Generative AI-based character-oriented dialogue service device

Abstract

This invention provides a character-oriented dialogue service device based on generative AI. [Solution] The generative AI-based character-oriented dialogue service device of the present invention includes: a character model generation unit that learns the characteristics of a character as content and generates a character model for the generative AI-based device; a chat room generation unit that generates a chat room with the character; a message preprocessing unit that receives dialogue messages through the chat room and preprocesses the dialogue messages as question prompts based on the context of the chat room and the context of the content; a response message output unit that inputs the question prompts as input data to the character model of the generative AI-based device and outputs response messages that interact contextually with the character as output data; and a derived content generation unit that edits the dialogue messages in the chat room and generates derived content.

Description

Technical Field

[0001] The present invention relates to character dialogue generation technology, and more particularly to a character-oriented dialogue service device based on a generative AI platform that can generate a character model based on the generative AI platform and provide a response message that interacts contextually with the character in response to a user's query.

Background Art

[0002] In modern society, interactions with people are carried out through various digital platforms, and characters are particularly at the center of such interactions. Characters communicate with users in various digital environments such as online games, chats, and virtual spaces, and are used as a medium for conveying emotions. Conventionally, characters generally interact with users by simply operating according to a predetermined script or repeating pre-produced animations.

[0003] Existing characters are mainly developed based on computer graphics technology, through which a more realistic appearance and movement can be realized. However, the process of creating such characters requires a lot of cost and time, and advanced graphic technology is required. For example, the characters used in movies and commercials (CFs) are real and sophisticated, but personalizing them or generating them according to a specific user has the drawback of being less accessible to the general user.

[0004] Furthermore, with the surge in demand for non-face-to-face services since the COVID-19 pandemic, the need for digital characters that users can interact with without direct face-to-face contact has increased. In particular, 3D character content can provide an interactive experience similar to that of a human, but its production process remains complex and costly, limiting its widespread adoption. Against this backdrop, technologies that utilize deep learning-based natural language processing techniques, such as generative AI, to enable characters to engage in more natural and intelligent conversations with users are attracting attention. Generative AI models, based on contextual understanding capabilities learned from large amounts of data, can maintain contextual consistency in conversations with users and generate customized responses tailored to user input.

[0005] In particular, unlike existing static characters, generative AI-based character models can provide dialogue that reflects the character's personality and unique characteristics by learning a specific worldview, personality, way of speaking, and experiences. This allows users to experience immersive interaction with characters rather than simple conversation, and such interactions can be applied to various content fields such as games, novels, dramas, and advertisements.

[0006] Korean Published Patent No. 10-2021-0156742 relates to a method for providing a dialogue service with an artificial intelligence character, and includes the steps of: generating an artificial intelligence character selected by user input; providing a dialogue service with the generated artificial intelligence character; and, if the function of providing a heart gift or a star gift is selected by the user input among the dialogue services with the artificial intelligence character, providing a heart gift or a star gift to the artificial intelligence character. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Korean Published Patent No. 10-2021-0156742 (December 27, 2021) [Overview of the project] [Problems that the invention aims to solve]

[0008] One embodiment of the present invention aims to provide a generative AI-based character-oriented dialogue service device that can learn the characteristics and worldview of a specific character and reflect the character's personality in dialogue with the user.

[0009] One embodiment of the present invention aims to provide a character-oriented dialogue service device based on generative AI that can generate a chat room in which a user can interact with a character, generate contextually appropriate question prompts based on messages entered in the chat room, and generate response messages that maintain the character's personality and contextual consistency. [Means for solving the problem]

[0010] Among the embodiments, the generative AI-based character-oriented dialogue service device includes: a character model generation unit that learns the characteristics of a character as content and generates a character model for the generative AI; a chat room generation unit that generates a chat room with the character; a message preprocessing unit that receives dialogue messages through the chat room and preprocesses the dialogue messages as question prompts based on the context of the chat room and the context of the content; a response message output unit that inputs the question prompts as input data to the character model of the generative AI and outputs response messages that contextually interact with the character as output data; and a derived content generation unit that edits the dialogue messages in the chat room and generates derived content.

[0011] The character model generation unit can input the character's worldview, personality, way of speaking, and experiences as expressed through webtoons, movies, dramas, novels, or games as learning data into the character model of the generative AI platform and perform the learning process.

[0012] The character model generation unit receives advertising content through the character model of the generative AI platform, and can generate character-oriented advertising content by reflecting the characteristics of the character in the description of the advertising content.

[0013] The message preprocessing unit can extract keywords from the dialogue message, determine which context to preferentially apply among the context of the chat room or the context of the featured content, and generate the question prompt.

[0014] The response message output unit analyzes the dialogue content composed of the question prompt and the response message to determine an emotional index of joy, anger, sadness, and pleasure, and can modify the character so that the emotional index of joy, anger, sadness, and pleasure is reflected in the character's way of speaking or appearance.

