Artificial intelligence dialogue processing device and method

Abstract

The present invention provides an artificial intelligence dialogue processing device and method that provides a response appropriate to the context of the dialogue. [Solution] The artificial intelligence dialogue processing device 130 includes: an artificial intelligence chat room generation unit that generates an artificial intelligence chat room in which an artificial intelligence chat agent that responds to user dialogues participates; a search extension interface unit that generates a search extension interface for setting up a search extension chat room used in the artificial intelligence chat room; a search extension query word generation unit that receives user chat messages and generates search extension query words used to search past chat history; an artificial intelligence chat message generation unit that searches past chat history and generates an artificial intelligence chat message by inputting it to the artificial intelligence chat agent; and an artificial intelligence chat response unit that provides an artificial intelligence chat message as a response to the user's chat message.

Description

Technical Field

[0001] The present invention relates to artificial intelligence and communication technology, and more particularly, to an artificial intelligence dialogue processing apparatus and method that can provide a response adapted to the context of a dialogue by performing question and answer in an artificial intelligence chat room based on information and content of a search extension chat room selected by a user, and causing a plurality of artificial intelligence chat agents generated according to a search option to participate in a user chat room. Background Art

[0002] In modern society, the utilization of digital communication tools is rapidly increasing in various fields such as remote work, global collaboration, and customer support. These tools provide text-based chat rooms capable of transmitting and receiving messages in real time, file sharing, video conference functions, and the like, and support smooth communication between users. Particularly, in project management and team collaboration environments, information sharing and decision making are efficiently performed through chat rooms in which a large number of team members participate simultaneously.

[0003] However, conventional dialogue tools have several limitations in supporting efficient information exchange and collaboration between users. For example, in a chat room in which a large number of users participate, it may be difficult to quickly search for and provide necessary information, and there is a lack of a function for providing useful information at an appropriate timing without interfering with the flow of chat. Further, since real-time support that immediately responds to a user's request is insufficient, important information may be missing, or a delay may occur in a decision-making process.

[0004] In particular, the ability to selectively utilize text, image, audio, or file-based data stored in specific chat rooms according to search options, and to clearly provide the rationale behind AI-generated responses in conjunction with relevant dialogue blocks, is not adequately addressed by conventional technologies. Therefore, there is a need for dialogue processing technology for a search-enhanced platform that can search and present information related to user questions in real time while maintaining the continuity of the conversation.

[0005] Korean Published Patent No. 10-2024-0007989 describes an integrated chatbot service that combines multiple chatbot groups into one, accurately searches for and provides content that matches the user's intent, and improves the relevance of search results through additional search term recommendations.

[0006] The integrated chatbot platform service provision method includes the process of the integrated chatbot receiving a question from a user device, extracting key keywords from the question, transmitting these to a first chatbot and a second chatbot, each chatbot deriving content search results related to the key keywords, and providing a response to the question to the user device based on the derived search results. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Korean Published Patent Publication No. 10-2024-0007989 [Overview of the project] [Problems that the invention aims to solve]

[0008] The present invention aims to provide an artificial intelligence dialogue processing device and method that generates an artificial intelligence chat room and enables an artificial intelligence chat agent to generate a response that is appropriate to the context of the conversation based on the information and content of the search-enhanced chat room selected by the user.

[0009] The present invention aims to provide an artificial intelligence dialogue processing device and method that can configure a search environment that suits the user's requirements by providing a search extension interface that allows the user to set up a search extension chat room among multiple chat rooms in which the user has participated, and to select a search interval or search theme.

[0010] The present invention aims to provide an artificial intelligence dialogue processing device and method that can improve the response accuracy and contextual relevance of an artificial intelligence chat agent by generating search-extended query terms based on a user's chat message and searching the past chat history of a search-extended chat room using the generated search-extended query terms.

[0011] The present invention aims to provide an artificial intelligence dialogue processing device and method that allows users to confirm evidence-based responses within the flow of a chat by assigning response justification links to dialogue blocks selected according to search extended query terms, thereby enabling users to directly access relevant information through response justification links presented with artificial intelligence chat messages.

[0012] The present invention aims to provide an artificial intelligence dialogue processing device and method that ensures user control in the artificial intelligence response generation process and enables the generation and reflection of real-time information in a collaborative environment by providing a function that allows the user to edit the artificial intelligence's response or their own chat message during the artificial intelligence chat message provision process.

[0013] The present invention aims to provide an artificial intelligence dialogue processing device and method that can flexibly utilize multiple artificial intelligence chat agents suitable for various collaboration scenarios by providing a structure that can generate a new artificial intelligence chat room that reflects the changed settings when search options are changed or when the configuration of the search-enhanced chat room is changed. [Means for solving the problem]

[0014] To achieve the above objective, an artificial intelligence dialogue processing device according to one aspect of the present invention includes: an artificial intelligence chat room generation unit that generates an artificial intelligence chat room in which an artificial intelligence chat agent capable of responding to user dialogue participates; a search extension interface unit that generates a search extension interface for setting up a search extension chat room to be used in the artificial intelligence chat room, wherein the search extension chat room is composed of the user's participating chat room; a search extension query word generation unit that receives the user's chat message as input in the artificial intelligence chat room and generates a search extension query word used to search the past chat history of the search extension chat room; an artificial intelligence chat message generation unit that searches the past chat history of the search extension chat room using the search extension query word and generates an artificial intelligence chat message by inputting the user's chat message and the searched past chat history as an integrated input prompt to the artificial intelligence chat agent; and an artificial intelligence chat response unit that provides the artificial intelligence chat message to the artificial intelligence chat room as a response to the user's chat message by the artificial intelligence chat agent.

[0015] The artificial intelligence chat agent may be invited by the user to simultaneously participate in other chat rooms that the user is participating in.

[0016] The search extension interface unit can set a search interval or search theme for the search extension chat room via the search extension interface.

[0017] The search extension interface unit can connect to a search extension agent that controls searches and responses to an external database via the search extension interface.

[0018] The search extension interface unit can generate a new artificial intelligence chat room when the search interval or search theme for the search extension chat room is changed, or when the search extension agent is changed, via the search extension interface.

[0019] The search extension interface unit can add or remove the search extension chat rooms via the search extension interface.

[0020] The search extension query term generation unit can determine whether the user's chat message belongs to the search theme of the search extension chat room and decide whether or not to generate the search extension query term. If the search extension query term generation unit fails to generate the search extension query term, it can decide whether or not to provide assistance from other artificial intelligence chat agents present in other artificial intelligence chat rooms set to other search themes related to the search extension query term.

[0021] The artificial intelligence chat message generation unit can select dialogue blocks corresponding to the search extension query terms from the past chat history of the search extension chat room, and assign response justification links to the selected dialogue blocks, thereby utilizing the dialogue blocks and response justification links to generate the integrated input prompt. The artificial intelligence chat message generation unit can provide the artificial intelligence chat message along with the response justification links through the artificial intelligence chat agent, and can confirm the content of the dialogue blocks through the user's interaction.

[0022] The artificial intelligence chat response unit can directly modify the artificial intelligence chat message through user editing control during the process of providing the artificial intelligence chat message.

[0023] Said artificial intelligence chat response unit may, in the process of providing said artificial intelligence chat message, modify said user's chat message under the user's editing control, so that said artificial intelligence chat agent indirectly modifies said artificial intelligence chat message.

