Method and system for providing interactive artificial intelligence response services

Abstract

The present invention relates to a method and system for providing an interactive artificial intelligence response service. More specifically, the method for providing an interactive artificial intelligence response service according to the present invention may include the steps of: receiving a user query via a service page; identifying location information from the user query; searching for satellite imagery related to the user query from at least one database using the user query and the location information; performing analysis related to the user query using the satellite imagery; generating a prompt using at least one of the user query, the location information, the analysis results from the analysis, and the satellite imagery; processing the prompt as input to a large-scale language model (LLM); obtaining an answer to the user query from the large-scale language model; and providing the answer to the user query to the service page.

Description

Technical Field

[0001] The present invention relates to a method and system for providing an interactive artificial intelligence answering service.

Background Art

[0002] Recently, with the development of artificial intelligence, especially deep learning that extracts data characteristics using a deep neural network structure, there has been a rapid increase in cases of excellent achievements in various fields such as speech recognition, image recognition, natural language processing, and autonomous driving.

[0003] With the development of such deep learning technology, recently, generative artificial intelligence (Generative AI) technology has attracted attention. More specifically, a generative artificial intelligence model can generate new data in various forms such as text, images, and sounds from given data, which provides a different level of applicability from simply classifying or predicting existing data.

[0004] That is, it has become possible to automatically generate articles, images, sounds, etc. that were previously generated by humans by using a generative artificial intelligence model, and services using generative artificial intelligence (e.g., ChatGPT) show initiative and accuracy differentiated from existing chatbot services and have attracted great interest worldwide.

[0005] On the other hand, with the development of satellite technology, the utilization of satellites has become increasingly important in fields such as earth observation, weather prediction, and communication. Satellites can collect various forms of data such as high-quality images and videos, geographical information, and environmental data through sensors and cameras and provide various information about the earth in real time.

[0006] With the advancement of satellite technology, there has been a surge in research and initiatives recently to apply artificial intelligence technology to satellite data and utilize it in various fields. The analysis and processing of the vast amounts of images, videos, and data collected from satellites are extremely important, and high-resolution satellite imagery, in particular, is proving useful in a wide range of fields, including agriculture, disaster management, urban planning, and environmental monitoring.

[0007] Thus, research into applying artificial intelligence technology to the satellite field is actively underway, and this has created a need for methods to provide various satellite-related services using generative artificial intelligence models. [Overview of the project] [Problems that the invention aims to solve]

[0008] The present invention provides a method and system for providing an interactive artificial intelligence response service that can efficiently analyze satellite data and provide users with information that can be utilized in various fields.

[0009] More specifically, the present invention provides a method and system for providing an interactive artificial intelligence response service that can analyze satellite data at the request of a user and provide the user with the information they desire via an interactive artificial intelligence chatbot.

[0010] Furthermore, the present invention provides a method and system for providing an interactive artificial intelligence response service that can provide users with satellite images collected from satellites.

[0011] In other words, the present invention provides a method and system for providing an interactive artificial intelligence response service that can sell high-resolution satellite imagery to users. [Means for solving the problem]

[0012] To solve the above-mentioned problems, the interactive artificial intelligence response service provision method according to the present invention may include the steps of: receiving a user query via a service page; identifying location information from the user query; searching for satellite imagery related to the user query from at least one database using the user query and the location information; performing analysis related to the user query using the satellite imagery; generating a prompt using at least one of the user query, the location information, the analysis results from the analysis, and the satellite imagery; processing the prompt as input to a large-scale language model (LLM); obtaining an answer to the user query from the large-scale language model; and providing the answer to the user query to the service page.

[0013] Furthermore, the service page may include a first area for receiving the user query and a second area for providing the satellite imagery retrieved in response to the user query.

[0014] Furthermore, the first region may provide text-based information related to the user query, and the second region may provide image-based information related to the user query.

[0015] Furthermore, a map image is displayed in the second area, and when the satellite image is retrieved in response to the user query, the satellite image may overlap at least a portion of the map image.

[0016] Furthermore, the area in the map image where the satellite image overlaps may be identified based on the location information identified for searching the satellite image.

[0017] Furthermore, the system may further include the steps of: receiving user input specifying a particular area in the satellite imagery provided to the second area; receiving additional user queries related to the particular area in the first area; and generating a response to the additional user queries using the satellite imagery of the particular area, metadata related to the particular area, and at least one of the additional user queries.

[0018] Furthermore, the step of generating a response to the additional user query may include: analyzing the user intent of the additional user query; identifying a function corresponding to the user intent and selecting a specific analysis model that performs the identified function; processing satellite imagery of the specific region as input to the specific analysis model; obtaining output from the specific analysis model that corresponds to the user intent; generating a response to the additional user query using the output; and providing the response to the additional user query to at least one of the first and second regions.

[0019] Furthermore, the response to the additional user query provided to the second region may include highlighting at least a portion of the satellite image of the particular region provided to the second region.

[0020] Furthermore, the service page may provide at least one tool icon for receiving user input for the satellite imagery provided in the second area, wherein the tool icons are linked to different functions, with a first tool icon linked to a first selection function for selecting at least a portion of the satellite imagery according to a first criterion, and a second tool icon linked to a second selection function for selecting at least a portion of the satellite imagery according to a second criterion different from the first criterion.

[0021] Furthermore, upon receiving the user query, the system may further include the steps of: processing the user query as input to the large-scale language model; selecting a model from the large-scale language model that performs the analysis corresponding to the user query; inputting the satellite image as input to the selected model; and receiving the analysis results of the satellite image corresponding to the user query from the selected model.

[0022] Furthermore, in the step of generating the prompt, the analysis results received from the selected model may be included in the prompt, and the large-scale language model may use the analysis results and the satellite imagery to generate an answer to the user query.

[0023] On the other hand, the method for providing an interactive artificial intelligence response service according to the present invention further includes the steps of: receiving a user query via a service page including a first domain and a second domain; identifying a document to be used to generate a response to the user query; processing the user query and the identified document as input to a large-scale language model; obtaining a response to the user query from the large-scale language model; and providing the response to the service page, wherein the first domain of the service page is provided with the response, and the second domain of the service page is provided with information regarding the identified document.

[0024] On the one hand, the interactive artificial intelligence answer service providing system according to the present invention includes a memory and at least one processor, and the memory and the processor cooperate to receive a user query via a service page, identify location information from the user query, use the user query and the location information to search for satellite images related to the user query from at least one database, perform an analysis related to the user query using the satellite images, generate a prompt using at least one of the user query, the location information, the analysis result from the analysis, and the satellite images, process the prompt as an input to a large language model (LLM), obtain an answer to the user query from the large language model, and provide the answer to the user query on the service page.

[0025] On the other hand, the program according to the present invention is a program that is executed by one or more processes in an electronic device and can be stored in a computer-readable recording medium. The program includes steps of receiving a user query via a service page, identifying location information from the user query, searching for satellite images related to the user query from at least one database using the user query and the location information, performing an analysis related to the user query using the satellite images, generating a prompt using at least one of the user query, the location information, the analysis result from the analysis, and the satellite images, processing the prompt as an input to a large language model (LLM), obtaining an answer to the user query from the large language model, and providing the answer to the user query on the service page, and may include instruction words for executing these steps.

Effect of the Invention

[0026] As described above, according to the method and system for providing an interactive artificial intelligence answering service according to the present invention, it is possible to generate an answer with an improved hallucination phenomenon by utilizing the retrieved document and provide it to the user. Thereby, in the present invention, it is possible to contribute to reducing the provision of incorrect information and providing highly reliable information to the user.

[0027] Also, according to the method and system for providing an interactive artificial intelligence answering service according to the present invention, it is possible to analyze satellite data related to a user's query and provide a response to the user's query together with visual information. Thereby, the user can more intuitively confirm the response to the query.

[0028] Furthermore, according to the method and system for providing an interactive artificial intelligence answering service according to the present invention, by integrally analyzing various data including satellite images corresponding to a user's query and providing various information required by the user, the user can utilize the provided information in various fields such as agriculture, environmental monitoring, urban planning, etc., and can improve efficiency across various industries.

[0029] Furthermore, according to the method and system for providing an interactive artificial intelligence answering service according to the present invention, it is possible to provide an answer that conforms to the user's query by utilizing the information specified according to the user's request. That is, in the present invention, it is possible to provide accurate and highly reliable information corresponding to the user's requirements without performing a complicated information search process.

Brief Description of the Drawings

[0030] [Figure 1] It is a conceptual diagram for explaining an interactive artificial intelligence answering service providing system according to the present invention. [Figure 2A] It is a conceptual diagram for explaining various modes provided in an interactive artificial intelligence answering service providing system according to the present invention. [Figure 2B] It is a conceptual diagram for explaining various modes provided in an interactive artificial intelligence answering service providing system according to the present invention. [Figure 3] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 4A] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 4B] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 4C] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 4D] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 4E] This is a conceptual diagram illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. [Figure 5] This is a flowchart illustrating the method for providing an interactive artificial intelligence response service according to the present invention. [Figure 6A] This is a conceptual diagram illustrating the method for providing an interactive artificial intelligence response service according to the present invention. [Figure 6B] This is a conceptual diagram illustrating the method for providing an interactive artificial intelligence response service according to the present invention. [Figure 7] This is a conceptual diagram illustrating a method for selling satellite imagery and providing paid services related to one embodiment of the present invention. [Figure 8] This is a conceptual diagram illustrating a method for selling satellite imagery and providing paid services related to one embodiment of the present invention. [Figure 9] This is a conceptual diagram illustrating a method for selling satellite imagery and providing paid services related to one embodiment of the present invention. [Figure 10] This is a conceptual diagram illustrating a method for selling satellite imagery and providing paid services related to one embodiment of the present invention. [Figure 11]This is a conceptual diagram illustrating a method for selling satellite imagery and providing paid services related to one embodiment of the present invention. [Modes for carrying out the invention]

[0031] The embodiments disclosed herein will be described in detail below with reference to the accompanying drawings, but regardless of the reference numerals used in the drawings, identical or similar components will be given the same reference numerals, and redundant descriptions thereof will be omitted. The suffixes “module” and “part” used for components in the following description are added or mixed for the sake of ease of writing the specification and do not have any distinguishing meaning or role in themselves. Furthermore, when describing the embodiments disclosed herein, if it is determined that a detailed description of the relevant prior art would obscure the gist of the embodiments disclosed herein, such detailed description will be omitted. In addition, the accompanying drawings are intended solely to facilitate the understanding of the embodiments disclosed herein, and the technical ideas disclosed herein should not be limited by the accompanying drawings and should be understood to include all modifications, equivalents and substitutions that fall within the concept and technical scope of the present invention.