[0015] The response message output unit can check whether the response message contains advertising keywords, reflect the contextual interaction in the character-oriented advertising content, convert it into character-context-oriented advertising content, generate an advertising message together with the character-context-oriented advertising content and an advertising link, and provide it as another user-responsive dialogue message added following the response message.

[0016] The derivative content generation unit can group the dialogue messages in the chat room based on the context of the dialogue and process the derivative content into at least one of the following: synopsis, scenario, novel, still cuts, and video.

[0017] The derived content generation unit extracts continuous or discontinuous dialogue messages that contextually connect the dialogue messages in the chat room to perform the editing, and at least a part of the appearing content can be reflected in the derived content.

Advantages of the Invention

[0018] The disclosed technology can bring the following advantages. However, the scope of rights of the disclosed technology should not be understood to be limited thereby, because it does not mean that a specific embodiment must include all of the following advantages or only the following advantages.

[0019] The character-oriented dialogue service device of the generative AI base according to an embodiment of the present invention can learn the characteristics and worldview of a specific character and reflect the character's personality in the dialogue with the user.

[0020] The character-oriented dialogue service device of the generative AI base according to an embodiment of the present invention can generate a chat room where the user can interact with the character, generate a contextually appropriate questioning prompt based on the message input in the chat room, and generate a response message that maintains the character's personality and contextual consistency.

Brief Description of the Drawings

[0021] [Figure 1] It is a diagram for explaining the character-oriented dialogue service system according to the present invention. [Figure 2] It is a diagram for explaining the system configuration of the character-oriented dialogue service device in FIG. 1. [Figure 3] It is a diagram for explaining the configuration of the character-oriented dialogue service device in FIG. 1. [Figure 4] It is a flowchart for explaining the functional configuration of the character-oriented dialogue service device according to the present invention. [Figure 5]This is a diagram for explaining an embodiment of character information and an initial dialogue start screen according to the present invention. [Figure 6] This is a diagram for explaining an embodiment of a user comment creation and reaction interface according to the present invention. [Figure 7] This is a diagram for explaining an embodiment of a dialogue generation and dialogue management screen according to the present invention. [Figure 8] This is a diagram for explaining an embodiment of a story posting option interface utilizing dialogue with a character.

Mode for Carrying Out the Invention

[0022] The description of the present invention is merely an embodiment for structural or functional description, and the scope of rights of the present invention should not be construed as being limited by the embodiments described in this specification. That is, since the embodiments can be variously modified and can have various forms, the scope of rights of the present invention should be understood to include equivalents in which the technical idea can be realized. Also, the objects or effects presented in the present invention do not mean that a specific embodiment must include all of them or only such effects, and thus the scope of rights of the present invention should not be understood to be limited thereby.

[0023] On the other hand, the meanings of the terms described in this application should be understood as follows. <00001​​​​When one component is described as being "connected" to another, it should be understood that it may be directly connected to the other component, but other components may also exist in between. On the other hand, when one component is described as being "directly connected" to another, it should be understood that no other components exist in between. Similarly, other expressions describing relationships between components, such as "between" and "immediately between," or "adjacent to" and "directly adjacent to," should be interpreted in the same way.

[0026] A singular expression should be understood to include multiple expressions unless the context clearly indicates otherwise, and terms such as “includes” or “possesses” should be understood to indicate the presence of an implemented feature, figure, stage, action, component, part, or combination thereof, and not to preemptively rule out the possibility of the presence or addition of one or more other features, figures, stages, actions, components, parts, or combinations thereof.

[0027] The identification codes at each stage (e.g., a, b, c, etc.) are used for explanatory purposes only and do not indicate the order of the stages. The stages may occur in a different order than those listed unless the context clearly indicates a specific order. In other words, the stages may occur in the same order as listed, substantially simultaneously, or in reverse order.

[0028] The present invention may be embodied as computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes any type of recording device on which data readable by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage devices. Furthermore, the computer-readable recording media can be distributed across a network of computer systems, and computer-readable code can be stored and executed in a distributed manner.

[0029] All terms used herein have the same meaning as that generally understood by a person of ordinary skill in the art to which this invention pertains, unless otherwise defined. Terms defined in commonly used dictionaries should be interpreted in accordance with their meaning in the context of the relevant art and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined in this application.

[0030] Figure 1 illustrates the character-oriented dialogue service system according to the present invention.

[0031] Referring to Figure 1, the character-oriented dialogue service system 100 may include a user terminal 110, a character-oriented dialogue service device 130, and a database 150.

[0032] The user terminal 110 can be connected to the character-oriented dialogue service device 130 via a network, and can receive generative AI-based character-oriented dialogue services provided by the character-oriented dialogue service device 130, and can be considered a computing terminal operated by the user. Here, the character-oriented dialogue service can be, but is not limited to, solutions for user-customized interaction, immersive content creation, and advertising message delivery, and may further include game interfaces, educational interactive support services, or virtual character-based consultation services.