[0024] An artificial intelligence dialogue processing method according to one aspect of the present invention, which is made to achieve the above object, is an artificial intelligence dialogue processing method performed in an artificial intelligence dialogue processing apparatus, comprising: an artificial intelligence chat room generation step of generating an artificial intelligence chat room in which an artificial intelligence chat agent capable of responding to a user's dialogue participates; a search expansion interface generation step of generating a search expansion interface for setting a search expansion chat room that can be used in said artificial intelligence chat room, said search expansion chat room being composed of chat rooms that said user has participated in; a search expansion query term generation step of receiving said user's chat message as input in said artificial intelligence chat room, and generating a search expansion query term that can be used to search past chat histories of said search expansion chat room; an artificial intelligence chat message generation step of searching past chat histories of said search expansion chat room using said search expansion query term, and inputting said user's chat message and said searched past chat histories as an integrated input prompt to said artificial intelligence chat agent, thereby generating an artificial intelligence chat message; and an artificial intelligence chat response step of providing said artificial intelligence chat message to said artificial intelligence chat room as a response to said user's chat message from said artificial intelligence chat agent. [Effects of the Invention]

[0025] The present invention has the following effects. However, this does not mean that a specific embodiment must include all or only the following effects, and therefore the disclosed technical scope should not be understood as being limited thereby.

[0026] The artificial intelligence dialogue processing apparatus and method according to the present invention generates an artificial intelligence chat room, and enables an artificial intelligence chat agent to generate a response adapted to the dialogue context based on the information and content of a search-extended chat room selected by a user.

[0027] The artificial intelligence dialogue processing apparatus and method according to the present invention can configure a search environment adapted to a user's request by setting a search-extended chat room among a plurality of chat rooms that the user has joined, and providing a search extension interface that allows a user to select a search section or a search theme.

[0028] The artificial intelligence dialogue processing apparatus and method according to the present invention can improve the response accuracy and context adaptability of an artificial intelligence chat agent by generating a search-extended query term based on a user's chat message, and searching past chat histories of the search-extended chat room using the generated search-extended query term.

[0029] The artificial intelligence dialogue processing apparatus and method according to the present invention enables a user to check a ground-based response in the flow of a chat by assigning a response ground link to a dialogue block sorted according to a search-extended query term, and allowing the user to directly access relevant information through the response ground link presented together with an artificial intelligence chat message.

[0030] The artificial intelligence dialogue processing apparatus and method according to the present invention can ensure the user's controllability over the response generation process of artificial intelligence, and enable generation and reflection of real-time information in a collaborative environment by providing a function that allows the user to edit an artificial intelligence response or the user's own chat message during the process of providing an artificial intelligence chat message.

[0031] The artificial intelligence dialogue processing device and method according to the present invention provide a structure that can generate a new artificial intelligence chat room that reflects the changed settings when search options are changed or when the configuration of the search-enhanced chat room is changed, thereby enabling flexible use of multiple artificial intelligence chat agents that are suited to diverse collaboration scenarios. [Brief explanation of the drawing]

[0032] [Figure 1] This is a diagram illustrating an artificial intelligence chat processing system based on one embodiment of the present invention. [Figure 2] This diagram illustrates the system configuration of the artificial intelligence dialogue processing device shown in Figure 1. [Figure 3a] This diagram illustrates the functional configuration of the artificial intelligence dialogue processing device shown in Figure 1. [Figure 3b] This diagram illustrates the functional configuration of the artificial intelligence dialogue processing device shown in Figure 1. [Figure 4] This is a flowchart illustrating the artificial intelligence dialogue processing method performed in the artificial intelligence dialogue processing device shown in Figure 1. [Figure 5a] Figure 1 is a diagram illustrating the process of generating and utilizing dialogue data in the artificial intelligence dialogue processing device. [Figure 5b] Figure 1 is a diagram illustrating the process of generating and utilizing dialogue data in the artificial intelligence dialogue processing device. [Figure 6a] This figure shows an example of a user interface (UI) configuration implemented in the artificial intelligence dialogue processing device shown in Figure 1. [Figure 6b] This figure shows an example of a UI configuration implemented in the artificial intelligence dialogue processing device shown in Figure 1. [Figure 6c] This figure shows an example of a UI configuration implemented in the artificial intelligence dialogue processing device shown in Figure 1. [Modes for carrying out the invention]

[0033] Since the descriptions relating to the present invention are merely embodiments for structural or functional explanation, the technical scope of the present invention should not be construed as being limited by the embodiments described herein. That is, since embodiments are subject to various modifications and can take on various forms, the scope of the present invention should be understood to include equivalents that can realize the technical idea. Furthermore, since the purposes or effects presented in the present invention do not mean that a particular embodiment must include all of them or only such effects, the technical scope of the present invention should not be understood as being limited by them.

[0034] On the other hand, the meanings of the terms used herein should be understood as follows:

[0035] Terms such as "first," "second," etc., are used to distinguish one component from another, and these terms should not limit the scope of the technology. For example, the first component may be named the second component, and similarly, the second component may also be named the first component.

[0036] When one component is described as being "connected" to another, it should be understood that it can be directly connected to that other component, but that other components may exist in between. Conversely, when one component is described as being "directly connected" to another, it should be understood that no other components exist in between. On the other hand, other expressions describing the relationship between components, such as "between" and "immediately between," or "adjacent to" and "directly adjacent to," should be interpreted similarly.

[0037] A singular expression should be understood to include multiple expressions unless the context clearly indicates a different meaning, and terms such as "includes" or "possesses" should be understood to indicate the presence of a feature, number, step, action, component, part, or combination thereof, without pre-existing exclusion of the presence or possibility of adding one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0038] The identification codes (e.g., a, b, c, etc.) in each step are used for explanatory purposes only and do not indicate the order of the steps. Each step may occur in a different order than specified unless the context clearly indicates a specific order. That is, each step may occur in the same order as specified, substantially simultaneously, or in the opposite order.

[0039] The present invention is embodied by computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices that store data read by a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk (registered trademark), optical data storage devices, etc. Furthermore, the computer-readable recording media may be distributed across a network of computer systems, and computer-readable code may be stored and executed in a distributed manner.

[0040] All terms used herein have the same meaning as generally understood by a person of ordinary skill in the art to which the present 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 not in an ideal or overly formal sense unless expressly defined herein.

[0041] Figure 1 is a diagram illustrating an artificial intelligence chat processing system according to one embodiment of the present invention.

[0042] Referring to Figure 1, the artificial intelligence chat processing system 100 includes multiple user terminals 110, an artificial intelligence dialogue processing device 130, and a database 150.

[0043] User terminal 110 corresponds to a computing device operated by a user. User terminal 110 can be embodied as a desktop, laptop, tablet PC, or smartphone, and may consist of various electronic devices. User terminal 110 includes at least one of the first user terminal 110a, the second user terminal 110b, and the third user terminal 110c.

[0044] The user terminal 110 has the function of allowing the user to create a new chat room or to exchange and share information such as text, photos, audio, and files in an existing chat room. The user can set at least one of the existing chat rooms as a search-enhanced chat room, and configure the artificial intelligence dialogue processing device 130 to search the past chat history of that chat room and generate artificial intelligence chat messages.

[0045] The user terminal 110 is embodied by a mobile device or a desktop device and is connected to the artificial intelligence dialogue processing device 130 via cellular communication, Wi-Fi® communication, or internet communication.

[0046] The artificial intelligence dialogue processing device 130 is a computing device connected to the user terminal 110 via a network, and performs functions such as generating artificial intelligence chat rooms, providing a search extension interface, generating search extension query terms, and generating artificial intelligence chat messages. The artificial intelligence dialogue processing device 130 works in conjunction with the user terminal 110 to analyze the past chat history of the search extension chat room and generates artificial intelligence chat messages that are appropriate to the context of the conversation through an artificial intelligence-based response generation function.

[0047] The artificial intelligence dialogue processing device 130 analyzes the user's chat messages, searches for dialogue blocks stored in the search-enhanced chat room, constructs contextual information necessary for response generation, and generates an artificial intelligence chat message that includes a response justification link.

