[1130] The server removes unnecessary information and converts the extracted data into a format suitable for the generative AI model.

[1131] Step 7:

[1132] The server loads and initializes the pre-trained generative AI model.

[1133] Step 8:

[1134] The server inputs the converted data set into a generative AI model to analyze past conversation patterns.

[1135] Step 9:

[1136] The server uses the generative AI model to generate new, natural-sounding conversations.

[1137] Step 10:

[1138] The server outputs the generated conversation in text format and suggests it to the user.

[1139] Step 11:

[1140] The server compares the user's past conversation patterns with those of other users and automatically matches them with compatible conversation partners.

[1141] Step 12:

[1142] The server transmits information about conversation partners who are compatible with the proposed conversation to the user's terminal.

[1143] Step 13:

[1144] The user checks the proposed conversation content on the terminal and provides feedback if necessary.

[1145] Step 14:

[1146] The terminal sends the user's feedback to the server.

[1147] Step 15:

[1148] The server analyzes the feedback received and retrains and refines the generative AI model.

[1149] Step 16:

[1150] The server applies the improved generative AI model to future conversation generation.

[1151] Example 1

[1152] 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."

[1153] Conventional chat systems have not been able to generate natural conversations between users or match compatible conversation partners, making it difficult to achieve effective communication. Furthermore, the accuracy of the system has not been improved sufficiently because continuous improvements based on user feedback have not been made.

[1154] 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.

[1155] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and transmission date and time, means for using a natural language processing model to generate natural conversation based on the analysis results, means for suggesting the generated conversation to the user, means for selecting a compatible conversation partner based on the user's past conversation patterns, and means for receiving user feedback on the suggested conversation content and analyzing the feedback to retrain and improve the natural language processing model. This allows users to enjoy natural conversation and communicate effectively with compatible conversation partners. Furthermore, since the system is continuously improved based on the feedback, the accuracy of conversation generation can continue to improve.

[1156] "User" means an individual or organization that uses the system to upload chat logs.

[1157] A "chat log" is a data file that records messages exchanged between users.

[1158] A "server" is a computer system that receives requests from users and analyzes, stores, and processes chat logs.

[1159] "Receiving means" refers to the process and technical elements for receiving chat logs from users.

[1160] The "analysis means" is a process for analyzing the contents of the received chat log and extracting the message contents, sender, and sending date and time.

[1161] A "natural language processing model" is a model that uses AI technology to understand and generate natural human language.

[1162] The "generation means" is a process for generating new conversations using a natural language processing model based on the analysis results.

[1163] The "suggestion means" refers to the process and technical elements for suggesting the generated conversation to the user.

[1164] The "compatibility selection means" is a process for analyzing the user's past conversation patterns and selecting a conversation partner with good compatibility.

[1165] "Feedback" refers to the evaluation or opinion that a user provides regarding the proposed conversation content.

[1166] "Analysis means (feedback)" is the process of analyzing user feedback and reflecting it in improving the accuracy of the system.

[1167] The "restudy and refinement" means a process for retraining and refining a natural language processing model based on received feedback.

[1168] The system according to the present invention receives chat logs from users, analyzes them, generates natural conversations, makes suggestions, and improves models based on the feedback. How this system is implemented will be described below in detail.

[1169] Receive chat logs from users

[1170] The user accesses the system from their device and selects a chat log file. For example, they can use a web browser or a dedicated mobile app to click the "Select File" button and select a chat history file from LINE or another service. The device then sends the selected chat log file to the server as an HTTP POST request. The server receives the HTTP request and saves the chat log file in the " / uploads / chatlogs / " directory.

[1171] Chat log analysis

[1172] The server reads the saved chat log file. It parses the file line by line and extracts information such as the message content, sender, and date and time of sending. Using regular expressions and pattern matching, it extracts the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello." The extracted information is then converted into a format that is easy for the generative AI model to handle. For example, it can be formatted as "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[1173] Generating natural conversations using generative AI models

[1174] The server loads and initializes a pre-trained generative AI model (e.g., using TensorFlow or PyTorch). It inputs a formatted dataset into the model and performs inference to generate new, natural-sounding conversations. The generated conversations are output in text format. For example, the model might generate a new message such as, "What movie have you seen recently?"