[0033] The user terminal 110 may consist of one or more terminals, and if it consists of multiple terminals, it may include the first user terminal, the second user terminal, ..., the nth (n is a natural number) user terminal. For example, the user terminal 110 may be embodied in a smartphone, laptop computer, or other computer that can operate in conjunction with the character-oriented dialogue service device 130, but is not necessarily limited to these, and may also be embodied in various devices, including tablet PCs.

[0034] Furthermore, the user terminal 110 may be connected to the character-oriented dialogue service device 130 via a network, and multiple user terminals 110 may be connected to the character-oriented dialogue service device 130 simultaneously. The user terminal 110 can install and run a dedicated program or application for interacting with the character-oriented dialogue service device 130.

[0035] The character-oriented dialogue service device 130 can be implemented in a server corresponding to a computer or program that executes the generative AI-based character-oriented dialogue service according to the present invention. For example, the character-oriented dialogue service device 130 may be implemented in a server that generates dialogue messages that reflect the characteristics and context of a character, generates dialogue responses or derived content based on the dialogue messages, and provides them to the user terminal 110, and each stage of the generative artificial intelligence-based character-oriented dialogue service method according to the present invention may be executed on the server. The character-oriented dialogue service device 130 may be connected to the user terminal 110 by a wired network or a wireless network such as Bluetooth®, WiFi, or LTE, and can send and receive data with the user terminal 110 via the wired or wireless network.

[0036] Furthermore, the character-oriented dialogue service device 130 may be implemented to operate in conjunction with an independent external system (not shown in Figure 1). For example, the character-oriented dialogue service device 130 can operate in conjunction with an advertising management service system that provides users with advertising content delivery solutions that reflect the character's context, or with a content creation support service system that provides derivative content creation solutions that utilize dialogue data.

[0037] The database 150 can be a storage device that stores various information necessary for the operation of the character-oriented dialogue service device 130. For example, the database 150 can store the character's worldview, personality, way of speaking, and learning data, but is not limited to this. It may also store user dialogue history, context information, and generated derivative content data that occur during the operation of the character-oriented dialogue service device 130.

[0038] In Figure 1, the database 150 is shown as a separate device from the character-oriented dialogue service device 130, but it is not necessarily limited to this, and can of course be implemented as part of the character-oriented dialogue service device 130.

[0039] Figure 2 is a diagram illustrating the system configuration of the character-oriented dialogue service device shown in Figure 1.

[0040] Referring to Figure 2, the character-oriented dialogue service device 130 may include a processor 210, memory 230, user input / output unit 250, network input / output unit 270, and communication port unit 290.

[0041] The processor 210 can execute character-oriented dialogue service procedures of a generative AI platform according to embodiments of the present invention, manage the memory 230 that is read or created in such a process, and schedule synchronization times between volatile and non-volatile memory in the memory 230. The processor 210 can control the overall operation of the character-oriented dialogue service device 130, is electrically connected to the memory 230, the user input / output unit 250, and the network input / output unit 270, and can control the flow of data between them. The processor 210 may be embodied as the CPU (Central Processing Unit) or GPU (Graphics Processing Unit) of the character-oriented dialogue service device 130.

[0042] The memory 230 may include an auxiliary storage device that is implemented as a non-volatile memory such as an SSD (Solid State Disk) or HDD (Hard Disk Drive) and is used to store all the data necessary for the character-oriented dialogue service device 130, or it may include a main memory that is implemented as a volatile memory such as RAM (Random Access Memory). Furthermore, the memory 230 can store a set of instruction words that execute the character-oriented dialogue service method of the generative AI base according to the present invention, by being executed by an electrically connected processor 210.

[0043] The user input / output unit 250 includes an environment for receiving user input and an environment for outputting specific information to the user, and may include an input device including an adapter such as a touchpad, touchscreen, image keyboard, or pointing device, and an output device including an adapter such as a monitor or touchscreen. In one embodiment, the user input / output unit 250 may correspond to a computing device connected via a remote connection, in which case the character-oriented dialogue service device 130 may run as an independent server.

[0044] The network input / output unit 270 provides a communication environment for connecting with the user terminal 110 via a network, and may include adapters for communication such as LAN (Local Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), and VAN (Value Added Network). Furthermore, the network input / output unit 270 may be implemented to provide short-range communication functions such as WiFi and Bluetooth, or wireless communication functions of 4G or higher, for wireless data transmission.

[0045] The communication port section 290 may be implemented as a port mapping table that performs data routing during the process of sending and receiving data over the network. Here, the communication port section 290 can separate communication sessions between the user terminal 110 and the server by assigning a source port unique to the user terminal 110, thereby preventing data collisions during the data transmission and reception process.

[0046] Figure 3 is a diagram illustrating the configuration of the character-oriented dialogue service device shown in Figure 1.

[0047] Referring to Figure 3, the character-oriented dialogue service device 130 may include a character model generation unit 310, a chat room generation unit 320, a message preprocessing unit 330, a response message output unit 340, a derived content generation unit 350, and a control unit 360.

