Service providing apparatus and method for supporting creation of personalized web page for product sales

Abstract

The present invention relates to a service providing device and method for supporting the creation of a personalized web page for product sales. More specifically, it relates to a service providing device and method that automatically generates a personalized web page optimized for the type of product for sale by analyzing minimal information provided by a user, the user's SNS account, and the latest trends, and supports the user in intuitively and easily editing the generated web page through a chatbot. The present invention automatically extracts personalized information such as user profiles, post content, interest tags, and follower information through the analysis of minimal information provided by a user and SNS accounts, automatically classifies the type of product to be sold based on product registration information, and obtains trend information by analyzing the layout patterns of popular sales pages belonging to the same category as the user's product. By generating a personalized sales page with differentiated content based on personalized information, a competitive layout reflecting trend information, and sales options optimized for the product type automatically set, the invention has the effect of supporting primary producers, such as agricultural producers who are not familiar with the use of information technology, or novice sellers, to easily build a professional and competitive sales page without specialized knowledge of market research or competitor analysis.

Description

Technology Field

[0001] The present invention relates to a service providing device and method that supports the creation of a personalized web page for product sales, and more specifically, to a service providing device and method that automatically generates a personalized web page optimized for the type of product to be sold by analyzing minimal information provided by a user, the user's SNS account, and the latest trends, and supports the user in intuitively and easily editing the generated web page through a chatbot. Background Technology

[0002] With the recent rise in online commerce, there is an increasing number of cases where primary producers, such as agricultural producers, sell their products directly to consumers through open markets or their own shopping malls.

[0003] For such online sales, it is essential to open a web page (hereinafter referred to as the 'sales page') to introduce and sell products.

[0004] Existing open market platforms and e-commerce solutions provide template-based input methods to support sellers in creating sales pages. This method involves sellers selecting one of several predefined templates and constructing the sales page by entering information step-by-step for each component of the template.

[0005] However, this template-based method of generating sales pages requires a complex multi-step input process, which poses a problem as it is difficult for elderly farmers and fishermen or novice sellers who are not familiar with digital devices to use.

[0006] In particular, in the recent online market, differentiated sales strategies and personalized marketing are becoming increasingly important for each seller, even for identical products; however, it is difficult to implement such personalization using existing standardized template methods. Moreover, creating personalized pages tailored to a seller's characteristics or customer base requires additional design work or professional assistance, resulting in significant time and cost.

[0007] Furthermore, even when sellers wish to structure their sales pages by leveraging customer bases or personal brands established through social media, existing systems lack the functionality to integrate and utilize such social media information. Consequently, sellers face the inconvenience of having to manually re-enter images or content generated on social media into their sales pages.

[0008] Furthermore, despite the fact that the necessary sales options vary significantly depending on the type of product, existing systems lack the capability to automatically recognize and optimize them. For instance, while selling physical goods requires options such as quantity, color, and size, selling online courses necessitates options like curriculum, schedule, and duration, and selling accommodation services requires options such as reservation dates, number of guests, and room type. However, under existing systems, sellers must manually identify the product type and set appropriate options themselves, posing a significant barrier to entry for novice sellers.

[0009] Furthermore, despite the fact that configuring sales pages to reflect rapidly changing market trends has a significant impact on revenue, existing systems lack the functionality to automatically analyze the latest trends and incorporate them into sales pages. As a result, sellers must conduct market research and competitor analysis separately, and specialized marketing knowledge is required to reflect these findings on the sales pages.

[0010] Meanwhile, most open marketplaces restrict users from directly uploading content containing scripts such as HTML (HyperText Markup Language) or JavaScript to the area displaying detailed product information. Due to these restrictions, sellers use a method of creating and registering product descriptions as a single long image file; however, this presents a problem in that it is difficult to quickly update product information because the image file is difficult to modify once registered. Prior art literature

[0011] Korean Patent Publication No. 10-2022-0121337 The problem to be solved

[0012] The present invention aims to support users, such as primary producers who have difficulty utilizing information technology, in easily and automatically generating personalized product sales web pages with minimal information input, and to enable the configuration of differentiated sales pages without the need for separate additional information input by automatically analyzing the user's SNS account information and automatically generating personalized sales pages based on it.

[0013] Furthermore, the present invention aims to support the creation of a competitive sales page by automatically classifying the types of products to be sold, analyzing trends related to product types to identify the sales strategies and page configurations of popular products within the same category as the product type, and automatically reflecting this on the user's sales page. Additionally, by automatically setting sales options optimized for product types, it aims to support even novice sellers in configuring an appropriate sales page without professional knowledge.