[1175] Conversation suggestions and matching compatible conversation partners

[1176] The server then incorporates the generated conversation into a template and formats it in HTML or JSON to propose it to the user. Furthermore, the server reads the user's past conversation patterns from the server's database and analyzes their compatibility with other users. A matching algorithm is used to evaluate similarities and differences and select appropriate conversation partners. For example, it might suggest to user A, "Let's talk about movies you've seen recently," and match them with user B, who also likes movies. This suggestion is sent to the user's device as an HTTP response, and the user can view the content in their browser or app.

[1177] Gathering feedback and improving the system

[1178] The user provides feedback on the suggested conversation content. For example, they can enter an evaluation such as "helpful" or "not helpful" using a dedicated button. This feedback is sent to the server in JSON format. The server receives the feedback data, stores it in a database, and analyzes it to reflect in the retraining of the model. Retraining improves the accuracy of the generative AI model, and more appropriate conversation suggestions will be made the next time the system is used.

[1179] Prompt Sentence Examples

[1180] "How do I upload my LINE chat logs and receive new conversation suggestions?"

[1181] In this way, the system can continuously provide users with natural conversations and achieve effective communication with compatible partners. Furthermore, by incorporating user feedback, the system's accuracy can be continuously improved.

[1182] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1183] Step 1: Receive chat logs from users

[1184] A user accesses the system from his / her terminal and selects a chat log file.

[1185] Specifically, click the "Select File" button on your browser or dedicated app and select a chat history file such as LINE.

[1186] The terminal sends the selected chat log file to the server as an HTTP POST request.

[1187] The server receives the HTTP request and saves the chat log files in the " / uploads / chatlogs / " directory.

[1188] Input: Chat log file selected by the user

[1189] Output: Chat log file saved on the server

[1190] Step 2: Analyzing the chat logs

[1191] The server reads the saved chat log file.

[1192] Specifically, it analyzes the file line by line and extracts information such as the message content, sender, and date and time of sending.

[1193] Using regular expressions and pattern matching, extract the necessary information from a line such as "[2023 / 10 / 01 12:34] User A: Hello."

[1194] The extracted information is converted into a format that is easy for the generative AI model to handle, for example, "{'Sender': 'User A', 'Date': '2023-10-01 12:34', 'Message': 'Hello'}".

[1195] Input: Chat log file stored on the server

[1196] Output: Formatted chat data

[1197] Step 3: Generating natural conversation using a generative AI model

[1198] The server loads and initializes the pre-trained generative AI model.

[1199] Specifically, the generative AI model is loaded from disk into memory using libraries such as TensorFlow or PyTorch.

[1200] The formatted dataset is fed into the model for inference, which generates new, natural-sounding conversations. For example, the model generates new messages such as, "What movie have you seen recently?"

[1201] The generated conversation is output in text format.

[1202] Input: Formatted chat data

[1203] Output: The new conversational text generated by the generative AI model

[1204] Step 4: Conversation suggestions and matching with compatible conversation partners

[1205] The server proposes the generated conversations to the user.

[1206] Specifically, the generated conversation is incorporated into a template, configured in HTML or JSON format, and an HTTP response is prepared for sending to the user.

[1207] Furthermore, the server reads the user's past conversation patterns from a database and analyzes the similarities and differences between them and other users.

[1208] A matching algorithm is used to select an appropriate conversation partner. For example, if user A is asked to talk about a movie they recently saw, the system will match them with user B, who also likes movies.

[1209] The suggestions are sent to the user's device, and the user can view them in their browser or app.

[1210] Input: Generated conversation text, user's past conversation data

[1211] Output: Conversation suggestions for the user and selection of compatible conversation partners

[1212] Step 5: Gather feedback and refine the system

[1213] The user provides feedback on the proposed conversation content.

[1214] Specifically, the user selects a rating such as "useful" or "not useful" and clicks a dedicated button.

[1215] The device sends the feedback data in JSON format to the server as a POST request.

[1216] The server stores the received feedback in a database, analyzes it, and reflects it in the retraining of the model.

[1217] Re-learning improves the accuracy of the generative AI model, allowing for more appropriate conversation suggestions the next time it is used.

[1218] Input: User feedback data

[1219] Output: Highly accurate conversation suggestions from an improved generative AI model

[1220] (Application example 1)

[1221] 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."

[1222] Modern communication tools make it difficult for users to find content and topics that interest them. Finding compatible conversation partners and generating natural conversations based on topics of interest is also challenging, especially for busy users. Furthermore, there is a lack of feedback mechanisms to improve the quality of generated content and conversations. This leaves users with insufficient support for effective communication.

[1223] 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.