[0048] The character-oriented dialogue service device 130 does not necessarily have to simultaneously include all of the above-mentioned functional configurations. Depending on the embodiment, some of the above configurations may be omitted, or some or all of the above configurations may be selectively included. Furthermore, the character-oriented dialogue service device 130 may be implemented as an independent module that selectively includes some of the above configurations, and the character-oriented dialogue service method of the generative AI platform according to the present invention can be executed through the coordination between each module. The operation of each configuration will be described in detail below.

[0049] The character model generation unit 310 can generate a character model for a generative AI platform by learning the characteristics of characters that appear in the content. Here, the character model generation unit 310 can determine fictional characters that appear in various media such as novels, movies, dramas, webtoons, and games, and can quantitatively extract the characteristics of the character, including the worldview within the media, the character's personality, way of speaking, behavioral patterns, and experiences. For example, by applying text analysis techniques such as text tokenization and TF-IDF (Term Frequency-Inverse Document Frequency), the character model generation unit 310 can quantify character characteristics, including the character's lines, behavioral patterns, and emotions, and provide them as a training dataset for the character model of the generative AI platform.

[0050] In one embodiment, the character model generation unit 310 can qualitatively extract character characteristics within the medium. For example, the character model generation unit 310 can qualitatively extract specific character characteristics by representing the character's worldview and background information in a structured graph to grasp semantic relationships, or by classifying and linking the settings and worldview of the content in which the character appears and performing ontology-based analysis, and provide these as a training dataset for a character model on a generative AI platform. In one embodiment, the character model generation unit 310 can generate a training dataset by integrating the quantitatively and qualitatively extracted character characteristics, and provide the generated training dataset as input for a character model on a generative AI platform.

[0051] In one embodiment, the character model generation unit 310 can learn a character model on a generative AI platform by inputting the character's worldview, personality, way of speaking, and experiences expressed through webtoons, movies, dramas, novels, or games as learning data. Here, the character model generation unit 310 can extract character features, including worldview, personality, way of speaking, and experiences, from various media, including webtoons, movies, dramas, novels, or games, and generate a learning dataset. For example, the character model generation unit 310 can structure the worldview as background information text among the character features, and define specific reactions and decision-making methods for personality according to emotional characteristics (e.g., optimistic and lively). Furthermore, the character model generation unit 310 can learn the way a specific character speaks by inputting features of speech, including writing style, vocabulary selection, and tone, into the character model on the generative AI platform. However, it is not limited to this, and a learning dataset can be generated by constructing data in the form of a timeline including the causes and effects of past actions based on events experienced by a specific character in a story.

[0052] In one embodiment, the character model generation unit 310 can tokenize the training dataset and apply an embedding technique to train a character model on a generative AI platform. Here, the character model generation unit 310 can convert the training dataset into tokens and apply an embedding technique to understand the meaning and position of each token, thereby learning the relationships between the characteristics of each character, including their worldview, personality, way of speaking, and experiences. In one embodiment, the character model generation unit 310 can train the character model on a generative AI platform to reflect personality and way of speaking according to the characteristics of each character by performing interrelationship learning through a self-attention mechanism, but is not limited to this. By further learning the data of a specific character during the learning process (Fine-Tuning), it is possible to more precisely reflect the unique personality and dialogue patterns of the character.

[0053] In one embodiment, the character model generation unit 310 receives advertising content through a character model on a generative AI platform and can generate character-oriented advertising content by reflecting the characteristics of the character in the description of the advertising content. Here, character-oriented advertising content can refer to advertising messages or public relations content that have been restructured to reflect the unique characteristics of a particular character. The character model generation unit 310 can adjust the style, vocabulary, and expression of advertising content to reflect the characteristics of a particular character, such as their worldview, personality, way of speaking, and experiences. For example, if the character model generation unit 310 is a humorous character, it can add witty expressions and jokes to the advertising message, but is not limited to this. If the character has an old-fashioned way of speaking, it can adjust the advertising content to reflect the characteristics and way of speaking of the character by adjusting it to convey the advertisement in a formal way.

[0054] The chat room generation unit 320 can generate a chat room with a character. Here, the chat room generation unit 320 generates a virtual space for facilitating a conversation between the user and a specific character, and can provide the character's response in response to the user's input. For example, the chat room generation unit 320 can provide the user with an interface containing multiple character profiles, and when a specific character is selected by the user's input, it can generate the character by loading the character's characteristics from the database 150. Here, the chat room generation unit 320 can generate a chat room that reflects the chat room's theme, background, and interface design, etc., according to the character's characteristics. For example, if a character with a fantasy background is loaded, the chat room generation unit 320 can set the background to a room in a medieval castle, but is not necessarily limited to this; in the case of a character with a modern background, it may generate a chat room with a building theme such as a city cafe.

[0055] In one embodiment, the chat room generation unit 320 can generate character responses in real time and manage the flow of the conversation by receiving user input. For example, when a user inputs a question about a specific topic, the chat room generation unit 320 can maintain a consistent context so that the user input and the character's response are consistently linked by providing a response about that topic that reflects the character's characteristics.