[0014] In addition, the present invention aims to significantly enhance user convenience in web page management by supporting web page modification using a chatbot, thereby enabling the user to intuitively select the target to be modified on the web page when editing the web page through a chatbot interface, and by enabling editing of the target using natural language-based voice commands. means of solving the problem

[0015] A service providing device for supporting the creation of a personalized web page for product sales according to an embodiment of the present invention may include: a user analysis unit that generates personalized information by analyzing a user’s SNS account page; a product classification unit that automatically classifies the type of the product to be sold into one of a plurality of pre-set product types to create a sales page, which is a web page for the sale of the product to be sold, based on product registration information of the product to be sold that the user intends to sell; a trend analysis unit that generates trend information by analyzing one or more other sales pages in which product types belonging to the same category as the specific product type classified by the product classification unit are set; and a page generation unit that generates a personalized sales page based on the personalized information, product registration information, trend information, and the specific product type, wherein the page layout is determined based on the trend information and the product registration information, and a sales option corresponding to the specific product type is automatically set.

[0016] As an example related to the present invention, the user analysis unit may be characterized by extracting at least one of a user profile, post content, interest tags, and follower information from the SNS account page, and generating personalized information including at least one of the user's interest field, target customer base, and post content based on the extracted information.

[0017] As an example related to the present invention, the trend analysis unit may be characterized by analyzing a page layout pattern for one or more popular sales pages that satisfy a preset selection condition among one or more other sales pages in which a product type belonging to the same category as the specific product type is set, and generating trend information regarding the page layout pattern.

[0018] As an example related to the present invention, the page generating unit may be characterized by adding at least one of quantity, size, color, capacity, packaging method, sales unit, delivery method, payment method, availability period, license type, course schedule, course period, number of students, reservation date, number of guests, room type, schedule, and calendar to the sales options according to the specific product type.

[0019] As an example related to the present invention, the page generation unit may be characterized by generating a landing page to introduce a user's product for sale based on the personal information, generating product summary information based on the sales page when the sales page is generated, and automatically adding the product summary information, which includes link information connecting to the sales page, to the landing page.

[0020] As an example related to the present invention, a service providing device that supports the creation of a personalized web page for the sale of the product may further include: an editing interface providing unit that creates and provides an editing interface including an agent unit that identifies a selection area selected by a user on the sales page, and includes an editing request information for an editing request for the sales page, and the editing processing unit that receives editing content information regarding the editing target included in the selection area through the chatbot interface included in the editing interface, receives selection area information regarding the selection area identified by the agent unit, and updates the sales page by editing the editing target on the sales page through an artificial intelligence-based chatbot model based on the sales page information regarding the sales page, the selection area information, and the editing content information.

[0021] As an example related to the present invention, the editing interface is configured to display a page display area for displaying the sales page and the chatbot interface together on a single screen so that a user can input editing content through the chatbot interface while selecting an editing target in the page display area, and the agent unit is configured to identify a selection area containing the editing target selected by the user through the editing interface, generate selection area information for the selection area, and automatically input it into the chatbot interface or transmit it to the editing processing unit.

[0022] As an example related to the present invention, the editing processing unit may be characterized by receiving selection area information generated by the agent unit and editing content information entered by a user through the chatbot interface, applying the received selection area information, editing content information, and sales page information to the chatbot model to generate editing command information so that editing according to the editing content information is performed on a specific editing target within the selection area of ​​the sales page through the chatbot model, and updating the sales page by editing the editing target within the selection area according to the editing command information, or obtaining the updated sales page through the chatbot model in which editing according to the editing content information is performed on the editing target by applying the received selection area information, editing content information, and sales page information to the chatbot model.

[0023] As an example related to the present invention, the chatbot interface may be characterized by including a voice recognition function that converts voice input into text, thereby enabling a user to input the editing content information by voice.