[0032] Terms including ordinal numbers such as "1st," "2nd," etc., may be used to describe various components, but the components are not limited to those defined by these terms. These terms are used solely to distinguish one component from another.

[0033] When it is stated that one component is “connected” or “linked” to another component, it should be understood that it may be directly connected or linked to the other component, but there may also be another component between them. On the other hand, when it is stated that one component is “directly connected” or “directly linked” to another component, it should be understood that there is no other component between them.

[0034] A singular expression includes plural forms unless otherwise clearly indicated in the context.

[0035] In this application, terms such as “includes” or “having” are intended to specify the presence of features, figures, steps, actions, components, parts, or combinations thereof as described in the specification, and should be understood not to preemptively exclude the possibility of the presence or addition of one or more other features, figures, steps, actions, components, parts, or combinations thereof.

[0036] This invention relates to a method and system for providing an interactive artificial intelligence response service. The interactive artificial intelligence response service provision system according to the present invention generates responses based on generative artificial intelligence (AI) or a large language model (LLM), and is also called an interactive artificial intelligence response service provision system based on a large language model. However, for the sake of convenience of explanation, it will be referred to as the "response service provision system" below.

[0037] The answer service provision system according to the present invention may be a system that, upon receiving a user query (or query) related to a satellite and / or satellite imagery (or image), searches a database (or storage) for documents related to the user query, and uses the retrieved documents and a large-scale language model (LLM) to generate and provide an answer that matches the user's query intent.

[0038] Here, "satellite imagery" refers to images recorded by sensors (or detectors (e.g., image sensors, cameras, etc.)) mounted on artificial satellites, and in a broader sense, it may also include space photographs (CLFC) taken by cameras.

[0039] Furthermore, the answer service provision system according to the present invention may also be a system that, upon receiving a user query related to a satellite and / or satellite imagery, searches a database (or storage) for satellite imagery related to the user query, and uses the analysis results obtained from analyzing the retrieved satellite imagery and a large-scale language model to generate and provide an answer that matches the user's query intent. More specific details on this will be described later.

[0040] On the other hand, the answer service provision system according to the present invention aims to analyze satellite data at the user's request and provide the information desired by the user (for example, information corresponding to the user's query intent) via an interactive artificial intelligence chatbot.

[0041] The following will provide a more detailed explanation with reference to the attached drawings. Figure 1 is a conceptual diagram illustrating the interactive artificial intelligence response service provision system according to the present invention, and Figures 2a, 2b, 3, 4A, 4B, 4C, 4D, and 4E are conceptual diagrams illustrating the various modes provided in the interactive artificial intelligence response service provision system according to the present invention. Figure 5 is a flowchart illustrating the method for providing the interactive artificial intelligence response service according to the present invention, and Figures 6a and 6b are conceptual diagrams illustrating the method for providing the interactive artificial intelligence response service according to the present invention. Furthermore, Figures 7, 8, 9, 10, and 11 are conceptual diagrams illustrating a satellite image sales method and a paid service provision method according to one embodiment of the present invention.

[0042] On the other hand, as shown in Figure 1, the response service provision system 100 according to the present invention may include at least one of the following: an input unit 110, an output unit 120, a communication unit 130, a storage unit 140, and a control unit 150.

[0043] Although not shown in the figures, the answer service provision system 100 according to the present invention may include one or more processors, such processors may include one or more general-purpose processors and / or one or more special-purpose processors (e.g., digital signal processors, tensor processing units (TPUs), graphics processing units (GPUs), neural network processing units (NPUs), application-specific integrated circuits, application-specific semiconductors (ASICs), etc.). One or more processors may be configured to execute instruction words, computer-readable instructions, and / or other instructions described herein that are stored (or included) in the memory unit 140. The answer service provision system and method according to the present invention allows the memory and at least one processor to cooperate in performing data processing described later. The processor can perform a series of calculations and data processing using the data and information stored in the memory. In this case, the memory may be configured as the memory unit 140.

[0044] On the other hand, the input unit 110 is a means of data input and may be configured in various ways. For example, the input unit 110 may be configured to receive user input. The input unit 110 may be configured to receive user input from the user terminal 10. Here, "receiving input" may mean receiving an input signal (or selection signal) corresponding to the user input, based on the user input being made through the configuration of the input unit provided in the user terminal 10.

[0045] The input unit 110 is also called a user interface module. The input unit 110 may include a touchscreen, computer mouse, keyboard, keypad, touchpad, trackball, joystick, voice recognition module, or other similar device. However, the present invention does not limit the type of input unit 110. Furthermore, in the present invention, the input unit 110 does not necessarily mean hardware means, but can be understood as a channel for receiving input from the user.

[0046] Here, user input may include documents, text, images (or videos), audio, etc. In this case, the response service provision system 100 may further include a module that converts audio to text.

[0047] Next, the output unit 120 can output information through the configuration of an output unit (e.g., display unit, touchscreen, speaker, etc.) provided on the user terminal 10 which is linked to the answer service provision system 100 according to the present invention. For example, the output unit 120 can output a page 1000 (or service page) linked to the answer service provision system 100 according to the present invention to the display unit of the user terminal 10. Furthermore, the output unit 120 does not necessarily mean hardware means, but can be understood as a channel for outputting results to the user.

[0048] Next, the communication unit 130 may be connected wirelessly or via a wired network to the user terminal 10, LLM server 20 (or artificial intelligence server), satellite image DB 30, central server, external server, device, and at least one network, and may be configured to receive or transmit overall data and information necessary for the operation of the response service provision system 100 according to the present invention.

[0049] In this case, the communication unit 130 may include one or more communication modules to enable wireless and / or wired communication between the response service provision system 100 and the user terminal 10, between the response service provision system 100 and the artificial intelligence server 20, and between the response service provision system 100 and the satellite image DB 30. The communication unit 130 may also include one or more communication modules to connect the response service provision system 100 to one or more networks.

[0050] Here, the user terminal 10 may include at least one of the following: mobile phone, smartphone, notebook computer, laptop computer, slate PC, tablet PC, ultrabook, desktop computer, digital broadcasting terminal, PDA (personal digital assistant), PMP (portable multimedia player), navigation system, and wearable device (e.g., smartwatch, smart glass, HMD (head mounted display)).

[0051] Furthermore, the communication unit 130 can support various communication methods depending on the communication standard of the device it communicates with.

[0052] For example, the communication unit 130 supports WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Wi-Fi (Wireless Fidelity) Direct, DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband), WiMAX (World Interoperability for Microwave Access), HSDPA (High Speed ​​Downlink Packet Access), HSUPA (High Speed ​​Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), 5G (5th Generation Mobile Telecommunication), and Bluetooth (registered trademark). TMIt may be configured to communicate with a target using at least one of the following technologies: RFID (Radio Frequency Identification), Infrared Data Association (IrDA), UWB (Ultra-Wideband), ZigBee, NFC (Near Field Communication), Wi-Fi Direct, or Wireless USB (Wireless Universal Serial Bus).

[0053] On the other hand, the storage unit 140 plays a role in storing various data related to the present invention and may include one or more non-temporary computer-readable storage media that can be read and / or accessed by at least one of the one or more processors.

[0054] One or more computer-readable storage media may include volatile and / or non-volatile storage components such as optical, magnetic, organic, or other memory or disk storage devices. In some examples, the storage unit 140 can be embodied using a single physical device (e.g., one optical, magnetic, organic, or other memory or disk storage device), while in other examples, the storage unit 140 can be embodied using two or more physical devices.

[0055] The storage unit 140 may include computer-readable instructions and additional data. The storage unit 140 may include storage necessary to perform at least some of the methods, scenarios and techniques described herein and / or at least some of the functions of the apparatus and network.

[0056] Furthermore, at least a portion of the storage unit 140 may be cloud storage or a cloud server. The storage unit 140 may store at least a portion of the data corresponding to user input received from the input unit 110 and the training data.

[0057] In other words, the memory unit 140 can be understood as having no physical constraints, as long as it is a space that stores the information necessary for the operation of the answer service provision system 100 according to the present invention.

[0058] On the other hand, the control unit 150 can play a role in controlling the overall operation of the response service provision system 100 according to the present invention. The control unit 150 can process signals, data, information, etc. that are input or output via the above-mentioned components, or can perform a series of data processing to provide or process appropriate information and functions to the user.

[0059] When the control unit 150 receives a user query related to a satellite and / or satellite imagery, it can retrieve at least one document related to the user query from the storage unit 140 and / or at least one storage and / or document database, and use the retrieved document and the large-scale language model 160 to generate an answer to the user query and provide it to the user. In this case, the control unit 150 can provide the user with at least one document used (or referenced) to generate the answer to the user query, along with the answer.

[0060] Furthermore, when the control unit 150 receives a user query related to a satellite and / or satellite imagery, it can retrieve at least one satellite image related to the user query from the storage unit 140 and / or at least one storage and / or satellite imagery DB 30, and use the analysis results obtained from analyzing the retrieved satellite imagery and the large-scale language model 160 to generate an answer to the user query and provide it to the user. In this case, the control unit 150 can provide the user with at least one of the satellite imagery, location information, and map used (or referenced) to generate the answer to the user query, along with the answer.

[0061] In one embodiment, the analysis results obtained from the analysis of satellite imagery may relate to at least one of the following: search results for satellite imagery and coordinates, predicted satellite orbits, query results for satellite position information, calculation (or determination) of spectral indices, object detection, and segmentation results. However, in the present invention, the analysis results obtained from the analysis of satellite imagery are not necessarily limited to the examples described above.

[0062] In this regard, the control unit 150 may include at least one model (or module) used to generate a response to a user query (or analyze satellite imagery based on a user query). In the present invention, the model may be configured to perform a specified function and output a result value based on the execution result. In the present invention, the model may also be understood as a "tool." In this case, the tool may include functions or algorithms for performing a specific task.