[1224] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, senders, and sending dates and times, means for using a generative AI model to generate natural conversations and content based on the analysis results, and means for suggesting the generated conversations and content to users. This allows users to easily obtain natural conversations and content based on topics of interest and engage in appropriate communication.

[1225] A "chat log" is text data of the conversation content that a user sends and receives through a chat application.

[1226] "Analysis" is the process of extracting useful information from data and converting it into an understandable form.

[1227] A "generative AI model" is an artificial intelligence algorithm that learns from large amounts of data and generates new conversations and content.

[1228] "Suggestion" is the act of presenting information or content that is deemed useful to the user.

[1229] "Content" refers to information or entertainment that may be of interest to users, such as movies, music, articles, etc.

[1230] A "compatible conversation partner" is another user with whom effective communication can be expected based on the user's interests and past conversation patterns.

[1231] "Feedback" refers to opinions and comments provided by users regarding a system or service.

[1232] "Retraining" is the process by which an AI model is retrained based on new data and feedback to improve its performance.

[1233] The system realizing this invention analyzes a user's chat log and uses a generative AI model to suggest natural conversations and content. A specific embodiment of this system will be described below.

[1234] Hardware and Software Use

[1235] Hardware:

[1236] Smartphone (OS: Android, iOS)

[1237] Cloud server (e.g. AWS EC2)

[1238] software:

[1239] Frontend: React Native (cross-platform mobile app development)

[1240] Backend: Node.js and Express (Building the API server)

[1241] Database: MongoDB (for storing chat logs)

[1242] AI model: GPT-4 (using OpenAI API)

[1243] System operation explanation

[1244] Receive chat logs:

[1245] The user selects the chat log they want to upload using their smartphone and sends it to the server via an HTTP POST request. The server receives it and saves it in the specified directory.

[1246] Chat log analysis:

[1247] The server reads the saved chat logs and parses them line by line, including the message content, sender, and date and time, and then removes unnecessary parts and converts them into a format that is easy for the generative AI model to handle.

[1248] Conversation and content generation using generative AI models:

[1249] The server loads and initializes a pre-trained generative AI model, inputs the converted dataset into the model, and generates new, natural-sounding conversations and content based on past conversation patterns, which are then output in text format.

[1250] Conversation suggestions and compatible conversation partner matching:

[1251] The server then proposes the generated conversations and content to the user. It compares the user's past conversation patterns with those of other users and also suggests compatible partners. The proposals are sent to the user's device, where the user can view them.

[1252] Gathering feedback and improving the system:

[1253] Users provide feedback on the suggestions, which is sent to the server and analyzed to retrain and refine the generative AI model, allowing the system to continuously improve the accuracy of conversation generation.

[1254] Specific examples

[1255] For example, if User A uploads past chat logs to the app, the server receives and analyzes these chat logs. If User A has had many conversations about "movies" in past chats, the generative AI model will generate content to "talk about movies you've recently seen" and suggest it to User A. At the same time, User B, who is also interested in movies, may also be matched.

[1256] Example prompt sentence:

[1257] User's past chat logs:

[1258] 1. "Movies have been interesting lately."

[1259] 2. "What kind of movie would you like?"

[1260] 3. "I like action. I like it when it's really epic."

[1261] Generates prompt:

[1262] "The user likes action movies. Tell the AI ​​model to talk about the top 10 latest action movies."

[1263] In this embodiment, the user can easily obtain natural conversations and content based on topics of interest, and can carry out appropriate communication.

[1264] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1265] Step 1:

[1266] A user selects a chat log using a smartphone and sends it to the server through an HTTP POST request. The input is the locally stored chat log file, and the output is the HTTP request sent to the server.

[1267] Step 2:

[1268] The server saves the received chat log in the specified directory. It writes the received data (chat log file) to the disk and gets the save path as output.

[1269] Step 3:

[1270] The server reads the saved chat log file and analyzes the contents line by line. This analysis extracts the message content, sender, and sending date and time. The input is the text data of the chat log, and the output is structured data (e.g., JSON format) as the analysis result.

[1271] Step 4:

[1272] The analyzed data is converted into a format that is easy for the generative AI model to handle. It is converted into data with a specific structure (e.g., context, entities, keywords, etc.) and output as a dataset to be input into the generative AI model.

[1273] Step 5:

[1274] The server loads a pre-trained generative AI model and inputs the converted dataset into the model. The generative AI model generates new, natural-sounding conversations and content based on past conversation patterns. The input is the initialization parameters of the generative AI model and the converted dataset, and the output is the text data of the generated conversation or content.