[0024] A service provision method that supports the creation of a personalized web page for product sales by a service provision device according to an embodiment of the present invention may include: a step of generating personalized information by analyzing a user’s SNS account page; a step of automatically classifying the type of the product to be sold into one of a plurality of pre-set product types to create a sales page, which is a web page for selling the product to be sold, based on product registration information of the product to be sold that the user intends to sell; a step of generating trend information by analyzing one or more other sales pages in which product types belonging to the same category as the specific product type classified in the product classification step are set; and a step of creating a personalized sales page based on the personalized information, product registration information, trend information, and specific product type, wherein the page layout is determined based on the trend information, and sales options corresponding to the specific product type are automatically set, and the personalized sales page includes product description content generated based on the personalized information and product registration information. Effects of the invention

[0025] The present invention automatically extracts personalized information such as user profiles, post content, interest tags, and follower information through analysis of SNS accounts and minimal information provided by the user, automatically classifies the types of products to be sold based on product registration information, and obtains trend information by analyzing the layout patterns of popular sales pages belonging to the same category as the user's products. By generating a personalized sales page with differentiated content based on personalized information, a competitive layout reflecting trend information, and sales options optimized for product types automatically set, it has the effect of supporting primary producers, such as agricultural producers who are not familiar with using information technology, or novice sellers, to easily build a professional and competitive sales page without specialized knowledge of market research or competitor analysis.

[0026] In addition, the present invention provides an editing interface that simultaneously displays a page display area and a chatbot interface on a single screen, and by the agent unit automatically generating HTML element information, coordinate information, captured image, location description information, etc. of the editing target selected by the user and transmitting it to the chatbot model (138), it supports users without web programming knowledge to intuitively select the editing target and edit it using natural language commands or voice input, and even when the same content exists in multiple places on the sales page, it supports users to accurately identify and edit the editing target at a specific location intended by the user, thereby enabling all users, including elderly users who have difficulty using a keyboard, to easily manage web pages without complex code modifications, thus contributing to the activation of direct transactions between producers and small business owners and the expansion of their entry into the online market. Brief explanation of the drawing

[0027] FIG. 1 is a configuration diagram of a service providing system that supports the creation of a personalized web page for product sales according to an embodiment of the present invention. FIG. 2 is a configuration diagram of a service providing device that supports the creation of a personalized web page for product sales according to an embodiment of the present invention. FIGS. 3 to 6 are example diagrams of the operation of a service providing device that supports the creation of a personalized web page for product sales according to an embodiment of the present invention. FIG. 7 is a flowchart of a service provision method that supports the creation of a personalized web page for product sales according to an embodiment of the present invention. FIGS. 8 and 9 are example diagrams of a sales page generated by a service providing device that supports the creation of a personalized web page for product sales according to an embodiment of the present invention. FIGS. 10 to 13 are exemplary diagrams of a web page editing process using an editing interface provided by a service providing device according to an embodiment of the present invention. Specific details for implementing the invention

[0028] Hereinafter, detailed embodiments of the present invention will be described with reference to the drawings.

[0029] FIG. 1 is a configuration diagram of a service providing system (hereinafter, service providing system) that supports the creation of a personalized web page for product sales according to an embodiment of the present invention.

[0030] As described above, a service providing system according to an embodiment of the present invention may be configured to include a user terminal (10) possessed by a user, a service providing device (100) that communicates with the user terminal (10) through a communication network and performs the creation of a personalized web page for product sales, and one or more external servers such as an external SNS server that provides information related to the user's SNS (Social Network Service) account, and one or more external shopping mall servers that provide shopping mall-related websites.

[0031] At this time, the user terminal (10) described in the present invention may be configured to include a terminal communication unit, a terminal storage unit, a terminal input unit, a terminal display unit, a terminal control unit, a voice input unit, etc.

[0032] First, the terminal communication unit can communicate with the service providing device (100) through a communication network.

[0033] In addition, the terminal input unit may receive user input, generate input information, and provide the input information to the terminal control unit.

[0034] Additionally, the terminal display unit may display various information according to the control of the terminal control unit, and the terminal input unit and the terminal display unit may be composed of components that constitute a touch screen configured in the user terminal (10).

[0035] In addition, the terminal storage unit can store various data and programs necessary for the operation of the terminal control unit.

[0036] In addition, the voice input unit may receive the user's voice, generate a voice signal, and provide it to the terminal control unit, and the voice input unit may be configured to include a microphone.

[0037] In addition, the terminal control unit performs overall control functions of the user terminal (10) and may include RAM, ROM, CPU, GPU, and a bus, and the RAM, ROM, CPU, GPU, etc. may be connected to each other through the bus.

[0038] Additionally, the terminal storage unit stores application-related data for performing the function of processing information received from the service providing device (100), generating information to be transmitted to the service providing device (100), and transmitting it to the service providing device (100), and the terminal control unit may operate as an application unit by executing the application.