[0056] The message preprocessing unit 330 receives dialogue messages through the chat room and can preprocess them as question prompts based on the context of the chat room and the context of the featured content. Here, the question prompt can correspond to a user's question input to the character model of the generative AI platform, and can, for example, correspond to a combination of the user's message and contextual information. The message preprocessing unit 330 can preprocess dialogue messages as question prompts based on the context of the chat room, which includes the flow of messages previously exchanged in the chat room and the character's current state, and the context of the featured content, which includes setting information of the worldview to which the character belongs (e.g., webtoon, novel, and movie). Here, the message preprocessing unit 330 can decide which information from the dialogue context and the featured content context to apply preferentially, depending on the user's dialogue message. For example, if the user's dialogue message is related to the character's personal experience, the message preprocessing unit 330 can preferentially apply the dialogue context, and if the user's dialogue message is a question about the character's worldview, it can preferentially apply the context of the featured content.

[0057] In one embodiment, the message preprocessing unit 330 can receive text data entered by a user through a chat room, analyze the entered message, and perform a preprocessing process to refine the text by removing unnecessary spaces and special characters. For example, the message preprocessing unit 330 can apply natural language processing (NLP) techniques to analyze the semantic structure of the dialogue message, divide the dialogue message into words or grammatical units, identify the grammatical role of each word (noun, verb, adjective, etc.), and preprocess it as a question prompt.

[0058] In one embodiment, the message preprocessor 330 can extract keywords from the dialogue message, determine a preferred context from the context of the chat room or the context of the content appearing, and generate a question prompt. Here, the keywords can correspond to the main subject, object, and action in the dialogue message. The message preprocessor 330 can apply natural language processing techniques to analyze the words that make up the dialogue message, search for major parts of speech including nouns, verbs, and adjectives, and determine a preferred context from the context of the chat room or the context of the content appearing. Furthermore, the message preprocessor 330 can apply the TF-IDF (Term Frequency-Inverse Document Frequency) technique to calculate the frequency and importance of words, select relatively important words from the dialogue message, determine a preferred context, and generate a question prompt. The message preprocessor 330 is not necessarily limited to this, and can also perform dependency parsing on the dialogue message to analyze the relationships between words, grasp the central subject from the dialogue message, and generate a question prompt.

[0059] The response message output unit 340 can input a question prompt as input data to the character model of the generative AI platform and output a response message that contextually interacts with the character as output data. Here, the response message output unit 340 can generate a response that reflects the characteristics of a specific character to the user's question, based on the character model of the generative AI platform. For example, the response message output unit 340 can generate a continuous dialogue by analyzing the relationship with previous dialogue messages based on the context of the chat room of the question prompt through the character model of the generative AI platform. Furthermore, depending on the context of the content in which the character appears, the response message output unit 340 can generate a response message that reflects characteristics such as background information, worldview, character personality, and major events set in the original work (movie, novel, game, etc.) to which a specific character belongs. In one embodiment, the response message output unit 340 can analyze the intent of the user's question from the question prompt and generate a sentence that is appropriate to the user's intent, reflecting the characteristics of a specific character selected by the user.

[0060] In one embodiment, the response message output unit 340 can analyze the dialogue content, which consists of a question prompt and a response message, to determine an emotional index of joy, anger, sadness, and happiness, and can modify the character so that the emotional index of joy, anger, sadness, and happiness is reflected in the character's way of speaking or appearance. Here, the response message output unit 340 can apply natural language processing (NLP) techniques to analyze the dialogue content based on the question prompt entered by the user and the response message generated by the character, and can identify the emotional elements of the question prompt and the response message. For example, the response message output unit 340 can identify positive, negative, and emotions including joy, anger, sadness, and happiness from the question prompt and the response message, but is not necessarily limited to these. In one embodiment, the response message output unit 340 can quantify the emotional elements of the analyzed question prompt and the response message to determine an emotional index of joy, anger, sadness, and happiness, and can generate a response message that reflects the character's emotional state according to the emotional index of joy, anger, sadness, and happiness.

[0061] In one embodiment, the response message output unit 340 can check whether the response message contains advertising keywords and reflect the contextual interaction in the character-oriented advertising content to convert it into character-context-oriented advertising content. Here, character-context-oriented advertising content can be a customized advertisement that takes into account the characteristics of the character and the flow of the dialogue, and can be an advertisement that reflects the characteristics of the character, such as the character's way of speaking, personality, and worldview, to provide the user with an immersive experience.

[0062] For example, the response message output unit 340 can utilize a generative AI model to extract important keywords within the response message, determine whether those keywords are relevant to the advertisement, and extract advertising keywords that play a crucial role in the advertisement, including product names, service names, and specific brand names. The response message output unit 340 can analyze the extracted advertising keywords and, if the response message contains advertising keywords, decide whether or not to provide the advertisement. For example, if the character has a humorous personality, the response message output unit 340 can naturally provide the user with an advertisement by generating character-oriented advertising content that reflects the character's characteristics, allowing the character to naturally suggest a specific advertisement based on their humor and approachability.