[0063] In one embodiment, if a user query includes information about vehicle detection in satellite imagery, the control unit 150 can use the first model 170 to detect vehicles in the satellite imagery and provide the user with a response that includes information about the number of vehicles detected. The first model 170 may be an object detection model that has been pre-trained to detect at least one object.

[0064] In another embodiment, if a user query includes content relating to the NDVI (Vegetation Index) analysis of a specific area (or region) included in satellite imagery, the control unit 150 can use the second model 180 to analyze the vegetation index of the specific area and provide the user with a response that includes content relating to the vegetation index analysis results. The second model 180 may be a spectral index analysis model that has been pre-trained to calculate information related to spectral index analysis results by analyzing the imagery.

[0065] However, the models used for generating responses to user queries in the present invention are not necessarily limited to the models described above, and may include other models in addition to the first model 170 and the second model 180. Needless to say, the models used for generating responses to user queries in the present invention can be one or more, and can be varied in various ways, depending on the circumstances.

[0066] On the other hand, the response service provision system 100 according to the present invention can be embodied in the form of various platforms such as applications, software, and websites. For the convenience of explanation, this specification does not limit the form in which the response service provision system 100 is embodied to any one of these forms. The response service provision system 100 according to the present invention is also called a "conversational artificial intelligence chatbot service provision platform."

[0067] The user described above may have a user account pre-registered with the response service provision system 100 according to the present invention. In this case, the account may be generated via a page (or screen) linked to the response service provision system 100. Alternatively, the account may be generated in at least one other system linked to the response service provision system 100. However, in this specification, without distinguishing the system in which the user account was issued, any account that can use the various services (or functions) provided by the artificial intelligence chatbot service provision system 100 according to the present invention is referred to as an "account pre-registered with the response service provision system 100".

[0068] On the other hand, the response service provision system 100 can provide the user with multiple modes, each containing different functions. Here, the multiple modes may include a first mode and a second mode.

[0069] A graphic object may be provided in one area of ​​the service page 1000, which is linked to the response service provision system 100, allowing the user terminal 10 to select one of several modes.

[0070] For example, as shown in Figure 3, the control unit 150 can provide (or display or output) a graphic object 312a in a region 310 of the service page 1000 for selecting one of several modes.

[0071] The control unit 150 can output a GUI (Graphic User Interface) configured to select either the first mode or the second mode based on the selection of the graphic object 312a from the user terminal 10. Furthermore, the control unit 150 can control the system so that the mode corresponding to the user's selection is activated when either the first mode or the second mode is selected by the user.

[0072] In one embodiment, the output GUI may be a dropdown menu. However, it goes without saying that the GUI for selecting a specific mode is not necessarily limited to this and may be output in various forms.

[0073] In this case, the activation (or switching) of the first mode and / or the second mode can be understood as being based on the user's selection. Furthermore, the activation of the first mode and / or the second mode may be controlled autonomously by the control unit 150 based on the user account history information. The following describes several modes in more detail.

[0074] First, let's consider the first mode among the multiple modes.

[0075] The first mode (or knowledge mode) may also be called "Knowledge Mode," and may be configured to retrieve documents related to the user query from the memory unit 140 and / or at least one storage and / or document DB when a user query related to satellites and / or satellite imagery is received, and to generate and provide an answer to the user query using the retrieved documents and the large language model 160.

[0076] In one embodiment, the first mode may be configured to search for relevant content from a document database when a user needs information or documents related to a specific subject in the field of satellites and / or satellite imagery, and to generate and provide answers that match the user's query intent through the retrieved content.

[0077] In other words, in the first mode, the user may be provided with a response to a user query and at least one document used (or identified) to generate the response.

[0078] On the other hand, as shown in Figures 2a and 2b, the control unit 150 can receive a user query 200 (e.g., "Raw Query") corresponding to user input (e.g., "User Input"). Here, the reception of the user query 200 may be received via a service page 1000 linked with the response service provision system 100.

[0079] Referring to Figure 3, the service page 1000 may include at least one of a first area 310 that receives user queries and a second area 320 that provides documents retrieved in response to the user queries.

[0080] Specifically, the first area 310 may include at least one of the following: a first sub-area 311 that receives user queries, and a second sub-area where answers to user queries are provided (or displayed). In this case, the first sub-area 311 is the area where user input is performed, and is also called the input area.

[0081] The control unit 150 can receive user queries entered from the user terminal 10 via the service page 1000. In this case, the first sub-region 311 of the service page 1000 may include a graphic object 311a that is linked to the user query receiving function. For example, the control unit 150 can receive a user query 200 corresponding to user input (e.g., "Please tell me about the trends in reusable rockets.") from the user terminal 10 based on the selection of the graphic object 311a.

[0082] In one embodiment, the service page 1000 may further include a third area 330. The third area 330 may be configured to provide information related to conversational events that occur in the user account. The third area 330 may include graphic objects corresponding to conversational lists (or catalogs) generated by conversational events that occur in the user account. Based on the receipt of a user query 200, the control unit 150 can generate a chat list related to the user query 200 and provide graphic objects 331 corresponding to the chat list to the third area 330 of the service page 1000. The control unit 150 can also display (or output) the received user query 200 in a second sub-area 312 of the service page 1000.

[0083] On the other hand, the user query 200 can be rewritten using the large-scale language model 160 provided by the artificial intelligence server 20.

[0084] The control unit 150 can use a user query 200 to generate a prompt 201 (for example, "Rewrite Prompt Template") to be input to the large-scale language model 160. The elements (or information) constituting the prompt 201 can be understood as existing stored in the storage unit 140 (or memory).

[0085] In one embodiment, the present invention may further include a prompt generation unit (or model) that generates prompts; however, for the sake of explanation, this specification will describe the prompts as being generated by the control unit 150 itself.

[0086] The prompt 201 generated by the control unit 150 may include a user query 200 entered at the current time. In this case, if a dialogue (i.e., a previous dialogue) exists before the time the user query 200 was entered, the control unit 150 may generate the prompt 201 using at least one of the previous dialogue and the user query 200, taking into account the relationship with the previous dialogue, so that the user query 200 is recreated from the large language model 160 as an independent input value containing the relevant content. For the sake of explanation, the following description will assume that a prompt 201 including at least one of the previous dialogue and the user query 200 has been generated.

[0087] The control unit 150 can process the generated prompt 201 as input to the large-scale language model 160. Based on the input prompt 201, the large-scale language model 160 can generate a rewritten user query 202 by rewriting the user query 200, so that it can confirm (or understand) what intent the user query 200 contains. In this case, the large-scale language model 160 can be understood as an artificial intelligence model provided by the artificial intelligence server 20 (e.g., an OpenAI server).

[0088] On the other hand, the control unit 150 can process the user query 202 recreated by the large-scale language model 160 as input to the embedding model 210 (e.g., "Embedding Model"). The embedding model 210 can vectorize the recreated user query 202 and generate an embedded query vector 211 (e.g., "Embedded Query Vector" or embedding vector or query vector).

[0089] The control unit 150 can input the query 211, vectorized by the embedded model 210, into the vector DB 220 (e.g., "VectorDB" or document DB). Based on the embedded query vector 211 and the similarity (e.g., cosine similarity) between documents in the vector DB 220, the control unit 150 can perform a search for at least one document.

[0090] Furthermore, the control unit 150 can process the documents 221 retrieved from the vector DB 220 as input to the document sorting model 230 (or document resorting model). The document sorting model 230 can sort (or resort) the retrieved documents 221. For example, for resorting the retrieved documents 221, the document sorting model 230 can calculate the similarity between the retrieved documents 221 and the user's query using a sigmoid function and represent it with a score between 0 and 1. In this case, the document sorting model 230 can filter only the documents whose scores meet a predetermined criterion (e.g., 0.7 or higher) and extract the sorted documents 231 (e.g., "Reranked Documents" or resorted documents).

[0091] The control unit 150 can generate a prompt 240 (e.g., "RAG Prompt Template") to be input to the large language model 160 using at least one of the recreated user query 202 and the sorted document 231. For example, the prompt 240 may include at least one of the recreated user query 202, the sorted document 231, and the user's previous dialogue (or chat). The elements (or information) constituting the prompt 240 can be understood as existing stored in the storage unit 140 (or memory).

[0092] Furthermore, the control unit 150 can process the generated prompt 240 as input to the large-scale language model 160. Based on the input prompt 240, the large-scale language model 160 can generate an answer 250 (e.g., "Output Answer") to the user query 200. For example, the answer 250 to the user query 200 may be a text-formatted answer generated in the large-scale language model 160 using the RAG (Retrieval-Augmented Generation) method based on the sorted document 231.

[0093] The response 250 to the user query 200 and at least one document 251 (e.g., "Output Documents") used in the generation process of the response 250 may be transmitted to the backend 260 (e.g., "backend" or backend system) and the frontend 270 (e.g., "Front-end"). The backend 260 manages the user's conversation, the response 250 and the document 251, etc., and can store information related to the user's conversation history in the DB 261 (e.g., "Conversation MongoDB"). The frontend 270 can provide (or display) the response 250 and the document 251 to the user terminal 10.

[0094] On the other hand, the control unit 150 can provide a response 250 to the user query 200 via the service page 1000.

[0095] The control unit 150 can provide a response 250 to a user query 200 in a region of the service page 1000 output to the user terminal 10. For example, as shown in Figure 3, the control unit 150 can provide a response 250 to a user query 200 (for example, "Reusable rocket technology is developing mainly in the United States. SpaceX and Blue Origin are developing large reusable rockets, BFR and New Glenn, and the United States is conducting research on small reusable rockets through the XS-1 project. The development of reusable rockets is bringing new vitality to the space industry.") in a second sub-region 312 included in the first region 310 of the service page 1000.

[0096] In this case, the answer 250 may include at least one of the following graphic objects: 250a, which is linked to a function that allows copying the text contained in the answer 250; 250b, which is linked to a function that allows selecting whether the answer 250 is liked (or satisfactory); and 250c, which is linked to a function that allows selecting whether the answer 250 is disliked.