[0039] Additionally, the terminal storage unit may store web browser-related data, and the terminal control unit may access a main website provided by the service providing device (100) using the web browser, and may process information provided by the service providing device (100) through the main website using the web browser.

[0040] The operation configuration of the user terminal (10) described below may be performed by the terminal control unit or application unit.

[0041] In addition, the user terminal (10) may be composed of various terminals such as a smartphone, portable terminal, mobile terminal, personal computer, laptop computer, tablet PC, etc., equipped with communication functions.

[0042] In addition, the communication network described in the present invention may include various widely known wired / wireless communication networks.

[0043] In addition, the service providing device (100) may be configured as a server.

[0044] Additionally, the external SNS server may be configured as a server providing social network services (SNS) such as KakaoTalk, Naver, Instagram, Google, Facebook, etc., and may process user authentication through standard protocols such as OAuth 2.0, and when the service providing device (100) checks the access rights to an SNS account page, which is a web page corresponding to the user's SNS account stored and provided on the external SNS server, it may provide various resource information included in the SNS account page to the service providing device (100).

[0045] Accordingly, the service providing device (100) automatically generates a sales page, which is a web page for selling products personalized to the user based on information collected from the user's SNS account page (collected information), thereby supporting the user in easily configuring a product sales site, which is a website for selling products, and also supports the user in easily managing such management, such as editing, through conversation. The configuration of the service providing device (100) related thereto will be explained in detail below with reference to the drawings.

[0046] FIG. 2 is a detailed configuration diagram of a service providing device (100) according to an embodiment of the present invention, FIG. 3 to 6 are example diagrams of operation of a service providing device (100) according to an embodiment of the present invention, and FIG. 7 is a flowchart of operation for a service providing method that supports the creation of a personalized web page for product sales of a service providing device (100) according to an embodiment of the present invention.

[0047] As illustrated in FIG. 2, a service providing device (100) according to an embodiment of the present invention may be configured to include a communication unit (110), a storage unit (120), a control unit (130), etc.

[0048] The above communication unit (110) can transmit and receive various information by communicating with one or more user terminals (10), various external devices, various external servers (external SNS servers, external shopping mall servers, etc.) through a communication network.

[0049] The above storage unit (120) can store various types of information, and the storage unit (120) can be configured in various forms such as an HDD (Hard Disk Drive), an SSD (Solid State Drive), etc., and may be configured as a DB or may include one or more DBs.

[0050] For example, the storage unit (120) may include a service DB in which service information for providing services related to an Open Market platform is stored, a member DB in which member information for each user is stored, a product DB in which information related to the user's products is stored, and a site DB in which information about the user's product sales site including the user's sales page is stored.

[0051] At this time, the above service DB, the above member DB, the above product DB, the above site DB, etc., may each be configured as individual database servers and configured to link (communicate) with the above service providing device (100).

[0052] In addition, any one of the above service DB, above member DB, above product DB, and site DB may be included in the other.

[0053] Additionally, the control unit (130) performs overall control functions of the service providing device (100), and the control unit (130) may include RAM, ROM, CPU, GPU, and a bus, and the RAM, ROM, CPU, GPU, etc. may be connected to each other through a bus.

[0054] Additionally, the communication unit (110) and storage unit (120) may be configured to be included in the control unit (130).

[0055] Additionally, the control unit (130) may be configured to include a service management unit (131), a user analysis unit (132), a product classification unit (133), a trend analysis unit (134), a page generation unit (135), an editing interface provision unit (136), an editing processing unit (137), etc., and at least one of the components constituting the control unit (130) may be included in another.

[0056] In addition, a plurality of components constituting the control unit (130) may be implemented by a processor capable of data processing, and each may be implemented by a different processor separately, or may be functionally separated within a single processor.

[0057] In addition, at least one of the control unit (130) and the various components constituting the control unit (130) can communicate with one or more external devices, such as a user terminal (10), an external SNS server, an external shopping mall server, etc., through the communication unit (110), and the communication configuration through the communication unit (110) is omitted below.

[0058] First, as illustrated in FIG. 3, the service management unit (131) communicates with the user terminal (10) to perform member management functions such as user access management, membership registration, login, logout, and modification of member information, and can provide overall functions related to the service of the service providing device (100).