[0048] The artificial intelligence dialogue processing unit 130 interacts with the user terminal 110 through a dedicated program installed on the user terminal 110 or a web-based agent, and performs data synchronization and calculation processing in response to user approval.

[0049] The database 150 is a storage device that stores various data used in the process by which the artificial intelligence dialogue processing device 130 generates artificial intelligence chat messages. The database 150 stores user chat messages, search extension query terms, searched dialogue blocks, response basis links, user setting information, and analyzed information.

[0050] Furthermore, although the database 150 in Figure 1 is shown as a logical storage device included in the artificial intelligence dialogue processing device 130, it is not necessarily limited to this and can of course be realized by a separate, independent storage device or a cloud-based database.

[0051] Figure 2 is a diagram illustrating the system configuration of the artificial intelligence dialogue processing device shown in Figure 1.

[0052] Referring to Figure 2, the artificial intelligence dialogue processing unit 130 includes a processor 210, memory 230, user input / output unit 250, network input / output unit 270, and communication port unit 290.

[0053] The processor 210 controls the overall operation of the artificial intelligence dialogue processing unit 130 and interacts with the memory 230, user input / output unit 250, network input / output unit 270, and communication port unit 290 to manage the data flow. The processor 210 executes the data processing procedures of the artificial intelligence dialogue processing unit 130, manages the data in the memory 230 that is read or written in this process, and schedules the synchronization time between the volatile memory and non-volatile memory contained in the memory 230.

[0054] Furthermore, the processor 210 generates an artificial intelligence chat room in response to user requests and controls the process of setting up the search extension interface, selecting or changing the search extension chat room, generating search extension query terms, analyzing the past chat history of the search extension chat room, generating artificial intelligence chat messages, and providing responses. For example, when the search extension chat room settings are changed, the processor 210 determines whether a new artificial intelligence chat room has been created and performs the function of constructing a response basis link based on the searched dialogue block. The processor 210 is embodied by the CPU (Central Processing Unit) or GPU (Graphics Processing Unit) of the artificial intelligence dialogue processing device 130, and the amount of data processing is adjusted according to the computing performance of the device.

[0055] Memory 230 includes an auxiliary storage device embodied by non-volatile memory such as an SSD (Solid State Disk) or HDD (Hard Disk Drive), and a main storage device embodied by volatile memory such as RAM (Random Access Memory). Memory 230 also stores the instruction set of data processing procedures executed by the processor 210, and stores search-extended chat room setting information, search-extended query terms, searched dialogue blocks, response basis links, user chat messages, artificial intelligence chat messages, and user edit information, and assists in response generation by searching as needed. Memory 230 records user input data provided during the search-extended query term generation process, past chat history of the search-extended chat room, and response information generated by the artificial intelligence chat agent, and performs a function to support calculations of the artificial intelligence dialogue processing device 130.

[0056] The user input / output unit 250 is a device for receiving user input and outputting information, and includes an input device such as a touchpad, touchscreen, software keyboard, or pointing device, and an output device such as a monitor or touchscreen. The user input / output unit 250 receives user questions as input through the artificial intelligence dialogue processing unit 130 and performs the function of outputting a search extension interface, retrieved dialogue blocks, artificial intelligence chat messages, and response justification links. In addition, the user input / output unit 250 also provides a question and answer interface to the artificial intelligence dialogue processing unit 130 in conjunction with the user terminal, even in a remote connection environment.

[0057] The network input / output unit 270 provides a communication environment for connection with the user terminal and includes network interfaces such as LAN (Local Area Network), MAN (Metropolitan Area Network), WAN (Wide Area Network), and VAN (Value Added Network). The network input / output unit 270 also connects with the user terminal via short-range wireless communication functions such as Wi-Fi® and Bluetooth®, or via a 4G or higher mobile communication network, and performs the function of sending search extension query terms, search requests, retrieved dialogue blocks, artificial intelligence chat messages, and response basis links to the user terminal.

[0058] The communication port unit 290 is implemented by a module that performs port mapping or routing functions in the process of sending and receiving data over the network. The communication port unit 290 dynamically manages source and destination ports to divide sessions in the user question and response data transmission process, prevents data collisions, and optimizes network load. The communication port unit 290 also sets data packet priorities to optimize the transmission speed of AI chat messages and response links.

[0059] Figures 3a and 3b are diagrams illustrating the functional configuration of the artificial intelligence dialogue processing device shown in Figure 1.

[0060] Referring to Figure 3a, the artificial intelligence dialogue processing device 130 includes an artificial intelligence chat room generation unit 310, a search extension interface unit 320, a search extension query word generation unit 330, an artificial intelligence chat message generation unit 340, an artificial intelligence chat response unit 350, and a control unit 360. Also, as shown in Figure 3b, the artificial intelligence chat room generation unit 310 includes an artificial intelligence chat agent 311.

[0061] In this case, embodiments of the present invention do not necessarily have to simultaneously include all of the above 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 to realize the invention. Furthermore, one embodiment of the present invention can be realized by independent modules that selectively include some of the above configurations, and the artificial intelligence dialogue processing method can be executed through the coordination between these modules. The operation of each configuration will be described in detail below.

[0062] The artificial intelligence chat room generation unit 310 generates an artificial intelligence chat room in which an artificial intelligence chat agent 311 capable of responding to user conversations participates. Here, the artificial intelligence chat agent 311 is a response generation module based on a language model pre-trained to process user conversations, and is registered as an initial participant in the chat room by the artificial intelligence chat room generation unit 310. The artificial intelligence chat agent 311 receives user messages in real time within the chat room, performs message analysis, intent interpretation, conversation context tracking, etc., and generates artificial intelligence chat messages. Furthermore, even when the artificial intelligence chat agent 311 participates in multiple chat rooms with different purposes simultaneously, it is configured to independently maintain the context of each chat room, and its operating mode is set according to the role required in each chat room (business consultation, knowledge retrieval platform explanation, meeting summary, etc.). The artificial intelligence chat agent 311 is controlled to either dynamically join a new chat room in response to invitation information, or to generate additional responses while maintaining the conversational context in existing chat rooms.

[0063] The artificial intelligence chat room generation unit 310 receives an artificial intelligence chat room generation request from the user terminal 110 and configures an artificial intelligence chat room with the artificial intelligence chat agent 311 registered as an initial participant. The artificial intelligence chat room generation unit 310 also has a function to invite the artificial intelligence chat agent 311 to other chat rooms that the user has joined, allowing them to participate simultaneously.

[0064] The artificial intelligence chat room generation unit 310 interprets the chat room generation request input from the user terminal 110 and configures chat room attributes. The configured attributes include the type of artificial intelligence chat agent 311, a chat room identifier, user participation information, etc. At the time of chat room generation, the artificial intelligence chat room generation unit 310 registers the artificial intelligence chat agent 311 as an initial participant and configures it to process responses to user chat messages. The artificial intelligence chat room generation unit 310 can also manage multiple artificial intelligence chat rooms simultaneously and configures invitation information so that the artificial intelligence chat agent 311 can participate in multiple chat rooms in parallel. The invitation information is generated in response to user requests or internal system dialogue flow analysis.

[0065] In one embodiment of the present invention, the artificial intelligence chat room generation unit 310 receives a chat room generation request as input from the user terminal 110. Based on the request, the artificial intelligence chat room generation unit 310 registers a new artificial intelligence chat room and sets the artificial intelligence chat agent 311 as the initial participant. The artificial intelligence chat room generation unit 310 assigns a chat room identifier to the generated artificial intelligence chat room and manages it so that it can be operated independently.