[0097] Furthermore, the control unit 150 may provide, along with the response 250, information about at least one document 251 identified to generate the response 250 (e.g., a PDF link to the document, the location of a document chunk, the source of the document, etc.). The control unit 150 may provide to the second area 320 of the service page 1000 information about a plurality of documents 321, 322, 323 identified to generate the response 250 to the user query 200. Here, the documents 321, 322, 323 provided to the service page 1000 may correspond to documents contained in at least one of the retrieved document 221 and / or sorted document 231.

[0098] Thus, in the present invention, through the first mode, it is possible to provide the latest information related to space, satellites, and aviation (e.g., news) that meets the user's requests, and to provide the user with answers that improve the hallucination phenomenon of the artificial intelligence model by utilizing the retrieved documents.

[0099] Next, we will examine the second mode among the multiple modes.

[0100] The second mode (or map mode) may also be called "Map Mode," and may be configured to retrieve at least one satellite image related to the user query from the memory unit 140 and / or at least one storage and / or satellite image DB 30 when a user query related to a satellite and / or satellite image is received, and to generate and provide an answer to the user query using the analysis results from the analysis of the retrieved satellite image and the large-scale language model 160.

[0101] Unlike the first mode, which provides text-based answers, the second mode can provide answers to user queries in a geographical format or visually display responses to user queries on a map. For example, the second mode can provide satellite imagery of a specific location (e.g., a location to respond to a user request) or visually represent (or display) information about a specific location.

[0102] In other words, in the second mode, the user may be provided with a response to a user query, along with at least one of the satellite imagery, location information, or maps used (or identified) to generate the response.

[0103] In this regard, referring to Figures 4A, 4B, 4C, 4D, and 4E, in the present invention, the second mode can operate based on the Re-Act (Reasoning Action) + Agent method.

[0104] In one embodiment, the Re-Act+Agent method may include: i) a "Thought process" of analyzing (or understanding) the query (or request) entered by the user and deciding what model (or tool) to use to generate (or resolve) the answer to the user query; ii) an "Action process" of performing the necessary tasks to generate the answer to the user query using the specific model determined through the thought process; and iii) an "Observation process" of observing (or confirming) the results of the tasks performed using the specific model. In this case, the thought process may also be understood as a "reasoning (or inference or thinking) process."

[0105] The Re-Act+Agent method proceeds in the order of "think → act → observe," and the cyclical process may be repeated in the order of "think → act → observe → think again" as needed. For example, if the control unit 150 determines that a particular model decided in the thinking process is not suitable for generating a response to a user query, or if additional information is needed, it may return to the thinking process and repeat the cyclical process of determining a new particular model.

[0106] On the other hand, in the second mode, since the response to the user query needs to be displayed on a map (or map), the information (or query) requested by the user can be represented as geometry information (or location information or coordinate information).

[0107] In one embodiment, when the control unit 150 receives a user query requesting satellite imagery of a specific region, it can generate geometry information for the specific region, or extract (or identify) geometry information for the specific region using a linked server and / or at least one model.

[0108] Geometry information is transmitted (or transmitted) in a specific format to coordinates on a map, and if satellite imagery is provided first, geometry information related to the satellite imagery may be included and input (or transmitted) to the response service provision system 100 together with user input (or user query).

[0109] Referring again to Figures 4A, 4B, 4C, 4D, and 4E, the user query 400 (e.g., "User Input"), satellite image (e.g., "Image name"), and geometry information (e.g., "Geometry") may be used to generate prompts to be input to the large-scale language model 160. More specifically, the control unit 150 can use at least one of the user query, satellite image, and geometry information to generate a prompt 410 (e.g., "Re-Act Prompt Template") to be input to the large-scale language model 160.

[0110] Prompt 410 may include various pieces of information (or elements) necessary to generate a response to the user query 400. For example, prompt 410 may include i) a list of available models 411 (or tool list) and a description of the role each model plays, ii) the date prompt 410 is executed, iii) satellite imagery, the name of the satellite imagery, geometry information (or geographic location information), iv) the user's previous interactions (e.g., queries previously entered by the user, information mentioned, requests, etc.), and v) at least one of the queries 400 currently entered by the user. However, the information included in prompt 410 is not necessarily limited to these, and may include various other pieces of information beyond the examples given above.

[0111] The control unit 150 can process the generated prompt 410 as input to the large-scale language model 160. Based on the input prompt 410, the large-scale language model 160 can determine how to solve the user query 400 and output (or generate) a format 420 (format, or output data) containing the elements (or information) necessary to execute this.

[0112] The format 420 output from the large language model 160 may include at least one of the following: i) a first element (e.g., "Thought") representing the reasoning related to the thoughts and / or decisions to be considered in order to resolve the user query 400, based on the user query 400 and the list of available models 411; ii) a second element (e.g., "Action Name") representing the name of the model identified (or selected) from the large language model 160 to resolve the user query 400; and iii) a third element (e.g., "Action Input") representing the input values ​​(or input data) required for the identified model to perform its work.

[0113] In one embodiment, let's assume that user query 400 contains the content "I want you to detect vehicles from worldview-2 video footage of Seoul Station taken in 2023." In this case, the first element may also contain the content "In order to satisfy the user's request, we must search for worldview-2 video footage of Seoul Station taken in 2023 and perform the task of detecting vehicles from the worldview-2 video footage using a model capable of vehicle detection."

[0114] In another embodiment, suppose user query 400 contains the following: "I want you to detect vehicles from worldview-2 video footage of Seoul Station taken in 2023." In this case, the second element may include the name of a model identified to perform a search process on satellite imagery corresponding to user query 400 (e.g., "worldview-2 video") (e.g., "SearchSatelliteImageTool", or satellite imagery search model), and the name of a model identified to detect vehicles from the searched (or identified) satellite imagery (e.g., "VehicleDetectionTool", or object detection model).

[0115] In another embodiment, suppose user query 400 contains the following: "I want you to detect a vehicle from worldview-2 video footage of Seoul Station taken in 2023." In this case, the third element may include input data necessary for the identified model (e.g., first model 170) to perform a specific task (e.g., vehicle detection) (e.g., object detection target information (e.g., vehicle), satellite imagery of a specific area (e.g., worldview-2 video)).

[0116] In this regard, if the name of a specific model included in the second element does not exist in the available model list 411, the control unit 150 can generate a prompt 410 again and input it to the large language model 160, causing the large language model 160 to generate the format 420 again. On the other hand, if the name of a specific model included in the second element exists in the available model list 411, the control unit 150 can input the input data included in the third element to the specific model and cause it to execute a specific function for generating a response to the user query 400.

[0117] Thus, the present invention uses a large-scale language model 160 to identify which model to use to resolve user input and to determine which input values ​​the identified model must use to perform its tasks. In other words, the present invention allows for the selection of the optimal model capable of resolving user queries, thereby providing appropriate answers that meet user requirements by performing only the necessary data processing.

[0118] On the other hand, as described above, in order to generate a response to user query 400, the choice of which specific model from the models included in the model list 411 to use for a particular function may be determined based on the model name (or second element) identified from the large-scale language model 160.

[0119] In this case, the required input data (e.g., Action Input) may differ for each model included in the model list 411. That is, the input data processed for each model may be different from one another.

[0120] In one embodiment, a satellite image search model (e.g., "SearchSatelliteImageTool") can search for satellite images in a satellite image DB30 and / or at least one storage (e.g., a PostgreSQL database) where the satellite images are stored. In this case, the satellite image search model may require input data processing that includes at least one of the following: the name of the satellite image, the date it was taken, and geographical information.

[0121] In other embodiments, an object detection model (e.g., "VehicleDetectionTool") can detect at least one object to be detected in satellite imagery. In this case, the object detection model may require input data processing that includes at least one of the target information to be detected and satellite imagery retrieved from a satellite imagery retrieval model.

[0122] In another embodiment, a satellite image rendering model (e.g., "RenderSatelliteImageTool") can render satellite imagery using link information (e.g., Tile JSON URL) provided from a pre-identified server (e.g., Tile Server) or satellite imagery DB30. In this case, the size of the image displayed at once is determined based on the zoom level (or degree) of the map (or map image), and as zooming out reveals a wider area, this can be divided and stored in multiple tiles. Because the servers are configured separately, the information called from the servers may differ depending on the imagery and map images used to perform specific functions of a particular model.

[0123] On the other hand, the term used for the aforementioned model list 411 may be referred to as either "function 411" or "function list 411," or at least one of these. In this case, each function included in the function list 411 may have a pre-defined algorithm that performs a different function (or a specific function) from the others. That is, the term "function" can be understood as a function that processes input data (or input values) according to a pre-defined algorithm and outputs output data (or changed values).

[0124] In one embodiment, when a specific function (e.g., "SearchCoordinateNaverTool") is determined from the large-scale language model 160, the control unit 150 can process input data (e.g., geometry information) according to an algorithm pre-set for the specific function and output output data (e.g., location information).

[0125] On the other hand, the observation process allows us to observe (or monitor) the results after a specific model performing a identified function has been executed. The observed process may differ for each model, and since the observation process is related to user input, it can be used as evidence to determine whether a response that matches the user input has been generated.

[0126] For example, the control unit 150 can observe that after the object detection model is executed, the object detection model outputs a result value such as "A total of XX vehicles were detected from the video footage at that location."

[0127] In this case, the results observed through the observation process can be converted into content through post-processing. Such content can then be used (or reflected) to generate answers to user query 400.

[0128] The control unit 150 can generate a final result (e.g., "finish answer") if, through the observation process, it is confirmed that the large-scale language model 160 has appropriately generated a name for a specific model and, through this, has used a model that is suitable for the user query. The final result is transmitted to the front-end, which can generate an answer to the user query in the form of a description of the results of the analysis using the specific model.

[0129] On the other hand, if the control unit 150 fails to generate an appropriate result for user input, it can execute the Re-Act+Agent process again. For example, the control unit 150 can use the large-scale language model 160 to identify a suitable model for resolving the user query, and it can repeatedly perform the process of generating an appropriate result for user input based on that specific model.

[0130] The control unit 150 may, based on the output data (e.g., analysis results) output by the identified model, generate an appropriate final answer that matches the user's query, and then transmit the generated final answer to the backend 430 (e.g., "backend").

[0131] In this regard, referring to Figure 4E, we can see the relationship between the artificial intelligence server 20, the backend 430, the frontend 440, and the pre-specified server 450 (hereinafter referred to as the tile server).