[0063] In one embodiment, the response message output unit 340 can check whether the response message contains advertising keywords, reflect the contextual interaction in the character-oriented advertising content, convert it into character-context-oriented advertising content, generate an advertising message with the character-context-oriented advertising content along with an advertising link, and provide it as another user-responsive dialogue message added following the response message. Here, advertising keywords may, but are not limited to, specific brand names, product names, and service names, and may also be specific words detected based on a database 150 that stores pre-set advertising keywords. If advertising keywords are detected, the response message output unit 340 can analyze the context of the conversation and provide an advertisement that matches the character's way of speaking and the situation. For example, the response message output unit 340 can generate an advertising message according to the character's characteristics, including the character's personality, way of speaking, and situation, and generate an advertising message that connects to the existing conversation flow. In one embodiment, after providing an advertising message to the user, the response message output unit 340 can generate a dialogue message that connects to the existing conversation flow in response to the user's reaction, thereby allowing the conversation between the character and the user to continue even after the advertisement has been exposed.

[0064] In one embodiment, the response message output unit 340 can generate an advertising message along with an advertising link for character context-oriented advertising content, and provide it as another user-responsive dialogue message added after the response message. Here, in the process of generating the character context-oriented advertising content, the response message output unit 340 can provide the user with an advertising message that introduces the advertised product and service, reflecting the characteristics of the character and the flow of dialogue, as well as links that allow the user to move to a web page or product page related to the advertisement. Furthermore, after providing the response message, the response message output unit 340 can generate a user-responsive dialogue message that guides the user to respond to the advertising message. For example, the response message output unit 340 can generate a user-responsive dialogue message that guides the user to take action on the advertising message, such as clicking on an advertising link to purchase an advertised product or pressing a button to indicate interest. Furthermore, the response message output unit 340 is not necessarily limited to this, and may also provide an advertising message to the user and provide a subsequent advertising message that guides the user to respond to the character context-oriented advertising content.

[0065] The derivative content generation unit 350 may generate derivative content by editing the dialogue messages in the chat room. Here, derivative content can refer to existing content or content obtained by transforming or reconfiguring existing content into a new format using the dialogue messages. For example, the derivative content generation unit 350 can generate new derivative content in the form of a scenario, novel, video, or synopsis based on a theme that emerged from the conversation. Here, the derivative content generation unit 350 can analyze the messages exchanged between the user and a specific character in the chat room, extract the theme, emotion, and style of the messages, and convert them into the form of a scenario, novel, or video according to the theme of the conversation to generate new derivative content.

[0066] In one embodiment, the derivative content generation unit 350 can group dialogue messages in a chat room based on the context of the dialogue and process the derivative content into at least one of a synopsis, scenario, novel, still cut, and video. Here, the derivative content generation unit 350 can group one or more messages that occurred in the chat room and group messages that feature similar themes, emotions, and events. The derivative content generation unit 350 can analyze the context and themes between the grouped messages and generate various forms of creative works according to the context and themes between the grouped messages.

[0067] For example, the derivative content generation unit 350 can create a brief summary of the story by summarizing the flow of the dialogue based on the context and themes between the grouped messages, and can generate a synopsis that summarizes the story based on the characteristics of the events and people that appear in the dialogue. The derivative content generation unit 350 can also analyze the context and themes between the grouped messages and generate a script in scenario format based on the dialogue messages. For example, the derivative content generation unit 350 can generate a scenario that describes how the locations and situations that appear in the dialogue messages actually unfold, by generating specific scenes in the scenario. The derivative content generation unit 350 is not necessarily limited to this, and may also generate still cuts by converting the scenes and situations depicted in the dialogue messages into image form.

[0068] In one embodiment, the derivative content generation unit 350 can extract and edit continuous or discontinuous dialogue messages from a chat room, contextually following the dialogue messages, and reflect at least a portion of the content in the derivative content. Here, the derivative content generation unit 350 can extract continuous dialogue messages from the chat room dialogue messages, including dialogues on themes such as conflict and problem-solving, develop them as a continuous story, and reflect them in the derivative content. Furthermore, in the case of discontinuous dialogue messages where the subject and direction of the dialogue are not constant, or where multiple diverse events appear in combination, the derivative content generation unit 350 can analyze the contextual connectivity of each conversation and generate derivative content based on that contextual connectivity. In one embodiment, the derivative content generation unit 350 can extract continuous or discontinuous dialogue messages, edit the flow of the dialogue to convert it into at least one continuous story, and generate derivative content.

[0069] The control unit 360 controls the overall operation of the character-oriented dialogue service device 130 and can manage the flow of control or data between the character model generation unit 310, the chat room generation unit 320, the message preprocessing unit 330, the response message output unit 340, and the derived content generation unit 350.

[0070] Figure 4 is a flowchart illustrating the functional configuration of the character-oriented dialogue service device according to the present invention.