[1275] Step 6:

[1276] The generated conversation and content are sent to the user's device as a suggestion. The server returns the generated result as an HTTP response, and the user receives the suggestion on the application. The input is the generated text, and the output is the suggestion displayed on the user's device.

[1277] Step 7:

[1278] The user provides feedback on the proposed content. The feedback is sent from the user's terminal to the server as an HTTP POST request. The input is the user's feedback text, and the output is saved as feedback data to the server.

[1279] Step 8:

[1280] The server analyzes the received feedback and retrains and improves the generative AI model. It analyzes the feedback data to extract important information and uses this information to retrain the generative AI model. The input is the feedback data, and the output is an improved generative AI model.

[1281] The specific operations at each step are carried out in sequence by the cooperation of various components, such as the server, terminal, and user, and through processes such as data transmission and reception, analysis, generation, and feedback, effective conversations and content are provided to users.

[1282] 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.

[1283] The system of this invention receives chat logs from users, analyzes them, and extracts message content, sender, and sent date and time. It also uses a generative AI model to generate natural conversations and suggest them to the user. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions based on the user's emotions.

[1284] Receive chat logs from users

[1285] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[1286] Chat log analysis

[1287] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[1288] Generating natural conversations using generative AI models

[1289] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the server generates new, natural-sounding conversations. The generated conversations are output in text format.

[1290] Conversation suggestions and matching compatible conversation partners

[1291] The server adjusts the generated conversation based on the user's emotions and makes suggestions. The suggestions include conversations with other users who are deemed compatible. The emotion engine reflects appropriate conversation content based on the emotional information recognized by the server. The server then sends the suggestions to the user's device, where the user can confirm them.

[1292] Gathering feedback and improving the system

[1293] The user provides feedback on the proposed conversation content and sends it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[1294] Specific examples

[1295] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes the chat log. The emotion engine also identifies the emotional state of the message content. For example, if past conversations have often focused on movies and the user appears to be enjoying themselves, the generative AI model will suggest, "Let's talk about an interesting movie you recently saw." If User B is also interested in movies and appears to be enjoying themselves, the server will suggest matching User A and User B, providing a conversation that is appropriate for both parties. By providing feedback on this suggestion, the system can reflect it in future conversation generation and matching, enabling it to provide even more accurate and natural conversations.

[1296] In this way, users can enjoy natural conversations without much effort and can achieve effective communication with people with whom they have good chemistry.

[1297] The processing flow will be explained below.

[1298] Step 1:

[1299] The user uses the terminal interface to select the chat log file to be uploaded.

[1300] Step 2:

[1301] The device sends the selected chat log file to the server as an HTTP POST request.

[1302] Step 3:

[1303] The server receives the chat log file sent from the device and saves it in the specified directory.

[1304] Step 4:

[1305] The server reads the saved chat log file and parses the contents line by line.

[1306] Step 5:

[1307] The server extracts message content, sender, and sent date and time from the analyzed chat log.

[1308] Step 6:

[1309] The server uses an emotion engine to analyze the emotional state of each message and add emotion information to the extracted data.

[1310] Step 7:

[1311] The server initializes the generative AI model and inputs the formatted dataset containing emotion information.

[1312] Step 8:

[1313] The server uses a generative AI model to generate new, natural conversations based on past conversation patterns and emotional information.

[1314] Step 9:

[1315] The server outputs the generated conversation in text format and adjusts it based on the user's emotions.

[1316] Step 10:

[1317] The server proposes tailored conversations to the user and compares them with other users' data to match compatible conversation partners.

[1318] Step 11:

[1319] The server identifies compatible conversation partners and sends suggestions to the user's device.

[1320] Step 12:

[1321] The user reviews the suggested conversation content on their device and adds feedback if necessary.

[1322] Step 13:

[1323] The terminal sends the user's feedback to the server.

[1324] Step 14:

[1325] The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine.

[1326] Step 15:

[1327] The server applies the improved generative AI model and emotion engine to future conversation generation.

[1328] Example 2

[1329] 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."

[1330] While conventional chat log analysis systems were able to extract message content, senders, and sending dates and times, they had problems in that they did not adequately generate natural conversations that took the user's emotional state into account or match compatible conversation partners. Furthermore, there was a lack of a way to effectively utilize user feedback on the generated conversation content to improve the system, making it difficult to consistently generate high-quality conversations.

[1331] 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.