[0132] The artificial intelligence server 20, which provides a large-scale language model 160, can process user queries 400, select a suitable model in response to the user query, and generate an appropriate response to the user query 400. The control unit 150 can transmit various status information generated during this process to the backend 430 in real time using the websocket method. For example, websockets can be understood as a technology that enables real-time bidirectional communication between a client and a server.

[0133] Next, the control unit 150 can transmit the status information transmitted to the backend 430 to the frontend 440. The frontend 440 can also be understood as a user interface (or user interface module) and can visually display information or results requested by the user. For example, the frontend 440 can display satellite imagery on a map image or display searched information.

[0134] Here, the front-end 440 can communicate directly with the tile server 450. The tile server 450 can provide information related to the tile images to be displayed on the map. For example, when a user zooms in, zooms out, or moves the map, the tile server 450 can provide the front-end 440 with a tile image that fits that location.

[0135] Furthermore, at least one of the user query 400 and / or the response to the user query and / or the user's dialogue and instructions may be linked to the user account and stored in the memory unit 140 and / or at least one storage (e.g., MongoDB 460). This can be used to understand the context of future interactions or to efficiently store and retrieve dialogue content.

[0136] The following will explain in more detail the method for providing the response service according to the present invention, assuming that the user has selected the second mode.

[0137] First, the present invention allows for the process of receiving user queries via a service page (see S510, Figure 5).

[0138] As shown in Figure 6a, the control unit 150 can provide the user terminal 10 with a service page 1000 linked to the response service provision system 100.

[0139] In the first area 610 of the service page 1000 when the second mode is activated, text-based information related to the user query (e.g., the user query, the answer to the user query, etc.) may be provided. In addition, in the second area 620, which is different from the first area 610, image-based information related to the user query (e.g., a map, satellite imagery, etc.) may be provided.

[0140] In this regard, the service page 1000 may include at least one of a first area 610 that receives user queries and a second area 620 on which a map image 621 is displayed.

[0141] Specifically, the first area 610 may include at least one of a first sub-area 611 for receiving user queries and a second sub-area 612 for providing (or displaying) answers to the user queries. In this case, the first sub-area 611 is the area where user input is performed and is also called the input area.

[0142] The control unit 150 can receive user queries entered from the user terminal 10 via the service page 1000. In this case, the first sub-area 611 of the service page 1000 may include a graphic object 611a that is linked to the user query receiving function. For example, based on the selection of the graphic object 611a from the user terminal 10, the control unit 150 can receive a user query 612a corresponding to user input (for example, "I would like you to detect trains at Itabashi Station from worldview-2 video footage of Itabashi Station taken in 2021.").

[0143] On the other hand, the control unit 150 can process the received user query 612a as input to the large-scale language model 160.

[0144] Specifically, the control unit 150 can generate prompts (e.g., Re-Act Prompt, or First Prompt) to be input to the large-scale language model 160 using user queries 612a (see Figures 4A to 4D).

[0145] The following will provide a more detailed explanation with reference to Figures 4A to 4d.

[0146] As described above, the prompts input to the large-scale language model 160 may include various elements necessary to generate an answer corresponding to the user query 612a. For example, the prompt may include i) a list of available models and a description of the role each model plays, ii) the date on which the prompt is executed, iii) satellite imagery, the name of the satellite imagery, geometry information (or geographic location information), iv) the user's previous interactions (e.g., queries previously entered by the user, information mentioned, requests, etc.), and v) at least one of the queries 612a currently entered by the user.

[0147] The control unit 150 processes the generated prompt as input to the large-scale language model 160 and can select a model from the large-scale language model 160 to perform the analysis corresponding to the user query.

[0148] Specifically, the large-scale language model 160 can determine how to solve the problem for user query 612a based on the input prompt and output output data containing various elements necessary to do so.

[0149] The output data from the large-scale language model 160 may include at least one of the following: i) a first element (e.g., "Thought") representing the reasoning related to the thinking and / or judgments to be considered in order to resolve the user query 612a, based on the user query 612a and the list of available models; ii) a second element (e.g., "Action Name") representing the name of the model identified from the large-scale language model 160 in order to resolve the user query 612a; and iii) a third element (e.g., "Action Input") representing the input data required for the identified model to perform a particular function.

[0150] In one embodiment, let's assume that user query 612a contains the content "I want you to detect trains at Itabashi Station from worldview-2 video footage of Itabashi Station taken in 2021." In this case, the first element may also contain the content "In order to satisfy the user's request, we must search for worldview-2 video footage of Itabashi Station taken in 2021 and perform the task of detecting trains from the worldview-2 video footage using a model capable of detecting trains."

[0151] Next, the second element output along with the first element may include the name of a model identified to perform a search process for satellite imagery corresponding to user query 612a (e.g., "worldview-2 imagery") (e.g., "SearchSatelliteImageTool", or satellite imagery search model), and the name of a model identified to detect vehicles from the searched (or identified) satellite imagery (e.g., "VehicleDetectionTool", or object detection model). For the sake of explanation, the model identified to search satellite imagery will be referred to as the "satellite imagery search model," and the model identified to detect vehicles will be referred to as the "first model 170."

[0152] Furthermore, the third element output along with the first and second elements may include input data necessary for the identified satellite imagery model to perform a specific function (satellite imagery search) (e.g., name of the satellite imagery, date of capture, satellite that took the image, location information, geometry information, etc.), and input data necessary for the first model 170 to perform a specific function (object detection) (e.g., target information for object detection (e.g., vehicle), satellite imagery retrieved from the satellite imagery search model (e.g., worldview-2 imagery)).

[0153] On the other hand, the present invention allows for the process of identifying location information from user queries (see S520, Figure 5).

[0154] As described above, the large-scale language model 160 can understand the content contained in the user query 612a based on the input prompt, and based on the result of the understanding, it can output the user intent analysis result of the user query 612a (for example, "In order to resolve the user's request, we must search for worldview-2 video footage taken at Itabashi Station in 2021 and perform the task of detecting vehicles from the worldview-2 video footage using a vehicle detection model," or the first element).

[0155] The control unit 150 can use the output data of the large-scale language model 160 to identify location information (or coordinate information or coordinate values) from the user query 612a.

[0156] Specifically, the control unit 150 can extract (or identify) information necessary for identifying location information based on the user intent analysis results. Here, examples of information necessary for identifying location information may include at least one of i) place name, ii) location name, and iii) address. However, in the present invention, the information necessary for identifying location information is not necessarily limited to the above examples, and may include any information that can be used to identify location information.

[0157] For example, the control unit 150 can extract "Itabashi Station," which corresponds to the place name, from the information necessary to identify the location, based on the user intent analysis results (for example, "In order to resolve the user's request, we must search for worldview-2 video footage taken at Itabashi Station in 2021 and perform the task of detecting vehicles from the worldview-2 video footage using a model capable of vehicle detection," or the first element).

[0158] Furthermore, the control unit 150 can identify the location information from the extracted information, based on the fact that the information necessary for identifying the location information has been extracted.

[0159] In this case, if information necessary for identifying location information is extracted, the control unit 150 may be configured to enable the identification of location information itself. Furthermore, the control unit 150 can identify (or extract) location information (e.g., latitude and longitude) using an external server linked to the response service provision system 100 and / or a model provided from the external server (e.g., "SearchCoordinateNaverTool" or a location information analysis model).

[0160] In one embodiment, the control unit 150 can identify location information (e.g., latitude: 37.3947, longitude: 127.1104) for a place name (e.g., "Itabashi Station") extracted from a user query 612a via a map API provided by an external server linked to the response service provision system 100, or via link information (URL) of a website linked to the map.

[0161] In another embodiment, the control unit 150 can use a location information analysis model to identify location information (e.g., latitude: 37.3947, longitude: 127.1104) for a place name (e.g., "Itabashi Station") extracted from the user query 612a.

[0162] However, although the process described above involves analyzing user intent using a large-scale language model 160, in the present invention, the analysis of user intent may also be performed by the control unit 150 itself.

[0163] On the other hand, the control unit 150 can generate geometry information using the identified location information. Here, "generating geometry information" can also be understood as converting (or representing) the identified location information into geometry information.

[0164] Geometry can refer to a method of representing spatial data in a Geographic Information System (GIS). Geometric information is used to define spatial characteristics such as location, shape, size, and direction in maps and geographical data, and can be represented in various forms such as points, lines, and polygons.

[0165] In this case, the conversion of geometry information is performed either i) by the control unit 150 itself, or ii) using a model (or module) configured to convert position information (or coordinate information or coordinate values) into geometry information. For the sake of clarity, the above cases will not be distinguished in the following explanation.

[0166] The control unit 150 can use a geometry information generation (or conversion) model to convert location information (e.g., latitude: 37.3947, longitude: 127.1104) for place names extracted from the user query 612a into geometry information.

[0167] Furthermore, the control unit 150 can process the location information for the place name as input to a geometry information generation model. In one embodiment, the geometry information generation model can convert the input location information into geometry information having at least one form of a point and / or polygon.

[0168] On the other hand, the present invention allows for the process of searching for satellite imagery related to a user query from at least one database using user queries and location information (see S530, Figure 5).

[0169] As described above, the model used to search for satellite imagery related to user query 612a may be determined by the large-scale language model 160. For example, the control unit 150 may use a satellite imagery search model to perform the process of searching for satellite imagery based on a second element output from the large-scale language model 160.

[0170] In other words, the satellite image search model may be a model selected from the large-scale language model 160 to perform the search process for satellite imagery corresponding to user query 612a (for example, "worldview-2 imagery").

[0171] The control unit 150 can identify the input data necessary for the satellite image search model to perform the satellite image search process and process it as input to the satellite image search model. In this case, the input data to be input to the satellite image search model may be identified based on a third element output from the large-scale language model 160. For example, the control unit 150 can input at least one of the following as input data to the satellite image model: the date of capture (e.g., "2021"), the name of the satellite image (e.g., "worldview-2 image"), identified location information, and transformed geometry information.

[0172] The satellite image search model can retrieve satellite imagery related to user query 612a from at least one storage and / or DB based on the input data. For example, the satellite image search model can retrieve satellite imagery from the satellite imagery DB30 related to the date the satellite imagery was taken (e.g., "2021"), the name of the satellite imagery (e.g., "worldview-2 imagery"), the identified location information, and the transformed geometry information.