[0071] Referring to Figure 4, the character-oriented dialogue service device 130 can generate a character model on a generative AI platform by learning the characteristics of the characters, which are the characters appearing in the story, as content, based on the character model generation unit 310 (step S410). The character-oriented dialogue service device 130 can generate a chat room with the characters through the chat room generation unit 320 (step S430).

[0072] The character-oriented dialogue service device 130 receives dialogue messages through the chat room via the message preprocessing unit 330 and can preprocess the dialogue messages as question prompts based on the context of the chat room and the context of the featured content (step S450). The character-oriented dialogue service device 130 can input the question prompts as input data to the character model of the generative AI platform via the response message output unit 340 and output a response message that contextually interacts with the character as output data (step S470). The character-oriented dialogue service device 130 can edit the dialogue messages in the chat room via the derived content generation unit 350 to generate derived content (step S490).

[0073] Figure 5 illustrates one embodiment of the character information and initial dialogue start screen according to the present invention.

[0074] Referring to Figure 5, the character-oriented dialogue service device 130 can provide an interface that allows the user to choose one of several characters. Here, the character-oriented dialogue service device 130 can select one of several characters in response to user input and visually represent the characteristics of the selected character on the screen. For example, the character-oriented dialogue service device 130 can provide the user with a profile image of the character, its name, and a brief description of the character (e.g., background, personality, and way of speaking).

[0075] In one embodiment, the character-oriented dialogue service device 130 can provide interaction buttons for the user to express feelings towards the character. For example, the character-oriented dialogue service device 130 can enable the user to express feelings towards the character by providing interaction buttons that can express feelings towards the character such as "like," "attractive," and "charismatic."

[0076] Furthermore, if an interaction button is selected in response to user interaction, the character-oriented dialogue service device 130 can switch to a dialogue interface with a character, along with the phrase "Start a conversation." Here, the character-oriented dialogue service device 130 can provide the user with an initial dialogue message through the character. For example, depending on the character's characteristics, the character-oriented dialogue service device 130 can provide an initial dialogue message including "Hello," "Good afternoon," and "How are you?" based on the character's worldview, speaking style, personality, and current situation, thereby initiating a dialogue that reflects the character's characteristics.

[0077] Figure 6 illustrates one embodiment of the user comment creation and response interface according to the present invention.

[0078] Referring to Figure 6, the character-oriented dialogue service device 130 can display the title of content related to a character at the top of the screen and display user-created comments in chronological order. Here, the character-oriented dialogue service device 130 can receive user opinions or questions through the comment input field at the bottom of the screen and register them as comments. In one embodiment, the character-oriented dialogue service device 130 can interact with other users by providing additional comment and interaction buttons for registered comments. For example, the character-oriented dialogue service device 130 can, through the additional comment function, allow users to create replies to specific comments and continue the conversation, but is not limited to this. It may also provide a function to express one's reaction by pressing a "like" button on comments that one likes, thereby providing a function for exchanging opinions and communicating with other users.

[0079] In one embodiment, the character-oriented dialogue service device 130 can provide a user with a preview of a dialogue record in which another user has interacted with a specific character. Here, the character-oriented dialogue service device 130 can provide the user with a preview of a portion of the dialogue record between a specific character and another user. This allows the character-oriented dialogue service device 130 to provide a screen that allows the user to check the progress of the conversation with the character and to check the conversation style, the character's way of speaking, and their reactions. Here, if a user who is interested in the preview wishes to be provided with additional dialogue, the character-oriented dialogue service device 130 can provide a payment screen via a "Continue" button to start a new dialogue with the character or to purchase the story that continues from the preview.

[0080] Figure 7 illustrates one embodiment of the dialogue generation and dialogue management screen according to the present invention.

[0081] Referring to Figure 7, the character-oriented dialogue service device 130 can provide an interface that allows users to generate and manage conversations they have with characters. Here, the character-oriented dialogue service device 130 can provide a list of user-generated conversations at the top or center of the screen, and can provide a selection box in each conversation list to allow the user to select conversations to modify or delete. Furthermore, when the user clicks on a selected conversation, the character-oriented dialogue service device 130 can display the entire content of that conversation at the bottom of the screen or in a new pop-up, providing an editing function that allows the user to check the character's conversation content and user messages, and edit the details as needed.

[0082] In one embodiment, the character-oriented dialogue service device 130 can generate a new conversation by inputting character statements and user messages, by providing a button at the bottom of the screen that allows the user to create a new conversation. The character-oriented dialogue service device 130 can insert the newly generated conversation at the top of an existing conversation list or at a position specified by the user, and can provide a delete function that allows the user to select an unwanted conversation and delete it by pressing a delete button. In one embodiment, the character-oriented dialogue service device 130 can provide a preview function for the generated conversation, allowing the user to examine the way the character's conversation is expressed and to check whether the character's conversation matches the character's characteristics. For example, the character-oriented dialogue service device 130 can check in the preview whether the content of the conversation, "(smiling) Thank you for understanding," matches a specific facial expression or atmosphere of the character.