[0066] In another embodiment, the artificial intelligence chat room generation unit 310 receives a request from the user terminal 110 to invite the artificial intelligence chat agent 311 as an additional participant to a specific chat room. In this case, the artificial intelligence chat room generation unit 310 configures the artificial intelligence chat agent 311 to participate in the new chat room in addition to the chat rooms it is already participating in. The artificial intelligence chat agent 311 configured in this way simultaneously receives user messages as input in multiple chat rooms and generates responses.

[0067] In another embodiment of the present invention, the artificial intelligence chat room generation unit 310 analyzes the user's recent message patterns and specific work themes to suggest the generation of a new artificial intelligence chat room. For example, if the user repeatedly mentions a specific project name or document work theme, the artificial intelligence chat room generation unit 310 provides the user with a guidance message recommending an artificial intelligence chat room configuration suitable for that theme. If the user approves the suggestion, the artificial intelligence chat room generation unit 310 automatically generates a new artificial intelligence chat room with the recommended configuration.

[0068] In other embodiments, the artificial intelligence chat room generation unit 310 generates artificial intelligence chat rooms conditionally by referring to external status information such as user location, schedule information, or device status. For example, when a user enters a meeting room according to the meeting schedule, the artificial intelligence chat room generation unit 310 automatically generates an artificial intelligence chat room for meeting recording and sets the artificial intelligence chat agent 311 as a participant. The artificial intelligence chat room generation unit 310 also configures a hierarchical dialogue environment in which higher-level chat rooms and lower-level chat rooms are operated in a multi-stage structure. The higher-level chat room handles the overall project summary, and the lower-level chat rooms are set to process individual business flows, and the artificial intelligence chat agent 311 provides responses by referring to both the higher-level and lower-level structures.

[0069] The search extension interface unit 320 is responsible for setting the search environment and search range for applying the search extension function in the artificial intelligence chat room, and functions as a higher-level setting module that defines the overall search range referenced by the search extension query term generation unit 330 and the artificial intelligence chat message generation unit 340.

[0070] The search extension interface unit 320 generates an interface (search extension interface) for setting up a search extension chat room, providing the user terminal 110 with an environment in which the user can select a search extension chat room to be used for the search extension function from among multiple chat rooms in which the user has joined. The search extension chat room is managed as a data set that the search extension query word generation unit 330 and the artificial intelligence chat message generation unit 340 search and refer to past chat history, and is used as a core element that constitutes the search target range.

[0071] The search extension interface unit 320 provides a user interface for setting conditions used in the search extension function, such as search intervals, search themes, and search categories. The search interval can be set as a recent period or a specific period, and the search theme can be configured using methods such as tag-based selection or text-based input. The set search options are stored as internal parameters and applied as filtering criteria when the search extension query term generation unit 330 generates search extension query terms.

[0072] The search extension interface unit 320 provides connection settings for a search extension agent that extends the search extension function in conjunction with an external database or external knowledge repository. The search extension agent includes various external information sources such as document repositories, business databases, and internal corporate knowledge repositories, and the search extension interface unit 320 configures its internal settings so that the extended information source selected by the user is included in the search scope.

[0073] The search extension interface unit 320 receives user input to add or delete search extension chat rooms in real time, updates the search chat room set, and provides it to the search extension query word generation unit 330. Furthermore, if search settings (search extension chat rooms, search options, extension agents, etc.) are changed, it updates internal parameters in conjunction with the control unit 360 so that the changes are immediately reflected in the search extension function.

[0074] The search extension interface unit 320 integrates and manages higher-level control elements of the search extension function, such as search range definition, search option settings, and search source configuration, and functions as a core module that constitutes the overall search environment used in the search extension query word generation step and the artificial intelligence chat message generation step.

[0075] In one embodiment of the present invention, the search extension interface unit 320 provides a list of chat rooms the user has joined in a table format through a UI executed on the user terminal 110. The user can select one or more chat rooms from this list to be used as search extension chat rooms. The search extension interface unit 320 also provides a date input UI (e.g., calendar-based input) for specifying a search interval and a text input UI for specifying a search theme. When the user changes the search interval or search theme, the search extension interface unit 320 stores the change as an internal setting value, and if the changed setting conflicts with an existing artificial intelligence chat room operation method, it requests the control unit 360 to generate a new artificial intelligence chat room.

[0076] Furthermore, the search extension interface unit 320 provides the user with a list of search extension agents for connecting to external databases. The user can select a specific database connection module, and the selected search extension agent is applied to the search structure of the artificial intelligence chat room. If the user requests that a search extension chat room be removed, the search extension interface unit 320 updates its internal parameters to remove the chat room from the search target set.

[0077] In another embodiment of the present invention, the search extension interface unit 320 analyzes the user's recent conversation patterns and automatically recommends search themes to be used for search extension. The search extension interface unit 320 extracts keywords that appear repeatedly from recent chat messages stored in the user terminal 110 and configures these keywords as search theme candidates. For example, if words such as "contract," "project schedule," and "technical specifications" appear at a certain rate or higher in the user's chat messages, the search extension interface unit 320 presents these to the user as a list of search theme candidates. If the user selects a recommended search theme, the search extension interface unit 320 stores the selected search theme as a search setting value and links this setting so that the search extension query word generation unit 330 can use it in the query generation process. As a result, the search extension interface unit 320 can improve the convenience of the search setting process and search accuracy through automatic recommendation, even without the user directly inputting search themes.

[0078] In another embodiment, the search enhancement interface unit 320 analyzes the user's dialogue flow or a specific work context to automatically recommend candidate chat rooms to be used as search enhancement chat rooms. The search enhancement interface unit 320 extracts unique identifying elements such as a specific project name, person in charge name, or document title from recent user chat messages, searches for existing chat rooms containing the same elements, and displays them as search enhancement chat room candidates. For example, if a user enters a message related to "A Project scheduling," the search enhancement interface unit 320 automatically identifies chat rooms that have previously mentioned "A Project" and includes them in the list of search enhancement chat room candidates for the user. As a result, the user is automatically recommended highly relevant chat rooms without having to directly search for them, simplifying the search enhancement setup process and improving search quality.

[0079] In yet another embodiment, the search extension interface unit 320 constitutes an advanced search setting interface that provides additional filter conditions in addition to the search interval and search theme. The search extension interface unit 320 provides the user terminal 110 with a multi-layer filtering UI that includes participant conditions, message type conditions, importance tag conditions, etc. The user can select, for example, at least one of the following conditions: a participant filter that searches only for messages containing a specific user, a message type filter that searches only for messages containing a document link or image, and an importance-based filter that searches only for messages with the importance tag. The search extension interface unit 320 stores the selected multi-layer filter conditions as internal setting values ​​and communicates these conditions to the chat history search process of the search extension query word generation unit 330 so that they are reflected. As a result, the search extension interface unit 320 can precisely adjust the search range even in complex business environments, enabling the generation of high-precision search infrastructure responses.

[0080] Additionally, the search enhancement interface unit 320 dynamically adjusts search settings based on the user's schedule information, location information, or device environment information. The search enhancement interface unit 320 automatically sets the search environment by referring to the schedule data, location data, and device status information provided by the user terminal 110. For example, if a user connects to an AI chat room at a meeting time registered in the calendar schedule, the search enhancement interface unit 320 automatically presents chat rooms containing meeting minutes or meeting-related tags as search enhancement candidates. Also, if a pattern is detected in which the user's work messages are concentrated in a specific time period (e.g., immediately after arriving at work), the search enhancement interface unit 320 automatically sets the search interval to "last day". If the user connects using a mobile device, the search options are switched to simplified mode and configured to display only the minimum number of selection items. As a result, the search enhancement interface unit 320 can adaptively respond to the user situation and automate search settings, significantly improving search efficiency and user convenience.