[0173] On the other hand, in an embodiment different from the one described above, the control unit 150 can use the identified location information to search for satellite imagery related to the user query 612a.

[0174] Specifically, the control unit 150 can use the converted geometry information and at least one storage and / or DB to search for satellite imagery related to the user query 612a. In this case, the DB used for searching satellite imagery stores various satellite imagery, and each satellite imagery may include spatial information (i.e., geometry information (e.g., location and extent)) indicating which region it includes.

[0175] In this invention, there may be one or more databases used for searching satellite imagery, depending on the need (or circumstances). However, for the sake of convenience in the following explanation, the databases used for searching satellite imagery will not be separately classified and will be referred to as "satellite imagery DB30".

[0176] Specifically, the control unit 150 can search for satellite images related to a user query by comparing the converted geometry information with the geometry information of satellite images stored in the satellite image DB 30, using the converted geometry information as a reference.

[0177] In one embodiment, the control unit 150 can use PostGIS (for example, the spatial query function or PostGIS functions of PostGIS) capable of storing, managing, and analyzing spatial data to search for at least one satellite image from among multiple satellite images stored in the satellite image DB 30 that corresponds to the converted geometry information.

[0178] In another embodiment, the control unit 150 can use PostGIS to search for at least one satellite image from among multiple satellite images stored in the satellite image DB 30 that contains location information (or coordinate information or coordinate values) of a place name (for example, "Itabashi Station") included in the user query 612a.

[0179] Thus, in the present invention, the search for satellite imagery related to a user query can be performed either by the satellite imagery search model or by the control unit 150 itself, and the system can be configured to perform either.

[0180] On the other hand, if multiple satellite images related to user query 612a are retrieved from the satellite image database, the control unit 150 can identify one of the multiple satellite images based on various criteria pre-set in the response service provision system 100. For example, the various criteria may include at least one of the following: i) the most recent date (or time), ii) the user's request (or selection), iii) a weight that includes the identified location information, and iv) a weight that includes the user's area of ​​interest (or designated). However, in the present invention, the criteria for identifying a satellite image are not necessarily limited to the examples described above, and may include various other criteria besides those described above.

[0181] In one embodiment, if the preset criterion is "most recent date," the control unit 150 can analyze the metadata matched to each of the searched satellite images and identify the most recently captured satellite image as the satellite image associated with the user query 612a.

[0182] Furthermore, the present invention allows for the use of satellite imagery to perform analysis related to user queries (see S540, Figure 5).

[0183] The control unit 150 can perform analysis corresponding to user queries using the retrieved (or identified) satellite imagery.

[0184] The control unit 150 can input satellite imagery retrieved in relation to user query 612a as input to a model identified for performing analysis corresponding to user query 612a.

[0185] As described above, the large-scale language model 160 can determine which model to use to perform the analysis corresponding to user query 612a. More specifically, the control unit 150 may use the first model 170 to execute the object discovery process corresponding to user query 612a based on the second element output from the large-scale language model 160.

[0186] In other words, the first model 170 may be a model selected from the large language model 160 to perform the process of detecting objects to be detected (e.g., vehicles) from satellite imagery retrieved in response to the user query 612a (e.g., "worldview-2 imagery").

[0187] The control unit 150 can identify the input data necessary for the first model 170 to perform the object detection process corresponding to the user query 612a and process it as input to the first model 170. In this case, the input data to be input to the first model 170 can be identified based on the third element output from the large language model 160. For example, the control unit 150 can input at least one of the retrieved satellite imagery (e.g., "worldview-2 imagery") or the object information to be detected (e.g., "vehicle") as input data to the first model 170.

[0188] The first model 170 can detect objects to be detected in retrieved satellite imagery based on input data. For example, the first model 170 can detect vehicles from input satellite imagery and output analysis results that include information about the number of vehicles detected (e.g., 614 vehicles) (e.g., "A total of 614 vehicles were detected.").

[0189] In another embodiment, the first model 170 can perform object detection work if, as appropriate, satellite imagery is rendered on the map image 621 of the service page 1000. In this case, the control unit 150 can render the satellite imagery retrieved in response to the user query 612a onto the map image 621 and control the first model 170 to detect target objects in the satellite imagery.

[0190] Furthermore, the control unit 150 can receive the analysis results of satellite imagery corresponding to user query 612a from the selected model. For example, the control unit 150 can receive the analysis results of satellite imagery corresponding to user query 612a from the first model 170 (for example, "A total of 614 vehicles were detected.").

[0191] On the other hand, the present invention can generate a prompt using at least one of the following: user queries, location information, analysis results, and satellite images (see S550, Figure 5).

[0192] The control unit 150 can generate a prompt to be input to the large-scale language model 160 using at least one of the following: the user query 612a, the identified location information, the analysis results received from the first model 170, and the retrieved satellite imagery. For example, the prompt may include at least one of the following: the user query 612a, the identified location information (e.g., latitude: 37.3947, longitude: 127.1104), the output data from the first model 170 (e.g., "A total of 614 vehicles were detected."), and the satellite imagery used by the first model 170 for analysis (or retrieved satellite imagery).

[0193] Furthermore, the present invention allows for the processing of prompts as input to a large-scale language model (see S560, Figure 5).

[0194] The control unit 150 can process the prompt as input to the large-scale language model 160. Based on the input prompt, the large-scale language model 160 can use the analysis results of the first model 170 and the retrieved satellite imagery to generate an answer 612b to the user query 612a (for example, "Rendered red, green, and blue for worldview-2_20211106_wydku8vn. Worldview-2 satellite imagery taken on November 6, 2021 was rendered in red, green, and blue. A total of 614 vehicles were detected from the imagery at that location.") (see Figure 6a).

[0195] Furthermore, the present invention enables the process of obtaining answers to user queries from a large-scale language model (see S570, Figure 5).

[0196] The control unit 150 can receive a response 612b to a user query 612a from the large language model 160 (for example, "Rendered red, green, and blue for worldview-2_20211106_wydku8vn. Rendered red, green, and blue from worldview-2 satellite imagery taken on November 6, 2021. Also detected a total of 614 vehicles from the imagery at that location.").

[0197] On the other hand, the present invention allows for the process of providing answers to user queries on a service page (see S580, Figure 5).

[0198] The control unit 150 can provide the service page 1000 with the answer 612b to the user query 612a, and at least one satellite image 622 retrieved in response to the user query 612a.

[0199] The first area 610 of service page 1000 may provide text-based information related to user query 612a, and the second area 620 may provide image-based information related to user query 612a.

[0200] As a result, the control unit 150 can provide the first area 610 of the service page 1000 with a response 612b generated from the large language model 160 (for example, "Rendered red, green, and blue for worldview-2_20211106_wydku8vn. Rendered red, green, and blue from worldview-2 satellite imagery taken on November 6, 2021. Also detected a total of 614 vehicles from the imagery at that location.").

[0201] Furthermore, the control unit 150 can provide the satellite imagery 622 retrieved in response to a user query to the second area 620 of the service page 1000. More specifically, based on the fact that satellite imagery has been retrieved in response to a user query 612a, the control unit 150 can provide (or display) the satellite imagery 622 by overlaying it on at least a portion of the map image 621 displayed in the second area 620.

[0202] In this case, the area where the satellite image 622 overlaps with the map image 621 may be identified based on location information identified for the search of the satellite image.

[0203] In one embodiment, the control unit 150 can identify a region corresponding to the identified location information from among a plurality of regions included in the map image 621, based on the fact that the identified location information corresponds to "latitude: 37.3947, longitude: 127.1104". The control unit 150 can overlay (or render) the satellite image 622 onto the identified region and display it.

[0204] In another embodiment, the control unit 150 may overlay satellite imagery 622 onto the map image 621 based on information about a specific place name (e.g., "Itabashi Station") extracted from the user query 612a. The control unit 150 may also overlay and display satellite imagery 622 on a specific area corresponding to a specific place name from among a plurality of areas included in the map image 621.

[0205] Furthermore, the satellite imagery 622 may also include the analysis results of the satellite imagery 622 of the first model 170. For example, the satellite imagery 622 may also display information (e.g., vehicles) about objects detected by the first model 170 through object detection analysis results.

[0206] On the other hand, the control unit 150 can receive user input specifying a particular area in the satellite imagery provided to the second region 620.

[0207] In the present invention, user input to satellite imagery is not necessarily limited to specifying a particular region. In addition to the user specifying a particular region in the satellite imagery, the present invention can provide a user environment that enables input related to the satellite imagery provided to the second region 620 through input to the entire satellite imagery or input of text queries to the satellite imagery. However, for the sake of explanation, this specification will be described on the premise that user input for a particular region has been received.

[0208] The service page 1000 may be provided with at least one tool icon (or graphic object) for receiving user input for satellite imagery provided in the second area 620. For example, as shown in Figure 6b, the control unit 150 may provide at least one of the first tool icon 622a and the second tool icon 622b in the second area 620 of the service page 1000.

[0209] In the present invention, the first tool icon 622a and the second tool icon 622b may be linked to different functions, respectively.

[0210] The first tool icon 622a may be linked to a first selection function that selects at least a portion of the satellite imagery 631 according to a first criterion. For example, the first selection function of the first tool icon 622a may be configured to allow the user to select a square area centered on a region (or point) selected in the satellite imagery 631, and to allow rotation of the selected area.

[0211] The second tool icon 622b may be linked to a second selection function that selects at least a portion of the satellite imagery 631 according to a second criterion different from the first criterion. For example, the second selection function of the second tool icon 622b may allow the user to specify a square rectangular area centered on (or based on) a region (or point) selected in the satellite imagery 631, and may be configured so that rotation of the specified area is not possible.

[0212] The control unit 150 can receive user input specifying a particular area in the satellite image 622 using one of the multiple tool icons 622a and 622b. For example, as shown in Figure 6b, suppose the second tool icon 622b is selected from the user terminal 10. The control unit 150 can receive user input from the user terminal 10, where the second tool icon 622b was activated, specifying a particular area 632 from among the multiple areas included in the satellite image 631.

[0213] On the other hand, the control unit 150 can receive additional user queries related to a specific region 632.

[0214] Specifically, the control unit 150 can receive additional user queries 632a related to a specific area 632 in the first area 610 of the service page 1000 (for example, "Please calculate the NDVI index for this area").