[0059] For example, the service management unit (131) can operate a main website by utilizing service-related data for providing a website related to an open market platform stored in the service DB, and can support receiving a membership request from a user terminal (10) through the main website to perform membership registration (S1).

[0060] At this time, the main website may be composed of one or more web pages, and the service management unit (131) may perform various functions provided through the main website while communicating with the user terminal (10) through the one or more web pages.

[0061] In particular, when the service management unit (131) receives a request for SNS easy login from the user terminal (10) as the user selects an SNS easy login button (KakaoTalk login, Naver login, Google login, Facebook login, etc.) displayed on the login page or sign-up page of the main website, it can redirect the user terminal (10) to an authentication page of an external SNS server corresponding to the selected SNS service and perform OAuth authentication in conjunction with the external SNS server.

[0062] For example, when a user selects Facebook easy login through a user terminal (10), the service management unit (131) redirects the user terminal (10) to a Facebook-related authentication server to perform user authentication, and when the user completes login and consent to provide information at the Facebook-related authentication server, it obtains an access token from the Facebook-related authentication server, stores it in the member DB, and transmits the access token to the user analysis unit (132).

[0063] Meanwhile, the user analysis unit (132) can use the access token received from the service management unit (131) to access the SNS account page, which is a web page corresponding to the user's SNS account provided by an external SNS server, and analyze the SNS account page to generate personalized information (S2).

[0064] At this time, the above SNS account page may include the user's SNS account information stored on an external SNS server.

[0065] For example, the user analysis unit (132) can call an API (Application Programming Interface) of an external SNS server using the access token, and access a web page corresponding to the user's SNS account through the API, and extract and collect user profile, post content (posted content), interest tags (hashtags), follower information, etc. from the SNS account page.

[0066] Additionally, the user analysis unit (132) can analyze information collected from the SNS account page using a natural language processing (NLP) algorithm and an image analysis model, generate personalized information including the user's interests, target customer base, and post content, and store it in the member DB by matching it with the user's member information or transmit it to the page generation unit (135).

[0067] For example, when the user analysis unit (132) analyzes the user's Facebook account page, it can access the user's Facebook account page through the Facebook API using the access token, collect content posted by the user and various information from the user's Facebook account page, analyze the collected images through an image analysis model or analyze the hashtags of the Facebook account page to extract areas of interest, and analyze the profiles of followers corresponding to the user to identify the characteristics of the target customer base, and generate personalized information including the areas of interest, target customer base, and post content.

[0068] At this time, the personal information may be configured to include one or more post content (images, text, etc.) posted (registered) by the user on the SNS account page, so that the page creation unit (135) can utilize the content when creating a sales page.

[0069] In the configuration described above, the service management unit (131) can support the use of SNS account information even when the user registers directly on the main website without using SNS simple login.

[0070] For example, when a user completes membership registration by entering the user’s SNS account (e.g., SNS username, profile URL, etc.) registered on an external SNS server as member identification information (ID) to be used on the main website on the membership registration page of the main website, the service management unit (131) can send an SNS account linkage request message to the user terminal (10) after membership registration is completed.

[0071] Afterwards, the service management unit (131) receives a response message to the SNS account linkage request message from the user terminal (10), and if it is confirmed that the user has agreed to the SNS account linkage based on the response message, the service management unit (131) connects the user terminal (10) to the authentication page of the external SNS server to perform OAuth authentication, and if the user agrees to provide information, it can obtain an access token from the external SNS server and store it in the member DB.

[0072] Afterwards, the user analysis unit (132) can use the issued access token to access the SNS account page corresponding to the SNS account ID entered by the user when signing up and collect information necessary for generating personalized information.

[0073] Alternatively, in the case of an SNS account set as a public profile by the user, the user analysis unit (132) may collect only limited information, such as profile information and public posts disclosed through the public API of the external SNS server without separate authentication, and use it to generate personalized information.

[0074] Additionally, as illustrated in FIGS. 3 and 6, the page generation unit (135) can be linked with the user analysis unit (132) to generate a landing page for introducing the user's product for sale based on the personalized information when generating personalized information corresponding to the user, and can generate a payment page, which is a web page for payment of the product to be sold, through a sales page connected via a link to the landing page (S3).

[0075] At this time, the page generation unit (135) can store the landing page and payment page generated through the page generation model in the site DB and set the landing page to be displayed as the main page when the user accesses the product sales site.