[0081] The search extension query term generation unit 330 analyzes chat messages entered by users in the artificial intelligence chat room and generates search extension query terms for searching the past chat history of the search extension chat room. The search extension query term generation unit 330 refers to search settings such as search themes, search intervals, and search categories set by the search extension interface unit 320, determines whether the user message falls within the search range, and decides whether to generate a search extension query term and the generation format based on the result of this determination.

[0082] The search extension query term generation unit 330 analyzes the semantic elements of the user message (theme, individual name, time information, work unit, contextual expression, etc.) to construct core query terms used in the search extension function. The search extension query term generation unit 330 applies analysis techniques such as individual name recognition, text extraction (OCR), and object recognition not only to text messages but also to unstructured data such as files, images, videos, and audio inputs to derive semantic core elements, which are then converted into search extension query terms.

[0083] If the search extension query term generation unit 330 determines during the search extension query term generation process that the search results are inappropriate or lack sufficient relevance to the search settings, it checks whether there are other artificial intelligence chat rooms outside the search extension chat room that are related to the search theme, and determines an artificial intelligence chat agent 311 participating in such a chat room as a candidate for auxiliary query processing. If an auxiliary query is needed, the search extension query term generation unit 330 selects one of the following methods: making a one-time query request to the artificial intelligence chat agent 311, or configuring a cooperative structure in a new chat room that enables continuous responses.

[0084] The search extension query term generation unit 330 transmits metadata such as the search target chat room information, search options, and filtering criteria, along with the finally generated search extension query term, to the artificial intelligence chat message generation unit 340. The transmitted search extension query term and metadata are referenced by the artificial intelligence chat message generation unit 340 when it constructs an integrated input prompt and are used as basic data for selecting dialogue blocks that fall within the search range from the past chat history of the search extension chat room. Through this data processing procedure, the search extension query term generation unit 330 accurately reflects the semantic characteristics and search settings of user messages, thereby improving the precision of the artificial intelligence response of the search infrastructure.

[0085] In one embodiment of the present invention, the search extension query word generation unit 330 analyzes a chat message input from the user terminal 110 and determines whether the semantic theme of the message is included in the search theme specified by the search extension interface unit 320. If it is determined that the message is included in the search theme, the search extension query word generation unit 330 extracts core words, contextual expressions, mentioned individual names, etc., from the user message and generates a search extension query word for searching past chat history in the search extension chat room. If the search extension query word generation unit 330 fails to generate a search extension query word, it analyzes the cause of the failure to generate the search extension query word and checks whether there are other artificial intelligence chat rooms with the same or similar search themes. In this case, the search extension query word generation unit 330 decides whether to request assistance from another artificial intelligence chat agent 311. For example, if a user message falls under the theme "Technical Review Request," but there is insufficient past history for that theme in the current search-enhanced chat room, the search-enhanced query term generation unit 330 searches for other artificial intelligence chat rooms that have search themes related to technical reviews, and selects either a temporary query request or continued participation in the invited platform to the artificial intelligence chat agent 311.

[0086] In another embodiment of the present invention, the search extension query term generation unit 330 generates search extension query terms by extracting semantic elements not only from text messages but also from unstructured data such as files, images, videos, and audio inputs. The search extension query term generation unit 330 analyzes individual names, repeated terms, and key visual information contained in the unstructured data, evaluates whether the analyzed information matches the search theme, and constructs a suitable search extension query term. For example, when a user uploads a technical document file, the search extension query term generation unit 330 analyzes the file name, internal item names, or visual notation elements to derive query term candidates such as "technical specifications," "implementation configuration," and "performance requirements." When an image is input, it extracts text or equipment names within the image through OCR or object recognition processing, determines the relevance to the search theme, and generates a query term. Furthermore, the search extension query term generation unit 330 combines semantic elements extracted from different input formats to generate an integrated query term. For example, the system automatically constructs highly accurate search-enhanced query terms such as "A Project Meeting Minutes" by combining the project name extracted from the image with the intended request confirmed from the text message. As a result, the search-enhanced query term generation unit 330 can stably maintain the quality of semantic-based query generation even in dialogue environments containing unstructured data, and can be effectively applied in environments where complex information is mixed, such as technical document processing or business document retrieval.

[0087] In another embodiment, if the search extension query term generation unit 330 determines that the generation of the search extension query term has failed or that it lacks relevance to the search theme, it evaluates the theme similarity between user-participating chat rooms to search for alternative search targets. If sufficient search history is not available in existing search extension chat rooms, the search extension query term generation unit 330 automatically includes other chat rooms with high similarity in the candidate group. The search extension query term generation unit 330 generates query term candidates for a single user message based on multiple chat room themes to complement or expand the search scope. For example, the search extension query term generation unit 330 generates query term candidates for the same message that reflect different chat room themes such as project schedule, contract change records, and meeting minutes summaries. The search extension query term generation unit 330 also determines whether or not to provide assistance from other artificial intelligence chat agents 311 participating in the candidate chat rooms. The search extension query term generation unit 330 temporarily utilizes the artificial intelligence chat agent 311 for query processing as needed, or, in environments requiring long-term dialogue cooperation, configures a new artificial intelligence chat room to enable continuous support. As a result, the search extension query term generation unit 330 can automatically expand the search target even in search failure situations and support the execution of auxiliary queries utilizing multiple agents, thereby increasing the reliability and flexibility of search infrastructure response generation even in complex business environments.

[0088] The artificial intelligence chat message generation unit 340 queries the past chat history of the search-enhanced chat room based on the search-enhanced query term generated by the search-enhanced query term generation unit 330, and uses the queried dialogue blocks and user chat messages to construct an integrated input prompt to be input to the artificial intelligence chat agent 311. The integrated input prompt represents a standardized input object that integrates the user message, the search-enhanced query term, the dialogue blocks selected from the search-enhanced chat room, the response basis link corresponding to each dialogue block, and the metadata necessary for response generation into a single structured input set, and functions as a unit for providing the necessary context, source, and search information in a consistent format during the process of generating a response by the artificial intelligence chat agent 311.

[0089] The artificial intelligence chat message generation unit 340 evaluates the degree of keyword matching, occurrence time, message type, etc., for the searched dialogue blocks and selects dialogue blocks that are highly relevant to the user message. A response basis link, including a unique identifier, dialogue time, and participant information, is generated and attached to each selected dialogue block, and this link is used as basis data that the artificial intelligence chat agent 311 refers to when generating a response within the integrated input prompt. The artificial intelligence chat message generation unit 340 combines the user message, search extended query words, and selected dialogue blocks to generate the final structure of the integrated input prompt and configures it in a normalized format.

[0090] Furthermore, the artificial intelligence chat message generation unit 340 receives artificial intelligence chat messages generated by the artificial intelligence chat agent 311 and controls them to be transmitted to the artificial intelligence chat response unit 350 so that they can be provided to the user. The artificial intelligence chat messages include a response basis link, and if the user selects a link, an interaction function is provided so that the content of the corresponding dialogue block can be immediately confirmed.

[0091] The artificial intelligence chat message generation unit 340 applies a prompt optimization policy to maintain the quality of input data during the process of constructing an integrated input prompt. If the amount or length of the retrieved dialogue block exceeds a predetermined standard, the artificial intelligence chat message generation unit 340 normalizes the dialogue block to limit the number of input tokens, taking into account the processing efficiency of the artificial intelligence chat agent 311. The artificial intelligence chat message generation unit 340 is further configured to extract thematic core sentences and reflect them in the integrated input prompt in a summarized form. Through such structural control, the artificial intelligence chat message generation unit 340 can maintain the core context necessary for response generation while preventing the accumulation of unnecessary information.

[0092] The artificial intelligence chat message generation unit 340 filters out messages from the searched dialogue blocks that have low semantic contribution or are noise, and excludes them from the prompt. Noise messages consist of greetings, interjections, emoji-centered messages, non-business dialogues, etc. This filtering refines the input environment so that the artificial intelligence chat agent 311 can generate more semantic-based responses.