[0215] If an additional user query 632a is received, a response to that additional user query 632a can be provided based on the Re-Act+Agent method described above. The specific details related to this have already been explained, so a brief explanation follows.

[0216] The control unit 150 can generate a prompt (e.g., Re-Act Prompt) to be input to the large-scale language model 160 using an additional user query 632a. Here, the control unit 150 can generate a prompt using at least one of the user query 612a corresponding to the previous dialogue and the additional user query 632a, such that the additional user query 632a is analyzed by the large-scale language model 160 as an independent input value containing the relevant content, taking into account the relevance to the previous dialogue.

[0217] Furthermore, the control unit 150 can process the generated prompts as input to the large-scale language model 160. Based on the input prompts, the large-scale language model 160 can analyze the user intent of additional user queries 632a, identify functions corresponding to the user intent, and select a specific analysis model to perform the identified functions.

[0218] The large-scale language model 160 determines, based on the fact that the additional user query 632a includes the statement "I want the NDVI index for the region to be calculated," that the user intent relates to the calculation of the NDVI index for a specific region 632. Based on this determination, it can identify (or determine) a satellite image search model for searching satellite imagery for the specific region 632, and a second model 180 that performs a specific function corresponding to the additional user query 632a (e.g., calculating the NDVI index).

[0219] The control unit 150 can use a satellite image search model determined from the large-scale language model 160 to search for satellite imagery for a specific area 632 from the satellite imagery DB 30, and identify satellite imagery from the searched images that meet pre-set criteria.

[0220] Furthermore, the control unit 150 can process the identified satellite imagery as input to a second model 180 corresponding to a specific analysis model. For example, the second model 180 can analyze satellite imagery of a specific region 632 and output analysis results including the NDVI index for that region 632.

[0221] Furthermore, the control unit 150 can obtain output (or output data) from the second model 180 that corresponds to the user intent of the additional user query 632a.

[0222] On the other hand, the control unit 150 can use satellite imagery of a specific region 632, metadata associated with the specific region 632, and at least one of the additional user queries 632a to generate a response to an additional user query 632a (for example, "Rendered NDVI for worldview-2_20211106_wydku8vn. After detecting a total of 614 vehicles from the region in the WorldView-2 satellite imagery taken on November 6, 2021, the NDVI index was calculated.") (see Figure 6b).

[0223] Here, examples of metadata associated with a particular region 632 may include at least one of the following: location information, geometry information, climate data, land use information, and environmental data for the particular region 632. In this case, the metadata can be understood as being collected by the control unit 150 from the storage unit 140 and / or a pre-specified storage (or DB).

[0224] Furthermore, the control unit 150 can provide answers to additional user queries 632a to at least one of the first and second regions.

[0225] As shown in Figure 6b, the control unit 150 can provide a text response 632b to an additional user query 632a in the first region 610 and provide satellite imagery 631 for a specific region 632 in the second region 620.

[0226] In this case, the response to the additional user query 632a provided to the second region 620 may be displayed in the satellite image 631 of a specific region 632 provided to the second region 620 with at least a portion of it highlighted.

[0227] In other words, the control unit 150 may highlight at least a portion of the satellite image 631 of a specific region 632 and display it on the second region 620.

[0228] Furthermore, satellite imagery for a specific region 632 may be provided with the analysis results of the second model 180. For example, satellite imagery for a specific region 632 may also be provided with information on the NDVI index analyzed by the second model 180 through the NDVI analysis results.

[0229] On the other hand, in the second mode, various functions can be provided to the user in addition to the functions mentioned above.

[0230] In the second mode, the user may be provided with functions related to at least one of the following: searching for satellite imagery and coordinates, predicting (or querying) satellite orbits, querying satellite position information, calculating (or determining) spectral indices, object detection and segmentation results, and purchasing (or selling) satellite imagery. However, it goes without saying that the functions provided in the second mode are not necessarily limited to the examples given above, and may include a variety of other functions.

[0231] In this case, the response service provision system 100 according to the present invention may include at least one model for performing the various functions provided in the second mode. However, the process for providing the various functions can also be performed by the control unit 150 itself. For the sake of explanation, the following description will assume that the process for providing the various functions is performed by the control unit 150 itself, without separately distinguishing the entity providing the various functions.

[0232] In one embodiment, as shown in Figure 7, the control unit 150 can receive a user query 711a via the service page 1000 requesting location information for a specific satellite (for example, "Please tell me the current location of sentinel-2a."). The control unit 150 can identify the location information of the specific satellite related to the user query 711a and provide a response 711b containing the identified location information (for example, "The current location of sentinel-2a is latitude -8.062, longitude -75.039.") to the first region 710. The control unit 150 can also overlay and display an object 721a of the specific satellite on a map image 721 provided to the second region 720. In this case, the position of the object 721a of the specific satellite displayed on the map image 721 may be determined based on the location information of the specific satellite.

[0233] In another embodiment, the control unit 150 may receive a user query 712a via the service page 1000 requesting a prediction of the satellite orbit of a specific satellite (for example, "Please propagate the orbit of the sentinel-2a satellite until June 24, 2024."). The control unit 150 may predict the satellite orbit of the specific satellite related to the user query 712a and provide a response 712b containing information about the predicted orbit (for example, "The orbit of the sentinel-2a satellite has been propagated until June 24, 2024.") to the first region 710. The control unit 150 may also overlay and display the predicted orbit information of the specific satellite on the map image 721 provided to the second region 720.

[0234] Furthermore, in the second mode, a user environment can be provided that allows users to purchase satellite imagery that they wish to buy.

[0235] In one embodiment, as shown in Figure 8, the control unit 150 can receive a user query 811a requesting the purchase of specific satellite imagery via the service page 1000 (for example, "I want to purchase satellite imagery."). Based on the receipt of the user query 811a, the control unit 150 can provide a response 811b related to the user query 811a in the first area 810 (for example, "Do you want to purchase the WorldView-2 satellite imagery of Itabashi Station taken on November 6, 2021? If you purchase this satellite imagery, you will be provided with high-resolution satellite imagery taken by the first satellite...") and provide the specific satellite imagery 821 that the user requested to purchase, and a graphic object 822 linked to a purchase request receiving function (or payment function for purchase) for the specific satellite imagery 821 in the second area 820. The specific satellite imagery 821 may be satellite imagery searched in response to the user query, or it may be satellite imagery identified by information input regarding the satellite imagery that the user wishes to purchase. Based on the selection of the graphic object 822, the control unit 150 can sell a specific satellite image 821 to the user.

[0236] In this case, the satellite imagery rendered on service page 1000 may be low-resolution, while the satellite imagery sold to users may be high-resolution. That is, the satellite imagery sold to users can be understood as having higher resolution and / or image quality than the satellite imagery rendered on service page 1000.

[0237] In another embodiment, as shown in Figure 9, the control unit 150 may receive a user query 911a via the service page 1000 (e.g., "I want to purchase all the satellite images you have shown me so far."). Based on the receipt of the user query 911a, the control unit 150 sends a response 911b related to the user query 911a to the first area 910 (e.g., "Do you want to purchase the worldview-2 satellite image of Itabashi Station taken on November 6, 2021? If you purchase this satellite image, you will be provided with high-resolution satellite images taken by the first satellite…" The control unit 150 can provide the user with a set of satellite images 921 and 922 that the user has requested to purchase, and a graphic object 923 linked to a function for receiving purchase requests for the satellite images 921 and 922. The satellite images 921 and 922 may be identified based on the user's dialogue history (for example, satellite images searched based on previous dialogues, current dialogues, etc.), and the control unit 150 can sell the satellite images 921 and 922 to the user based on the selection of the graphic object 923. In this case, the user may be given specific benefits depending on whether they purchase all of the satellite images 921 and 922 or only one.

[0238] On the other hand, in the second mode, various services can be offered through subscriptions to paid plans.

[0239] The control unit 150 can provide answers to user queries based on the available models.

[0240] In one embodiment, if the control unit 150 identifies a model for generating a response to a user query and the identified model is included in the list of models available to the user account, it can use the output of the identified model to generate a response to the user query.

[0241] On the other hand, in another embodiment, as shown in Figure 10, if the control unit 150 identifies a model for generating a response to a user query 1011a (e.g., "I would like you to analyze the NDWI for this region"), and the identified model is not included in the list of models available in the user account, it can provide the first region 1010 with a response 1011b requesting payment related to a paid plan subscription (e.g., "An NDWI analysis model is used to analyze the NDWI for this region, and this requires a subscription to an additional paid service"), and a graphic object 1012a linked to the paid plan subscription function. In this case, the paid plan subscription may relate to a subscription to use a specific model that performs a specific function (e.g., NDWI analysis). At this time, if the paid plan is charged in the user account in response to the selection of the graphic object 1012a, the control unit 150 can provide the user with the NDWI analysis results for a specific region 1022 analyzed using a specific model (NDWI analysis model). However, it goes without saying that the paid plan subscription is not necessarily limited to a subscription to use a specific model, and can be varied in various ways depending on the purpose of the present invention.

[0242] On the other hand, the present invention can provide various services (or personalized services) based on various information (or historical information) stored in conjunction with a user account.

[0243] The control unit 150 can perform clustering of various information stored in conjunction with user accounts based on various criteria. Here, clustering may mean grouping data with similar characteristics into one cluster (or group or ensemble), and separating data with different characteristics into different clusters.

[0244] In the present invention, the clustering criteria can be set in various ways. For example, when performing clustering on satellite imagery, the various criteria may include at least one of the following: i) location information matched to the satellite imagery, ii) date of capture matched to the satellite imagery, and iii) attributes of the satellite imagery (e.g., satellite imagery based on vegetation index analysis results, satellite imagery including orbital information, etc.).

[0245] In one embodiment, as shown in Figure 11, the control unit 150 can perform clustering of satellite images matched to a user account. In this case, the first group 1110 and the third group 1130 may include satellite images having the same location information (or coordinate information). In contrast, the second group 1120 may include satellite images based on satellite image attributes (for example, satellite images based on vegetation index analysis results).