[1332] In this invention, the server includes means for receiving a chat log from a user, means for analyzing the received chat log to extract message content, sender, and sending date and time, means for identifying the user's emotional state using an emotion engine that estimates emotions, means for generating a format-converted dataset based on the analysis result and the identified emotional state, means for using a generative AI model that uses the generated dataset as input to generate natural conversations, means for adjusting and suggesting the generated conversations based on the user's emotions, and means for receiving feedback and improving the system. This makes it possible to generate natural conversations that reflect the user's emotions and match users with compatible conversation partners, and further allows the system to be continuously improved using user feedback.

[1333] A "chat log" is a data file that shows the history of messages that a user has sent and received in the past.

[1334] An "emotion engine" is software or algorithm that analyzes message content to infer a user's emotional state.

[1335] A "generative AI model" is an artificial intelligence model that generates natural conversations based on pre-trained data.

[1336] A "formatted dataset" is a collection of input data that has been converted into a format acceptable to a generative AI model based on the results of analysis and emotion estimation.

[1337] The "means for suggesting" is a method or function for presenting the generated conversation content to the user.

[1338] "Feedback" is the act of a user providing an evaluation or comment on the proposed conversation content.

[1339] "Retraining" is a new learning process that uses feedback to improve the performance of generative AI models and emotion engines.

[1340] "Matching means" refers to a method or function that automatically connects users who are determined to be compatible with each other.

[1341] This invention is a system that receives chat logs from users, analyzes the logs, and generates and suggests natural conversations. The system also incorporates an emotion engine that recognizes the user's emotions and adjusts the generated conversations and suggestions. It also includes a means for matching compatible conversation partners and continuously improving the system based on feedback.

[1342] First, the user selects the chat log they want to upload using their device. The selected chat log is then sent from the device to the server. The server then saves the received chat log in a specified directory. This process uses standard HTTP POST requests and file I / O operations.

[1343] The server then reads the saved chat log file and analyzes it line by line. During the analysis process, it extracts the message content, sender, and date and time of sending. For example, this involves using regular expressions to extract this information from the text data. It then uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data includes the message content, sender, date and time, as well as emotional information.

[1344] Next, the server initializes a pre-trained generative AI model and inputs the converted dataset. Based on past conversation patterns and emotional information, the model generates new, natural-sounding conversations. For example, if the conversation revolves around "recent movies," the model might suggest, "Let's talk about some interesting movies you've seen recently."

[1345] The generated conversation is output in text format, and the server adjusts and suggests conversations based on the user's emotions. These suggestions include conversations with other users who are deemed compatible. Based on the emotional information recognized by the emotion engine, appropriate conversation content is generated and sent to the user's device.

[1346] Users can also provide feedback on the proposed conversation content and send it from their device to the server. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of conversation generation.

[1347] Specific examples

[1348] For example, if User A uploads a chat log exported from a social networking app to the system, the server receives and analyzes this chat log. The emotion engine also identifies the emotional state of the message content. For example, if the message "Hello" is evaluated as "positive," the generative AI model uses this information to generate a new message such as "How are you doing lately?"

[1349] Furthermore, if User B is also interested in movies and seems to enjoy them, the server will propose a match between User A and User B, providing a conversation suitable for both parties. Feedback on this proposal is provided, allowing the system to reflect this in future conversation generation and matching, enabling the system to provide even more accurate and natural conversations.

[1350] An example prompt is:

[1351] User A's emotional state: Happy

[1352] Past conversation topics: Movies, Travel, Music

[1353] Latest message: "What movie did you see last?"

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

[1355] Step 1:

[1356] The user selects the chat log they wish to upload on their device. This is done by using the device's file browser to find and select the specified chat log file. At this time, the system recognizes the selected file as the "selected chat log file."

[1357] Step 2:

[1358] The terminal sends the selected chat log file to the server. In this process, an HTTP POST request is used to send the file contents to the server. The input is the "selected chat log file" and the output is the "chat log file sent to the server."

[1359] Step 3:

[1360] The server saves the received chat log file in the specified directory. The file name is guaranteed to be unique using the user ID and timestamp. The input is the "chat log file sent to the server" and the output is the "saved chat log file."

[1361] Step 4:

[1362] The server reads and analyzes the saved chat log file line by line. Specifically, it uses regular expressions to extract the message content, sender, and sending date and time from each line. The input is the "saved chat log file," and the output is the "extracted message content, sender, and sending date and time."