[0093] Additionally, the artificial intelligence chat message generation unit 340 supports access control and security setting functions for response-based links. If a particular dialogue block contains sensitive information or confidential company information, the artificial intelligence chat message generation unit 340 applies security policies by verifying user permissions and restricting link access, or by providing guidance messages when permissions are insufficient.

[0094] In one embodiment of the present invention, the artificial intelligence chat message generation unit 340 receives a search extension query term as input from the search extension query term generation unit 330. The artificial intelligence chat message generation unit 340 uses the search extension query term to select dialogue blocks that match the query term from the past chat history of the search extension chat room. Each selected dialogue block is assigned a response basis link that includes a unique identifier and dialogue time information, and the response basis link is included in the integrated input prompt configuration.

[0095] In another embodiment, the artificial intelligence chat message generation unit 340 integrates the user chat message and the search extension query term to generate a structured, integrated input prompt that can be interpreted by the artificial intelligence chat agent 311. For example, the artificial intelligence chat message generation unit 340 constructs a standard template such as "user request summary → relevant dialogue block → response justification link → expected response format". The artificial intelligence chat agent 311 receives the prompt based on this template as input, generates an artificial intelligence chat message that matches the user message, and the generated message is transmitted by the artificial intelligence chat message generation unit 340 to the artificial intelligence chat response unit 350 along with the response justification link.

[0096] In another embodiment of the present invention, the artificial intelligence chat message generation unit 340 dynamically adjusts importance weighting based on the recency of the dialogue blocks. The artificial intelligence chat message generation unit 340 assigns higher importance to messages that occurred most recently among past dialogue blocks containing the same keyword, and reflects this in the integrated input prompt configuration. For example, if the search extended query term is "Project A schedule," and recent meeting minutes messages and older related messages are searched simultaneously, the recent meeting minutes messages are configured as a higher-priority dialogue block. This allows the artificial intelligence chat agent 311 to generate responses that more accurately match the user's intent.

[0097] In another embodiment, if there are multiple searched dialogue blocks, the AI ​​chat message generation unit 340 generates a "compressed dialogue context" that is grouped and summarized by theme. The AI ​​chat message generation unit 340 clusters the searched dialogue blocks based on recurring themes, discussions on the same document basis, issue processing flows, etc., generates a summary sentence for each theme, and includes the summary sentence in the integrated input prompt. For example, if there are eight dialogue blocks on the same theme, "project schedule delay," the AI ​​chat message generation unit 340 integrates and summarizes the eight messages to create a compressed context such as "summary of discussion records on the causes of schedule delays over the past three weeks." The configuration that applies the integrated input prompt can maintain the structuring of input information even in enterprise collaboration environments with a large amount of dialogue data or in project environments that are operated for a long period of time, and can help the AI ​​chat agent 311 generate stable and consistent responses.

[0098] The artificial intelligence chat response unit 350 receives artificial intelligence chat messages generated by the artificial intelligence chat agent 311 and provides those messages through the user interface of the artificial intelligence chat room. During the message provision process, the artificial intelligence chat response unit 350 detects the user's editing intent and either directly modifies the artificial intelligence chat message according to the editing method selected by the user, or assists in correcting the user's input message and requesting regeneration.

[0099] The artificial intelligence chat response unit 350 provides an editing interface so that the user can directly modify the text or specific content of the artificial intelligence chat message. When the user directly edits the message, the artificial intelligence chat response unit 350 immediately reflects the modified message in the record of the artificial intelligence chat room and stores it as the final message, and in this process, no additional calculations may be performed by the artificial intelligence chat agent 311.

[0100] The artificial intelligence chat response unit 350 controls the system to generate a new artificial intelligence chat message by transmitting the modified user message back to the artificial intelligence chat agent 311 if the user wishes to modify the artificial intelligence chat message by correcting their own input message. This indirect editing-based regeneration method is used to reconstruct the existing artificial intelligence chat message in a corrected form that reflects the user's intention to modify it.

[0101] The artificial intelligence chat response unit 350 displays a link containing the basis for the response within the artificial intelligence chat message to the user. If the user selects the link, it provides the corresponding dialogue block in a separate window or pop-up to help the user confirm the basis for generating the artificial intelligence chat message.

[0102] The artificial intelligence chat response unit 350 controls the scope of editing by referring to user permissions, chat room settings, and security policies. If direct editing is restricted in a collaborative environment, the artificial intelligence chat response unit 350 activates and applies only the indirect editing method.

[0103] Additionally, the artificial intelligence chat response unit 350 applies different processing paths depending on the user's editing method. When the user directly modifies the artificial intelligence chat message, the artificial intelligence chat response unit 350 performs a direct editing path, immediately reflecting the modified message in the dialogue record. Conversely, when the user modifies their own chat message and requests a new artificial intelligence chat message, the artificial intelligence chat response unit 350 performs an indirect editing path, transmitting the modified user message to the artificial intelligence chat agent 311 and replacing the existing message with the regenerated message. This path division structure is configured to select a processing procedure appropriate to the user's editing type, ensuring both accuracy and processing efficiency in response editing.

[0104] In one embodiment of the present invention, the artificial intelligence chat response unit 350 displays an artificial intelligence chat message provided by the artificial intelligence chat message generation unit 340 on the user terminal 110. During the message display process, the artificial intelligence chat response unit 350 detects a user's editing request, and if the user attempts to directly modify the message, it switches the message content to an editable format to reflect the user's input. Furthermore, if the user modifies their own chat message, the artificial intelligence chat response unit 350 transmits the modified message to the artificial intelligence chat agent 311 and controls it to generate a new artificial intelligence chat message. The generated message is provided to the user in a manner that replaces the existing artificial intelligence chat message.

[0105] In another embodiment of the present invention, the artificial intelligence chat response unit 350 provides editing options that allow the user to adjust expressive attributes such as the style, tone, length, and language used in the response. If the user selects a request such as "make it more concise," "change to formal language," or "reduce the level of jargon," the artificial intelligence chat response unit 350 transmits the option to the artificial intelligence chat agent 311 to request regeneration and provides the artificial intelligence chat message in the adjusted format. In conclusion, the artificial intelligence chat response unit 350 can improve the efficiency of the conversation and the customizability of expression by allowing the user to easily adjust the artificial intelligence chat message to their desired form.

[0106] In another embodiment, the artificial intelligence chat response unit 350 manages the editing history of artificial intelligence chat messages and provides a function that allows the user to roll back to a previous version of the message. The artificial intelligence chat response unit 350 is configured to record the message version for each modification step and to redisplay that version in the artificial intelligence chat room when the user selects "View previous message" or "Restore original message". In conclusion, the artificial intelligence chat response unit 350 is useful in collaborative environments where long-term and repetitive work records are important, from the standpoint of response quality management and record keeping.

[0107] The control unit 360 performs a central control function that manages the data flow and operation sequence between the artificial intelligence chat room generation unit 310, the search extension interface unit 320, the search extension query word generation unit 330, the artificial intelligence chat message generation unit 340, and the artificial intelligence chat response unit 350, which constitute the artificial intelligence dialogue processing device 130. The control unit 360 receives event, request, and status information generated in each component and performs control procedures such as inter-module cooperation processing, step switching, internal parameter synchronization, and data transmission path setting based on this information.

[0108] The control unit 360 adjusts the operation of each component so that the overall operation of the artificial intelligence dialogue processing device 130 is performed stably according to defined procedures. For example, the control unit 360 controls the search extension chat room information set in the search extension interface unit 320 to be transmitted to the search extension query word generation unit 330 in a timely manner, or processes the artificial intelligence chat messages generated by the artificial intelligence chat message generation unit 340 to be accurately linked to the artificial intelligence chat response unit 350. The control unit 360 also monitors the processing status of each component and performs recovery procedures when an error occurs, or resets the data transmission path to ensure the stability of the entire device.