[0246] In this case, a graphic object linked to a function that allows the user to purchase satellite imagery included in each group may be provided in one area of ​​the service page 1000. For example, based on the user terminal 10 selecting a first graphic object 1110a from among multiple graphic objects 1110a and 1110b included in one area of ​​the service page 1000, the control unit 150 can sell all the satellite imagery included in the first group 1110 to the user. As another example, if a second graphic object 1110b is selected, the control unit 150 can sell at least one satellite imagery selected by the user from among the satellite imagery included in the first group 1110 to the user.

[0247] On the other hand, the present invention may further include a third tool icon in addition to the first tool icon 622a and the second tool icon 622b shown in Figure 6b.

[0248] In one embodiment, as shown in Figure 12, the control unit 150 can provide a third tool icon 1210 linked to a specific function on the service page 1000. The third tool icon 1210 can work in conjunction with a specific function that allows the user to display specific coordinates they intend to include in their query on a map, in parallel with a method of extracting a specific area from a user query and searching for coordinates. Such a specific function can be used when searching satellite imagery containing specific coordinates.

[0249] When the third tool icon 1210 is selected from the user terminal 10, the control unit 150 can display the graphic object 1212 linked to the third tool icon 1210 on at least a portion of the map image 1211. When the third tool icon 1210 is selected, the graphic object 1212 may be displayed on the map image 1211 in a first visual appearance (for example, the shape of a blue circle). Furthermore, when the graphic object 1212 is displayed on the map, its geometry information can be stored in the response service provision system 100 (or front end).

[0250] As another example, when the user terminal 10 selects a specific area (or location) on the map image 1211 while the graphic object 1212 is not activated (or selected), the control unit 150 may display the graphic object 1212 on the map image 1211 with a second visual appearance (for example, a circle with a blue border on a white background).

[0251] On the other hand, the present invention can provide various functions related to the responses of the large-scale language model 160.

[0252] In one embodiment, as shown in Figure 13, the control unit 150 can, upon receiving a user query 1301 requesting a list of satellite images including a specific location ("Show me the satellite image of this location."), provide a response 1302 to the service page 1000 that includes a list of satellite images related to the location corresponding to the user query 1301. Here, the specific location may be an area specified by the user through at least one tool icon. At the same time as providing the response 1302, the control unit 150 may also render the most recently taken satellite image 1312 onto the map image 1311. This allows the user to set a specific region of interest to utilize the object detection function, or to render other satellite images on the map to continue interacting with the chatbot (or large language model).

[0253] In another embodiment, the control unit 150 may provide a graphic object 1320 in a region of the service page 1000 that is linked to a function that allows the user to select a chatbot model that provides answers to queries. When the graphic object 1320 is selected from the user terminal 10, the control unit 150 may provide a list 1321 containing a plurality of chatbot models 1321a, 1321b, and 1321c, and the user may select at least one of the different chatbot models 1321a, 1321b, and 1321c included in the list 1321. In this case, the control unit 150 can use the specific model selected by the user to provide answers to queries entered by the user.

[0254] As described above, the interactive artificial intelligence response service provision method and system according to the present invention can generate and provide to the user a response that improves the hallucinatory phenomenon by utilizing the retrieved document. This allows the present invention to reduce the provision of false information and contribute to providing users with reliable information.

[0255] Furthermore, according to the interactive artificial intelligence response service provision method and system of the present invention, satellite data related to the user's query can be analyzed, and a response to the user's query can be provided along with visual information. This allows the user to more intuitively confirm the response to their query.

[0256] Furthermore, according to the interactive artificial intelligence response service provision method and system of the present invention, by comprehensively analyzing various data, including satellite imagery, in response to user queries and providing various information that the user needs, the user can utilize the provided information in various fields such as agriculture, environmental monitoring, and urban planning, thereby improving efficiency across various industries.

[0257] Furthermore, according to the interactive artificial intelligence response service provision method and system of the present invention, it is possible to provide answers that match the user's queries by utilizing information identified in response to the user's request. In other words, the present invention can provide accurate and reliable information that corresponds to the user's requirements without performing a complex information retrieval process.

[0258] On the other hand, the present invention described above can be embodied as a program that is executed by one or more processes on a computer and can be stored on a medium (or recording medium) that is readable by such a computer.

[0259] Furthermore, the present invention described above can be embodied as computer-readable code or instruction words on a medium on which a program is recorded. That is, the present invention can be provided in the form of a program.

[0260] On the other hand, computer-readable media include all types of recording devices that store data readable by a computer system. Examples of computer-readable media include HDDs (Hard Disk Drives), SSDs (Solid State Disks), SSDs (Silicon Disk Drives), ROMs, RAMs, CD-ROMs, magnetic tapes, floppy disks, and optical data storage devices.

[0261] Furthermore, the computer-readable medium may include storage and may be a server or cloud storage accessible by electronic devices via communication. In this case, the computer can download the program according to the present invention from the server or cloud storage via wired or wireless communication.

[0262] Furthermore, in this invention, the computer described above is an electronic device equipped with a processor, i.e., a CPU (Central Processing Unit), and its type is not particularly limited.

[0263] On the other hand, the above detailed description should not be interpreted restrictively in any way, but should be considered illustrative. The scope of the invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the scope of the equivalents of the invention are included within the scope of the invention.

Claims

1. The steps include receiving user queries via a service page, The steps include identifying location information from the user query, The steps include: using the user query and location information, searching for satellite imagery related to the user query from at least one database; The steps include: performing analysis related to the user query using the aforementioned satellite imagery; A step of generating a prompt using at least one of the user query, location information, analysis results from the analysis, and satellite imagery, The steps include processing the aforementioned prompt as input to a Large-Scale Language Model (LLM), The steps include obtaining the answer to the user query from the large-scale language model, A method for providing an interactive artificial intelligence response service, characterized by comprising the step of providing a response to the user query on the service page.

2. The aforementioned service page is The first area that receives the user query, A method for providing an interactive artificial intelligence response service according to claim 1, characterized by including a second region that provides the satellite imagery retrieved in response to the user query.

3. In the first domain, information based on text related to the user query is provided. The method for providing an interactive artificial intelligence response service according to claim 2, characterized in that the second domain is provided with information based on images related to the user query.

4. The second area displays a map image. When the satellite imagery is retrieved in response to the user query, The method for providing an interactive artificial intelligence response service according to claim 3, characterized in that the satellite imagery is superimposed on at least a portion of the map image.

5. The method for providing an interactive artificial intelligence response service according to claim 4, characterized in that the area in the map image in which the satellite image overlaps is identified based on the location information identified for searching the satellite image.

6. The steps include receiving user input specifying a particular area in the satellite imagery provided to the second area, The steps include receiving additional user queries related to the specific area in the first area, A method for providing an interactive artificial intelligence response service according to claim 2, further comprising the step of generating a response to an additional user query using satellite imagery of the specific region, metadata related to the specific region, and at least one of the additional user queries.

7. The step of generating responses to the aforementioned additional user queries is: The steps include analyzing the user intent of the aforementioned additional user queries, The steps include identifying a function that corresponds to the user intent and selecting a specific analytical model that performs the identified function, The steps include processing satellite imagery of the aforementioned specific region as input to the aforementioned specific analysis model, The steps include obtaining output corresponding to the user intent from the aforementioned specific analysis model, Using the output, the step of generating a response to the additional user query, A method for providing an interactive artificial intelligence response service according to claim 6, characterized by comprising the step of providing a response to the additional user query to at least one of the first and second domains.

8. The response to the additional user query provided in the second area is: The method for providing an interactive artificial intelligence response service according to claim 7, characterized in that it includes highlighting at least a portion of the satellite image of the specific region provided to the second region.

9. The service page provides at least one tool icon for receiving user input for the satellite imagery provided in the second area, The aforementioned tool icons are linked to different functions, Of the tool icons, the first tool icon is linked to a first selection function that selects at least a portion of the satellite imagery according to a first criterion. The method for providing an interactive artificial intelligence response service according to claim 6, characterized in that the second tool icon among the tool icons is linked to a second selection function that selects at least a portion of the satellite imagery according to a second criterion different from the first criterion.

10. When the user query is received, the user query is processed as input to the large-scale language model, and a model is selected from the large-scale language model to perform the analysis corresponding to the user query. The steps include inputting the satellite image as input to the selected model, The method for providing an interactive artificial intelligence response service according to claim 1, further comprising the step of receiving the results of an analysis of the satellite image corresponding to the user query from the selected model.

11. In the step of generating the aforementioned prompt, The analysis results received from the selected model are included in the prompt. In the aforementioned large-scale language model, A method for providing an interactive artificial intelligence response service according to claim 10, characterized in that it generates a response to the user query using the analysis results and the satellite imagery.

12. The steps include receiving user queries via a service page that includes the first and second domains, The steps include identifying the document used to generate the answer to the user query, The steps include processing the user query and the identified document as input to a large-scale language model, The steps include obtaining the answer to the user query from the large-scale language model, The steps include providing the aforementioned answer on the aforementioned service page, and further, A method for providing an interactive artificial intelligence response service, characterized in that the first area of ​​the service page is provided with the response, and the second area of ​​the service page is provided with information about the identified document.

13. A system for providing interactive artificial intelligence response services, The system includes memory and at least one processor, The memory and the processor cooperate, We receive user queries via the service page. Location information is identified from the aforementioned user query, Using the user query and location information, a search is performed for satellite imagery related to the user query from at least one database. Using the aforementioned satellite imagery, perform analysis related to the user query. A prompt is generated using at least one of the user query, location information, analysis results from the analysis, and satellite imagery. The aforementioned prompt is processed as input to a Large-Scale Language Model (LLM), From the aforementioned large-scale language model, obtain the answer to the user query. An interactive artificial intelligence response service provision system characterized by providing answers to user queries on the aforementioned service page.

14. A program executed by one or more processes in an electronic device and stored on a computer-readable recording medium, The steps include receiving user queries via a service page, The steps include identifying location information from the user query, The steps include: using the user query and location information, searching for satellite imagery related to the user query from at least one database; The steps include: performing analysis related to the user query using the aforementioned satellite imagery; A step of generating a prompt using at least one of the user query, location information, analysis results from the analysis, and satellite imagery, The steps include processing the aforementioned prompt as input to a Large-Scale Language Model (LLM), The steps include obtaining the answer to the user query from the large-scale language model, A program stored on a computer-readable recording medium, characterized by including a command to perform the step of providing a response to a user query on the service page.

Images