[0076] For example, when the page generation unit (135) receives user profile information (profile photo, self-introduction text, etc.) included in the personal information from the user analysis unit (132), it can generate a landing page that applies the profile information to the seller introduction area.

[0077] For example, if the user’s SNS account page, such as a Facebook account page, has an introduction phrase “OO Farm” and a user image, the page creation unit (135) can use the introduction phrase and user image to create a seller introduction section at the top of the landing page, and can select images related to farms or agricultural products from the post content included in the personal information and place them in the background image or gallery section of the landing page.

[0078] Additionally, the page generation unit (135) stores the generated landing page in the site DB and transmits the URL information of the landing page to the service management unit (131), and the service management unit (131) can store the URL information in the member DB by matching it with the user's member information.

[0079] At this time, the landing page may be configured to include a product list area to link to the sales page of an individual product to be created in the future.

[0080] Additionally, the page generation unit (135) may generate a payment page along with the landing page generation. The payment page may include a basic payment interface that links with sales pages to be generated in the future, enabling a buyer to pay for the product.

[0081] For example, the page generation unit (135) can generate a payment page in which seller information (seller name, contact information, etc.) is pre-entered based on user information included in the personalized information, and can generate a standard payment page including a buyer information input field, a shipping address information input field, and a payment method selection area (credit card, bank transfer, simple payment, etc.) and store it in the site DB.

[0082] Additionally, the page generation unit (135) transmits the URL information of the generated payment page to the service management unit (131), and the service management unit (131) can store the payment page URL information in the member DB by matching it with the user's member information.

[0083] The landing page and payment page generated in this way are configured to be automatically linked when the user registers individual products to create a sales page later.

[0084] Meanwhile, as illustrated in FIG. 4, the service management unit (131) can receive product registration request information including product information, product images, and product descriptions for products to be sold from the user terminal (10) after the landing page is created (S4).

[0085] For example, the service management unit (131) can provide a product registration interface to the user terminal (10) when a user logged in with administrator privileges selects the "Product Registration" button while accessing the landing page, and can receive product registration request information including a product name, product image, and a brief product description entered through the product registration interface from the user terminal (10) and transmit it to the product classification unit (133).

[0086] At this time, the product registration interface may be composed of a separate web page, a modal window, or a pop-up window.

[0087] In addition, the product classification unit (133) can generate product registration information based on product registration request information received from the service management unit (131), and can automatically classify the type of the product to be sold into one of a plurality of pre-set product types based on the generated product registration information (S5).

[0088] For example, the product classification unit (133) may include a deep learning-based product classification model that has been pre-trained with a plurality of different product types, and by inputting the product name, product description, and product image included in the product registration request information into the product classification model, the product classification model can automatically classify which of the plurality of product types, such as food, clothing, electronic products, household goods, digital content, online courses, offline courses, accommodation services, etc., the product belongs to.

[0089] At this time, the product type may be composed of a hierarchical structure of major, medium, and minor categories, and the product classification unit (133) may classify the product to be sold into at least one hierarchy among the major, medium, and minor categories.

[0090] For example, if a user registers product registration information including a product name for a product to be sold called "strawberry chip" and a product image related to strawberry chip, the product classification unit (133) can classify the product type corresponding to the product registration information as a "processed fruit product" that is subdivided from the "food" category through the product classification model.

[0091] Alternatively, if the user registers product registration information including “yoga class” and images of products related to yoga classes as products for sale, the product classification unit (133) can classify the product type as ‘offline course’.

[0092] Additionally, the product classification unit (133) can transmit product type information for a specific product type classified for a product to be sold corresponding to the product registration information to the trend analysis unit (134), and store the product registration information in the product DB.

[0093] At this time, the product registration information may include the user identification information of the above user.

[0094] Additionally, when the trend analysis unit (134) receives product type information classified from the product classification unit (133), it can generate trend information by analyzing one or more popular sales pages that satisfy pre-set selection conditions among one or more other sales pages where a specific product type is set according to the product type information, or where a product type belonging to the same category as the specific product type is set (S5).

[0095] At this time, the aforementioned other sales page may be a sales page corresponding to other users different from the user.

[0096] For example, the trend analysis unit (134) includes a web crawling engine and, through the web crawling engine, collects one or more sales pages for products belonging to the same category as the specific product type from one or more external servers that provide open market or shopping mall websites, and can select popular sales pages among the collected one or more sales pages that have a sales volume or view count above a preset threshold.