[0083] Figure 8 illustrates one embodiment of a story display option interface that utilizes conversations with characters.

[0084] Referring to Figure 8, the character-oriented dialogue service device 130 can provide settings to display stories based on the content of conversations generated by the user. For example, the character-oriented dialogue service device 130 can receive a story title and story description from the user, input the user-generated story title into the story name input field, and briefly describe the plot and core content of the story through the story description input field. In one embodiment, the character-oriented dialogue service device 130 can provide a category setting function so that the user can classify the genre and character of the stories.

[0085] In one embodiment, the character-oriented dialogue service device 130 can provide users with a function to set the public and private status of a story, allowing them to either make the story public to all users or keep it private to maintain it for personal use or modification. Here, the character-oriented dialogue service device 130 can also provide users with a function to switch a story to paid content, allowing the device to set a credit price and determine the price according to the value of the story when the user switches a particular story to paid content.

[0086] In one embodiment, the character-oriented dialogue service device 130 can display customized advertisements based on user behavior data. For example, the character-oriented dialogue service device 130 can analyze the theme and keywords of the dialogue and automatically recommend and display advertisements for relevant products on the screen. The character-oriented dialogue service device 130 can also, but is not limited to, create a scene in which a specific product is recommended while the user is continuing a conversation with a character, and may provide an interface in which the user can click on or explore the advertisement. In one embodiment, the character-oriented dialogue service device 130 can present various advertisement screens, including recommendation-based advertisements, story-linked advertisements, reward-based advertisements, and customized advertisements, in response to user interaction.

[0087] Although preferred embodiments of the present invention have been described above with reference to the present invention, those skilled in the art will understand that the present invention can be modified and altered in various ways without departing from the spirit and scope of the invention as set forth in the following claims. [Explanation of Symbols]

[0088] 100 ···Character-Oriented Dialogue Service System 110 ···User terminal 130 ···Character-Oriented Dialogue Service Device 150 ···Database 210 ... Processor 230...memory 250 ···User Input / Output Section 270 ···Network Input / Output Section 290 ···Communication port section 310 ···Character Model Generation Unit 320 ···Chat Room Generation Unit 330 ···Message preprocessing unit 340 ···Response message output section 350 ···Derived Content Generation Unit 360 ···Control Unit

Claims

1. A character model generation unit that learns the characteristics of the characters appearing in the story as content and generates a character model based on generative AI; A chat room generation unit that generates a chat room with the aforementioned character; A message preprocessing unit receives a dialogue message through the chat room and preprocesses the dialogue message as a question prompt based on the context of the chat room and the context of the content being presented; A response message output unit inputs the aforementioned question prompt as input data to the character model of the generative AI platform and outputs a response message that interacts contextually with the character as output data; and A derived content generation unit that edits the dialogue messages in the aforementioned chat room to generate derived content; A character-oriented dialogue service device based on generative AI, including the following.

2. The aforementioned character model generation unit, The character-oriented dialogue service device for a generative AI platform according to claim 1, characterized in that the worldview, personality, way of speaking, and experiences of the character expressed through webtoons, movies, dramas, novels, or games are input as learning data into the character model of the generative AI platform to perform the learning.

3. The aforementioned character model generation unit, The character-oriented dialogue service device for a generative AI platform according to claim 2, characterized in that advertising content is input through the character model of the generative AI platform, and character-oriented advertising content is generated by reflecting the characteristics of the character in the description of the advertising content.

4. The message preprocessing unit, The character-oriented dialogue service device based on a generative AI platform according to claim 1, characterized in that it extracts keywords from the dialogue message, determines a preferred application context from among the context of the chat room or the context of the appearing content, and generates the question prompt.

5. The response message output unit is, The character-oriented dialogue service device based on generative AI according to claim 1, characterized in that it analyzes the content of the dialogue composed of the question prompt and the response message to determine an emotional index of joy, anger, sadness, and pleasure, and modifies the character so that the emotional index of joy, anger, sadness, and pleasure is reflected in the character's way of speaking or appearance.

6. The response message output unit is, The generative AI-based character-oriented dialogue service device according to claim 1, characterized in that it checks whether the response message contains advertising keywords, reflects the contextual interaction in character-oriented advertising content and converts it into character-context-oriented advertising content, generates an advertising message together with the character-context-oriented advertising content and provides it as another user-responsive dialogue message added following the response message.

7. The aforementioned derived content generation unit, The generative AI-based character-oriented dialogue service device according to claim 1, characterized in that it groups dialogue messages in the chat room based on the context of the dialogue and processes the derived content into at least one of a synopsis, scenario, novel, still cut, and video.

8. The aforementioned derived content generation unit, The character-oriented dialogue service device based on generative AI according to claim 7, characterized in that it extracts contextually continuous or discontinuous dialogue messages from the dialogue messages of the chat room, performs the editing, and reflects at least a portion of the appearing content in the derived content.

Images