[1363] Step 5:

[1364] The server uses an emotion engine to identify the user's emotional state from the analyzed message content. For example, it classifies the content into categories such as positive, negative, and neutral. The input is the extracted message content, and the output is the identified emotional state.

[1365] Step 6:

[1366] The server generates a format-converted dataset for the generative AI model based on the identified emotional state and analysis results. This is the process of formatting the data into a format that the generative AI model can understand. The input is the "analysis results and identified emotional state," and the output is the "format-converted dataset."

[1367] Step 7:

[1368] The server initializes the generative AI model and inputs the converted dataset, which generates natural-sounding conversations. The input is the converted dataset, and the output is the generated natural-sounding conversations.

[1369] Step 8:

[1370] The server adjusts the generated natural conversation based on the user's emotions and suggests it to the user. The input is the generated natural conversation and the identified emotional state, and the output is the suggested conversation content.

[1371] Step 9:

[1372] The user provides feedback on the proposed conversation content and sends it to the server via their device. The feedback can include ratings and specific comments. The input is the "proposed conversation content" and the output is "feedback from the user."

[1373] Step 10:

[1374] The server analyzes the received feedback and retrains and improves the generative AI model and emotion engine based on the content. The input is "feedback from the user" and the output is "an improved generative AI model and emotion engine."

[1375] These processing steps enable the system to generate natural conversations that reflect the user's emotions, match users with compatible conversation partners, and continuously improve the system.

[1376] (Application example 2)

[1377] 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."

[1378] Conventional chat analysis systems do not take into account users' emotional information when proposing advertisements, making it difficult to effectively present advertisements that match users' interests and needs. Furthermore, due to a lack of means to promote joint purchases and information exchange among users with similar interests, the relevance of advertisements tailored to users' preferences and the improvement of engagement are not sufficiently addressed.

[1379] 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.

[1380] In this invention, the server includes means for receiving chat logs from users, means for analyzing the received chat logs to extract message content, sender, and sending date and time, means for using a generative AI model to generate natural conversations based on the analysis results, means for generating advertisements based on the content of the generated conversations and the user's emotional information and proposing them to the user, means for encouraging joint purchases and information exchanges with other users with similar interests based on the proposed advertisements, and means for receiving feedback on the advertisements from users and retraining and improving the generative AI model and the emotional engine. This makes it possible to propose advertisements that reflect the user's emotional state and interests, further increasing engagement among users with similar interests and maximizing the effectiveness of advertisements.

[1381] A "user" is a user who provides a chat log to the system and receives the analysis results and suggestions.

[1382] A "chat log" refers to messages and conversation history sent and received by a user, and is data that includes the content, sender, and date and time of sending.

[1383] "Emotion information" refers to the user's emotional state identified based on an analysis of the chat log, and includes positive, negative, neutral, and other emotions.

[1384] A "generative AI model" refers to a pre-trained artificial intelligence model, an algorithm for generating natural conversations and advertisements based on chat logs and emotional information.

[1385] "Advertisement" refers to marketing messages and product information that are generated based on the user's interests and emotional information and are proposed to the user.

[1386] "Suggestion" refers to the act of presenting generated conversations or advertisements to the user, which contain content that corresponds to the user's interests and emotions.

[1387] "Feedback" refers to user reactions and opinions on proposed conversations and advertisements, which are used to improve the system and retrain the generative AI model.

[1388] "Joint purchasing" refers to the act of purchasing products in collaboration with other users who share similar interests, and includes means by which the system encourages this.

[1389] "Information exchange" refers to the sharing of knowledge and opinions among other users with similar interests, and includes the means by which the system facilitates this.

[1390] To implement this invention, it is necessary to build a system that analyzes users' chat logs, generates optimal advertisements based on their emotional information, and proposes them to users. This system is composed of hardware and software such as a server, user terminals, a generative AI model, an emotion engine, and a feedback database.

[1391] The server receives and analyzes chat logs sent from users' devices. Specifically, it uses a natural language processing library (e.g., NLTK) and a sentiment analysis engine (e.g., Microsoft Azure Text Analytics) to extract message content, sender, and date and time from the chat logs. This also identifies the user's sentiment information.

[1392] The analyzed data is then fed into a generative AI model (e.g., OpenAI GPT-4), which then generates optimal ads based on past conversation patterns and emotional information. This generation process ensures that the ads include content that reflects the user's interests and emotions.

[1393] The generated advertisements are then presented to the user's device, where they can be viewed using a smartphone or smart glasses. The system also matches users with similar interests, facilitating joint purchases and information exchange.