[0109] The control unit 360 performs control functions such as protocol management, message transmission timing adjustment, and internal stored parameter updates during communication and data transmission with the user terminal 110. The control unit 360 also manages control operations that occur throughout the entire device, such as creating new chat rooms, reflecting changes in search extension settings, and processing response regeneration requests, according to a consistent policy as needed. In conclusion, the control unit 360 functions as a core control layer that oversees the overall control procedure so that all components of the artificial intelligence dialogue processing device 130 can work together in coordination and in harmony.

[0110] Figure 4 is a flowchart illustrating the artificial intelligence dialogue processing method performed in the artificial intelligence dialogue processing device shown in Figure 1.

[0111] In Figure 4, the artificial intelligence dialogue processing device 130 performs an artificial intelligence chat processing procedure, which includes an artificial intelligence chat room generation step S410, a search extension interface step S430, a search extension query word generation step S450, an artificial intelligence chat message generation step S470, and an artificial intelligence chat response step S490.

[0112] In the AI ​​chat room generation step S410, the AI ​​dialogue processing device 130 generates an AI chat room in which an AI chat agent 311 capable of responding to user dialogues participates. The AI ​​dialogue processing device 130 receives a chat room generation request from the user terminal 110 and configures an AI chat room in accordance with the request, registering the AI ​​chat agent 311 as an initial participant.

[0113] In the search extension interface step S430, the artificial intelligence dialogue processing device 130 generates a search extension interface for setting up a search extension chat room through the search extension interface unit 320. The search extension interface provides a UI environment for setting up a specific chat room from among multiple chat rooms in which a user participates as a search extension chat room, and receives a search interval, a search theme, and search extension agent settings as input to configure the search target referenced in the search extension query word generation step S450 and the artificial intelligence chat message generation step S470.

[0114] In the search extension query term generation step S450, the artificial intelligence dialogue processing device 130 analyzes the user chat message through the search extension query term generation unit 330 and generates a search extension query term for searching the past chat history of the search extension chat room. The artificial intelligence dialogue processing device 130 determines whether the user message belongs to a search theme set in the search extension interface, and decides whether or not to generate a search extension query term according to the determination result. If query term generation fails, it evaluates the possibility of utilizing chat rooms with similar themes.

[0115] In the AI ​​chat message generation step S470, the AI ​​dialogue processing device 130 searches the past chat history of the search-enhanced chat room by applying the generated search-enhanced query term, and constructs an integrated input prompt based on the selected dialogue block and user chat message as search results. The constructed integrated input prompt is a standardized input set consisting of the user message, search-enhanced query term, selected dialogue block, response basis link, and related metadata, which is input to the AI ​​chat agent 311 and used to generate the final AI chat message.

[0116] In the artificial intelligence chat response step S490, the artificial intelligence dialogue processing device 130 provides the artificial intelligence chat message generated by the artificial intelligence chat agent 311 to the artificial intelligence chat room. The artificial intelligence dialogue processing device 130 displays the artificial intelligence chat message, which includes a response justification link, to the user and provides editing support so that the user can directly modify the message content or indirectly correct the response by recreating the user message.

[0117] In conclusion, the artificial intelligence chat processing procedure in Figure 4 has a structure in which the artificial intelligence chat room generation step S410, the search extension interface step S430, the search extension query word generation step S450, the artificial intelligence chat message generation step S470, and the artificial intelligence chat response step S490 are sequentially linked, and it shows the overall operation process for the artificial intelligence dialogue processing device 130 to improve the search base response and improve the dialogue flow based on user utterances.

[0118] Figures 5a and 5b are diagrams illustrating the process of generating and utilizing dialogue data in the artificial intelligence dialogue processing device shown in Figure 1.

[0119] Figure 5a shows the user interaction data processing and storage structure, and Figure 5b shows the current data utilization interface structure in the chat room.

[0120] In Figures 5a and 5b, the artificial intelligence dialogue processing device 130 receives user dialogue data 510 as input and generates and manages chat message data 540a and theme-based classification data 540b using a user dialogue assistant model 530 configured to execute an artificial intelligence chat agent 311. The generated data is used as basis data for contextual analysis of the chat room 540 selected by the user, generation of search extension queries, and generation of artificial intelligence chat messages.

[0121] Figure 5a is a diagram illustrating the process by which the artificial intelligence dialogue processing device 130 generates and stores chat message data 540a and theme-based classification data 540b based on user dialogue data.

[0122] The artificial intelligence dialogue processing device 130 receives user dialogue data 510 as input and generates chat message data 540a and theme-based classification data 540b using a user dialogue assistant model 530 configured as an interlocking model for the artificial intelligence chat agent 311. The chat message data 540a includes the original text of the user's utterance, and the theme-based classification data 540b is configured as a data structure for classifying and storing specific themes, business units, and category information included in the user dialogue.

[0123] The generated chat message data 540a and theme-based classification data 540b are stored in the database 150, and the stored data is referenced in the contextual analysis, search extension query generation, and artificial intelligence chat message generation processes of the chat room 540 selected by the user.

[0124] Furthermore, the artificial intelligence dialogue processing device 130 analyzes the user dialogue data 510 in real time through the user dialogue assistant model 530, automatically updates the generated chat message data 540a and theme-based classification data 540b, and manages them so that they can be queried and searched in the chat rooms 540 selected by the user, i.e., the chat rooms targeted for search enhancement.

[0125] Figure 5a shows the data processing flow in which user dialogue data 510 is processed by the user dialogue assistant model 530, the results are stored in the form of chat message data 540a and theme-based classification data 540b, and referenced by the search extension query term generation unit 330 and the artificial intelligence chat message generation unit 340.

[0126] Figure 5b is a diagram illustrating the process by which the artificial intelligence dialogue processing device 130 currently provides dialogue data retrieval and assistance messages, centered around the chat room 550.

[0127] Currently, chat room 550 is configured as an active dialogue interface where user input messages are processed in real time, and includes an area where AI chat messages generated by the AI ​​chat agent 311 are displayed. Currently, chat room 550 is selectively connected to chat message data 540a and theme-based classification data 540b stored in the chat room 540 selected by the user.

[0128] When a user selects a specific sentence or theme element in the current chat room 550, the artificial intelligence dialogue processing device 130 queries the corresponding data 540a and 540b stored in the database 150 to provide the original text or related information. For example, if a user enters the query "Show me the past discussions of Project X" in the current chat room 550, the artificial intelligence dialogue processing device 130 searches the chat message data 540a of the chat room 540 selected by the user and provides the main content to the current chat room 550 in a summarized or refined artificial intelligence chat message format.

[0129] Figure 5b shows that the current chat room 550 is currently performing real-time dialogue and past dialogue data information retrieval functions simultaneously, and illustrates the overall operation flow in which the artificial intelligence dialogue processing device 130 works in conjunction with the participating chat room 540 and the database 150 to provide a search-enhanced dialogue assistance function.

[0130] In conclusion, Figures 5a and 5b show both (1) the generation, classification, and storage processing structure of user dialogue data 510 (Figure 5a) and (2) the UI processing structure in which search extension infrastructure queries and artificial intelligence responses are provided, centered on the current chat room 550 (Figure 5b). This explains the overall processing flow in which the artificial intelligence dialogue processing device 130 organically links the user dialogue assistant model 530, the participating chat room 540, the database 150, and the current chat room 550 to perform the search extension infrastructure artificial intelligence chat function.

[0131] Figures 6a to 6c show examples of user interface (UI) configurations implemented in the artificial intelligence dialogue processing device shown in Figure 1.