[0097] At this time, the trend analysis unit (134) may determine the ranking of popular sales pages based on at least one of sales volume and views, and may generate the trend information by analyzing only one or more popular sales pages within a preset standard ranking (e.g., top 10).

[0098] In addition, the trend analysis unit (134) can perform trend analysis on one or more popular sales pages to generate trend information regarding page layout patterns.

[0099] For example, if a product 'strawberry chip' according to product registration information received from a user terminal (10) is classified into a specific product type such as 'processed fruit product' by the product classification unit (133), the trend analysis unit (134) can crawl and collect one or more other sales pages belonging to the same category as the 'processed fruit product' from an external server.

[0100] Specifically, when a product to be sold corresponding to the product registration information is classified as a specific product type of ‘fruit processing product’, the trend analysis unit (134) can collect sales pages of strawberry chips, banana chips, dried fruit, etc., that have the same product type as the specific product type.

[0101] In addition, the trend analysis unit (134) may select one or more popular sales pages among the collected other sales pages that have a monthly sales volume of 100 or more or a daily view count of 1000 or more as the target for analysis.

[0102] In addition, the trend analysis unit (134) can analyze the page layout structure by parsing HTML (HyperText Markup Language) code for each of one or more selected popular sales pages, and can derive the most frequently used page layout pattern (or layout pattern) by extracting the position and size of product images, the arrangement pattern of text content, the position of buttons, etc.

[0103] Additionally, the trend analysis unit (134) can generate the most frequently occurring page layout pattern in the one or more popular sales pages through the trend analysis model by inputting the HTML code of one or more popular sales pages into a deep learning-based trend analysis model that has been pre-trained to analyze one or more HTML codes to derive page layout patterns, and through analysis of product image locations, text placement patterns, button locations, section configurations, etc., and can generate trend information including the page layout pattern.

[0104] In addition, the trend analysis unit (134) can transmit the trend information generated for the product registration information to the page generation unit (135).

[0105] Meanwhile, the page generation unit (135) can be linked with the service management unit (131), the product classification unit (133), and the trend analysis unit (134) to extract the user's personal information from the member DB when receiving the product registration request information from the user terminal (10), obtain trend information corresponding to the product registration request information, product registration information generated by the product classification unit (133), and the specific product type (product type information), and generate a personalized sales page for the product to be sold based on the personal information, product registration information, trend information, and specific product type (S6).

[0106] That is, the page generation unit (135) can generate a personalized sales page based on the personal information, product registration information, trend information, and specific product type, which includes product description content generated based on the personal information and product registration information, determines the page layout based on the trend information, and automatically sets sales options corresponding to the specific product type.

[0107] For example, the page generation unit (135) may include a deep learning-based page generation model, and the page generation model may automatically generate an optimized sales page by receiving the personal information, the product registration information, the trend information and the specific product type.

[0108] Specifically, the above page generation model may be composed of a Transformer-based generative AI model and may include an encoder module that processes input information and a decoder module that generates a sales page.

[0109] Accordingly, the page generation unit (135) converts the personal information, the product registration information, the trend information, and the specific product type into a vector form and inputs it into the encoder unit of the page generation model, and the encoder unit can extract features of each input information to generate an integrated feature vector.

[0110] Subsequently, the decoder unit of the above page generation model can sequentially generate HTML code for the sales page based on the integrated feature vector. At this time, the decoder unit determines the HTML structure according to the page layout pattern extracted from the trend information, generates product description content based on the personalization information and product registration information, and can automatically add HTML elements related to sales options corresponding to the product to be sold according to the product registration information and a specific product type.

[0111] In this case, the above sales options refer to detailed product-specific options (detailed attributes of the product, purchase conditions, etc.) that a buyer can select or specify during the product purchase process. For products related to physical goods, they include quantity, size, color, capacity, packaging method, sales unit, delivery method, etc.; for products related to digital content, they include availability period, license type, etc.; for products related to online or offline courses, they include course schedule, course period, number of students, etc.; and for products related to accommodation services, they may include reservation date, number of guests, room type, etc. Additionally, they may include payment method options for selecting a payment method, calendar options for selecting a date, and schedule options for selecting a schedule.

[0112] For example, the page generation unit (135) can generate product description content by processing the post content (image, text) of the user’s SNS account page included in the personal information or product registration information and the product name and product description included in the product registration information through the page generation model.