[1394] User feedback on ads is sent to the server through a feedback collection system. The server analyzes this feedback and retrains and refines the generative AI model and emotion engine. A database (e.g., MySQL) is used for this feedback process.

[1395] As a specific example, if a user uploads a chat log about "movies" to the system, the sentiment analysis engine will identify the user's positive emotions. The generative AI model will then refer to past conversation patterns about "movies" to generate ads for "interesting movies they've recently seen." These ads will be suggested to the user, encouraging them to share information and make joint purchases with other users who are interested in movies.

[1396] Here are some example prompts to input to a generative AI model:

[1397] "User conversation and sentiment: [{'content': 'I like movies', 'sender': 'user1', 'timestamp': '2023-10-24T14:53:00', 'sentiment': {'neg': 0.0, 'neu': 0.3, 'pos': 0.7, 'compound': 0.7}}]. Generate the best ad."

[1398] This allows for advertising suggestions that reflect a user's emotional state and interests, further increasing engagement between users with similar interests.

[1399] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1400] Step 1:

[1401] The user selects the chat log they want to upload using their own device. The selected chat log is sent from the device to the server. The server saves the received chat log in the specified directory.

[1402] Input: Chat log (data sent from the user's device)

[1403] Output: Saved chat log file (stored in a directory on the server)

[1404] Specific behavior:

[1405] 1. The user selects the chat log file through the terminal interface.

[1406] 2. The device sends the selected file to the server.

[1407] 3. The server saves the received file in the specified directory.

[1408] Step 2:

[1409] The server reads the saved chat log file and analyzes it line by line. During the analysis, it extracts the message content, sender, and date and time of sending. It also uses an emotion engine to identify the user's emotional state and creates input data for the generative AI model. This data also contains emotional information.

[1410] Input: Saved chat log file

[1411] Output: Analysis results (message content, sender, sending date and time, emotional information)

[1412] Specific behavior:

[1413] 1. The server reads the chat log file.

[1414] 2. Use a natural language processing library (e.g., NLTK) to extract the message content, sender, and date and time.

[1415] 3. Identify the sentiment of each message using a sentiment analysis engine (e.g., Microsoft Azure Text Analytics).

[1416] 4. Create a dataset to input the generative AI model.

[1417] Step 3:

[1418] The server initializes a pre-trained generative AI model and inputs the converted data set. Based on past conversation patterns and emotional information, the model generates new, natural-looking ads. The generated ads are output in text format.

[1419] Input: Analysis result dataset (including emotion information)

[1420] Output: Generated advertisement (text format)

[1421] Specific behavior:

[1422] 1. The server initializes a generative AI model (e.g., OpenAI GPT-4) and inputs the dataset.

[1423] 2. The generative AI model generates ads based on the input data.

[1424] 3. Create the ad copy generated as output.

[1425] Step 4:

[1426] The server then adjusts the generated advertisements based on the user's emotions and suggests them to the user. It also matches users with similar interests and shares the advertisements with them. The server then sends the suggestions to the user's device, where the user can confirm them.

[1427] Input: Generated Ad

[1428] Output: Tailored ads and matching information

[1429] Specific behavior:

[1430] 1. The server tailors the generated ads based on the user's emotions.

[1431] 2. Send the advertisement to the user's device through the interface.

[1432] 3. Using a matching algorithm to identify other users with similar interests.

[1433] 4. Share advertising information relevant to the user.

[1434] Step 5:

[1435] Users provide feedback on proposed ads and send it to the server from their devices. The server analyzes the received feedback and retrains and refines the generative AI model and emotion engine, allowing the system to continuously improve the accuracy of ad generation.

[1436] Input: User feedback

[1437] Output: Improved generative AI models and emotion engines

[1438] Specific behavior:

[1439] 1. The user submits feedback on the ad through the device interface.

[1440] 2. The server receives and analyzes the feedback.

[1441] 3. Based on the analysis results, the generative AI model and emotion engine are retrained.

[1442] 4. Reflect the learning results and improve the accuracy of the system.

[1443] 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.

[1444] 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.

[1445] 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.

[1446] 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.

[1447] 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.

[1448] 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.

[1449] 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).

[1450] 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.

[1451] 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."

[1452] 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.

[1453] 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).

[1454] 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.

[1455] 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.

[1456] 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.

[1457] 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.

[1458] 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.

[1459] 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.

[1460] 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.