[0132] Figure 6a shows the AI ​​chat room screen and the search extension options settings screen, Figure 6b shows the AI ​​chat room list screen and the current chat room screen, and Figure 6c shows the list of participating chat rooms screen.

[0133] Figure 6a shows the configuration of the artificial intelligence chat room screen and the search extension options settings screen.

[0134] Here, the AI ​​chat room screen refers to the user interface screen that displays the current chat room 550 in which the user is actually having a conversation.

[0135] In Figure 6a, the current chat room 550 is configured as a dialogue interface in which the artificial intelligence chat agent 311 participates in and provides responses to a specific dialogue theme selected or activated by the user. In the current chat room 550, chat messages entered by the user and artificial intelligence chat messages generated by the artificial intelligence chat agent 311 are displayed in real time, and if the search extension function is enabled, data from the participating chat room 540 is referenced and reflected in the message generation process.

[0136] Figure 6b shows examples of the AI ​​chat room list screen and the current chat room screen.

[0137] In Figure 6b, the artificial intelligence chat room generation unit 310 provides multiple artificial intelligence chat histories in list form, and the user selects one of the chat histories from the list to set the current chat room 550 based on the selected chat history. The set current chat room 550 is displayed through the artificial intelligence chat room screen, and the conversation proceeds with the artificial intelligence chat agent 311 automatically joining.

[0138] Figure 6c shows the configuration of the chat room list screen.

[0139] In Figure 6c, the participating chat rooms 540 consist of past chat rooms or user-specified chat rooms that can be referenced for generating artificial intelligence responses in the search extension platform, and multiple participating chat rooms 540 are provided in a list format. Users can select one or more participating chat rooms 540 from this list as targets for search extension, and the chat message data contained in the selected participating chat rooms 540 is used as reference data in the process of generating artificial intelligence chat messages currently being generated in the chat room 550.

[0140] Figures 6a to 6c show examples of UI configurations implemented by the artificial intelligence dialogue processing device 130 on the user terminal 110. Figure 6a shows the artificial intelligence chat room screen and search extension option setting screen that display the current chat room 550. Figure 6b shows a screen that displays the artificial intelligence chat room list screen and the current chat room 550 in conjunction. Figure 6c shows a list screen of participating chat rooms 540 used for search extension.

[0141] This configuration explains how the artificial intelligence dialogue processing device 130 currently links the chat room 550, the participating chat room 540, and the search extension settings UI in order to provide artificial intelligence responses for the search extension platform.

[0142] 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. [Explanation of symbols]

[0143] 100 AI Chat Processing Systems 110 User terminals 130 Artificial Intelligence Dialogue Processing Device 150 databases 210 processors 230 memory 250 User Input / Output Section 270 Network Input / Output Section 290 Communication port section 310 Artificial Intelligence Chat Room Generation Unit 311 Artificial Intelligence Chat Agent 320 Search Extension Interface Section 330 Search Extension Query Term Generation Unit 340 Artificial Intelligence Chat Message Generation Unit 350 AI Chat Response Unit 360 Control Unit 510 User interaction data 530 User Dialogue Assistant Model 540 users selected a chat room to join. 540a Chat message data 540b Thematic Classification Data 550 Currently chat rooms

Claims

1. An AI chat room generation unit generates an AI chat room in which an AI chat agent capable of responding to user conversations participates, A search extension interface is generated for setting up a search extension chat room to be used in the aforementioned artificial intelligence chat room, and the search extension chat room consists of a search extension interface unit and the user's participating chat rooms. A search extension query term generation unit receives the user's chat message as input in the artificial intelligence chat room and generates a search extension query term used to search the past chat history of the search extension chat room. An artificial intelligence chat message generation unit generates an artificial intelligence chat message by searching the past chat history of the search-extended chat room using the search-extended query term, and inputting the user's chat message and the searched past chat history as an integrated input prompt to the artificial intelligence chat agent. An artificial intelligence dialogue processing device characterized by including an artificial intelligence chat response unit that provides the artificial intelligence chat message to the artificial intelligence chat room as a response to the user's chat message by the artificial intelligence chat agent.

2. The aforementioned artificial intelligence chat agent, The artificial intelligence dialogue processing device according to claim 1, characterized in that it is invited by the user to simultaneously participate in other chat rooms of the user.

3. The aforementioned search extension interface unit is: The artificial intelligence dialogue processing device according to claim 1, characterized in that a search interval or search theme for the search extension chat room is set via the search extension interface.

4. The aforementioned search extension interface unit is: The artificial intelligence dialogue processing device according to claim 1, characterized in that it connects a search-enhanced agent that controls searching and responding to an external database via the search-enhanced interface.

5. The aforementioned search extension interface unit is: The artificial intelligence dialogue processing device according to claim 1, characterized in that a new artificial intelligence chat room is generated when the search interval or search theme for the search-enhanced chat room is changed, or when the search-enhanced agent is changed via the search-enhanced interface.

6. The aforementioned search extension interface unit is: The artificial intelligence dialogue processing device according to claim 1, characterized in that it adds or removes the search extension chat room via the search extension interface.

7. The aforementioned search extension query term generation unit, The artificial intelligence dialogue processing device according to claim 1, characterized in that it determines whether the user's chat message belongs to the search theme of the search-enhanced chat room, and determines whether or not to generate the search-enhanced query term.

8. The aforementioned search extension query term generation unit, The artificial intelligence dialogue processing device according to claim 7, characterized in that, if the generation of the search extended query term fails, it determines whether or not to provide assistance from other artificial intelligence chat agents present in an artificial intelligence chat room set to another search theme related to the search extended query term.

9. The aforementioned artificial intelligence chat message generation unit, The artificial intelligence dialogue processing device according to claim 1, characterized in that it selects dialogue blocks corresponding to the search extension query term from the past chat history of the search extension chat room, assigns response basis links to the selected dialogue blocks, and uses the dialogue blocks and the response basis links to generate the integrated input prompt.

10. The aforementioned artificial intelligence chat message generation unit, The artificial intelligence dialogue processing device according to claim 9, characterized in that it provides the artificial intelligence chat message along with the response basis link through the artificial intelligence chat agent, and allows the user to confirm the content of the dialogue block through user interaction.

11. The aforementioned artificial intelligence chat response unit is The artificial intelligence dialogue processing device according to claim 1, characterized in that, in the process of providing the artificial intelligence chat message, the artificial intelligence chat message is directly modified by the user's editing control.

12. The aforementioned artificial intelligence chat response unit is The artificial intelligence dialogue processing device according to claim 1, characterized in that, in the process of providing the artificial intelligence chat message, the user's chat message is modified by the user's editing control, and the artificial intelligence chat agent indirectly modifies the artificial intelligence chat message.

13. An artificial intelligence dialogue processing method performed in an artificial intelligence dialogue processing device, An AI chat room generation step that generates an AI chat room in which an AI chat agent capable of responding to user conversations participates, A search extension interface is generated for setting up a search extension chat room to be used in the aforementioned artificial intelligence chat room, and the search extension chat room is composed of the user's participating chat rooms, and the search extension interface step is... A search extension query term generation step in which the chat messages of the user are received as input in the artificial intelligence chat room and search extension query terms are generated to be used to search the past chat history of the search extension chat room, An artificial intelligence chat message generation step involves searching the past chat history of the search-enhanced chat room using the search-enhanced query term, and generating an artificial intelligence chat message by inputting the user's chat message and the searched past chat history as integrated input prompts to the artificial intelligence chat agent. An artificial intelligence dialogue processing method characterized by including an artificial intelligence chat response step of providing the artificial intelligence chat message to the artificial intelligence chat room as a response to the user's chat message by the artificial intelligence chat agent.

Images