[1461] 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.

[1462] 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.

[1463] 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.

[1464] The following is further disclosed regarding the above embodiment.

[1465] (Claim 1)

[1466] means for receiving a chat log from a user;

[1467] A means for analyzing the received chat log to extract message content, sender, and sending date and time;

[1468] A method using a generative AI model that generates natural conversations based on the analysis results, and

[1469] means for suggesting the generated conversation to a user;

[1470] A system including:

[1471] (Claim 2)

[1472] 10. The system of claim 1, further comprising means for automatically matching compatible conversation partners based on the generated conversation suggestions.

[1473] (Claim 3)

[1474] 10. The system of claim 1, further comprising means for receiving feedback from a user and retraining and improving the generative AI model.

[1475] "Example 1"

[1476] (Claim 1)

[1477] means for receiving a chat log from a user;

[1478] A means for analyzing the received chat log to extract message content, sender, and sending date and time;

[1479] A means for using a natural language processing model to generate natural conversations based on the analysis results;

[1480] means for suggesting the generated conversation to a user;

[1481] A means for selecting a compatible conversation partner based on the user's past conversation patterns;

[1482] means for receiving user feedback on the proposed conversation content and analyzing the feedback to retrain and refine the natural language processing model;

[1483] A system including:

[1484] (Claim 2)

[1485] 10. The system of claim 1, further comprising means for automatically matching compatible conversation partners based on the generated conversation suggestions.

[1486] (Claim 3)

[1487] 10. The system of claim 1, further comprising means for receiving feedback from a user and retraining and improving the generative AI model.

[1488] "Application Example 1"

[1489] (Claim 1)

[1490] means for receiving a chat log from a user;

[1491] A means for analyzing the received chat log to extract message content, sender, and sending date and time;

[1492] Using generative AI models to generate natural conversations and content based on the analysis results,

[1493] a means for suggesting generated conversations and content to a user;

[1494] A system including:

[1495] (Claim 2)

[1496] 10. The system of claim 1, further comprising means for automatically matching with compatible conversation partners or other users based on the generated conversation and content suggestions.

[1497] (Claim 3)

[1498] 10. The system of claim 1, further comprising means for receiving feedback from a user and retraining and improving the generative AI model.

[1499] "Example 2: Combining Emotion Engines"

[1500] (Claim 1)

[1501] means for receiving a chat log from a user;

[1502] A means for analyzing the received chat log to extract message content, sender, and sending date and time;

[1503] means for identifying an emotional state of a user using an emotion engine for estimating emotions;

[1504] means for generating a formatted dataset based on the analysis results and the identified emotional state;

[1505] a means for using a generative AI model that generates natural-sounding conversations using the generated dataset as input;

[1506] A means for adjusting and suggesting the generated conversation based on the user's emotions;

[1507] a means of receiving feedback and improving the system;

[1508] A system including:

[1509] (Claim 2)

[1510] 10. The system of claim 1, further comprising means for automatically matching compatible conversation partners based on the generated conversation suggestions.

[1511] (Claim 3)

[1512] 10. The system of claim 1, further comprising means for receiving feedback from a user and retraining and improving the generative AI model and emotion engine.

[1513] "Application example 2 when combining emotion engines"

[1514] (Claim 1)

[1515] means for receiving a chat log from a user;

[1516] A means for analyzing the received chat log to extract message content, sender, and sending date and time;

[1517] A method using a generative AI model that generates natural conversations based on the analysis results, and

[1518] means for suggesting the generated conversation to a user;

[1519] means for generating an advertisement based on the content of the generated conversation and the user's emotional information and proposing the advertisement to the user;

[1520] A system including:

[1521] (Claim 2)

[1522] 10. The system of claim 1, further comprising means for encouraging joint purchasing or information exchange with other users with similar interests based on the suggested advertisements.

[1523] (Claim 3)

[1524] 10. The system of claim 1, further comprising means for receiving feedback on the advertisement from a user and retraining and improving the generative AI model and the emotion engine. [Explanation of symbols]

[1525] 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 chat log from a user; A means for analyzing the received chat log to extract message content, sender, and sending date and time; A method using a generative AI model that generates natural conversations based on the analysis results, and means for suggesting the generated conversation to a user; A system including:

2. The system of claim 1 further comprising means for automatically matching compatible conversation partners based on the generated conversation suggestions.

3. 10. The system of claim 1, further comprising means for receiving feedback from a user and retraining and improving the generative AI model.

Images