[0113] At this time, when the specific product type is a food product such as 'processed fruit', sales options such as capacity (30g, 50g, 100g), packaging method (standard packaging, gift packaging), and delivery method (standard courier, early morning delivery) are automatically set, and when the specific product type is an 'offline course', sales options such as a calendar for selecting the course schedule, selection of the number of students, and selection of the course period can be automatically set.

[0114] In addition, the page generation unit (135) can generate a sales page in HTML format so that each element (text, image, button, layout, etc.) has a structure that allows for individual editing, and can store the sales page in the site DB.

[0115] Additionally, as illustrated in FIG. 6, the page generation unit (135) can generate product summary information from the corresponding sales page when the sales page generation is complete, and automatically add the generated product summary information to the landing page corresponding to the user.

[0116] For example, when the page generation unit (135) completes the generation of a sales page for a product to be sold corresponding to ‘strawberry chips’, it can generate card-type product summary information by extracting a representative image (strawberry chip product image), product name ("OO Strawberry Chip"), price ("from 8,000 won"), and brief description ("healthy snack made with 100% domestic organic strawberries") from the sales page.

[0117] In addition, the page generation unit (135) can be configured to set the URL of the corresponding sales page as a link to the product summary information so that when a user selects the product summary information on the landing page, they can move to the corresponding sales page.

[0118] Additionally, the page generation unit (135) can update the landing page so that when adding the product summary information to the product list area of ​​the landing page, it is arranged in a grid form together with the product summary information of other previously registered products.

[0119] For example, if a user has already registered products such as 'tomato' and 'lettuce' and additionally registers a product called 'strawberry chip', the page generation unit (135) can generate product summary information for 'strawberry chip' and update the landing page so that it is displayed in a grid form along with existing products in the product list area of ​​the landing page.

[0120] Additionally, the page generation unit (135) can link the sales page and the payment page. Specifically, when generating the sales page, the URL of the payment page is connected to the "Buy Now" button included in the sales page, and the sales page and the payment page can be linked so that product option information selected by the buyer (product name, selected capacity, quantity, price, etc.) is automatically transmitted to the payment page.

[0121] Through this, when a buyer selects product options on the sales page and clicks the "Buy" button, the payment page is displayed with the selected product information automatically entered, enabling the buyer to proceed with payment.

[0122] Accordingly, as illustrated in FIGS. 8 and 9, the sales page generated by the page generation unit (135) may be configured to include a product purchase area and a product description area.

[0123] At this time, the product purchase area of ​​the sales page illustrated in FIG. 8 may include a product name ("OO Strawberry Chip"), discount price information ("8,000 won", "11,200 won 28%"), seller information ("OO Farm"), delivery information, option selection menu ("Select capacity: 1. 30g, 2. 50g"), a purchase button, etc.

[0124] Additionally, the product description area of ​​the sales page illustrated in FIG. 9 may include various product description content, such as a first product description image related to the sales unit description, a second product description image corresponding to the description of the unit product such as "There are about 40 pieces per bag (based on 30g)," and a third product description image corresponding to the description of the ingredients used in the product manufacturing such as "Made with strawberries larger than a water bottle cap." The sales unit description, the description of the unit product, and the description of the ingredients may each be included as product description content corresponding to text.

[0125] That is, one or more product description contents generated by the page generation unit (135) may each consist of text elements such as product introduction text, product feature description, marketing phrase, or product images within the HTML code of the sales page.

[0126] In addition, the above sales page can be created with a structure that allows the product purchase area and product description area to be viewed by scrolling.

[0127] In addition, the page generation unit (135) generates the sales page in HTML (HyperText Markup Language) format so that each element (text, image, button, layout, etc.) has a structure that allows for individual editing, stores the generated sales page in the site DB, and registers the sales page on the user's product sales site so that buyers can access it.

[0128] At this time, the page generation unit (135) can include the sales page on the product sales site.

[0129] Meanwhile, as illustrated in FIG. 5, the editing interface providing unit (136) can generate an editing interface for editing the sales page and provide it to the user terminal (10) when it receives editing request information regarding an editing request for the sales page from the user terminal (10) (S7) (S8).

[0130] For example, the above-described editing interface providing unit (136) can generate an editing interface for switching the sales page to an editable state when it receives editing request information from the user terminal (10) to request an editing mode switch for the sales page while the user terminal (10) is connected to the sales page.

[0131] At this time, the above-mentioned editing interface providing unit (136) can receive editing request information from the user terminal (10) when the user selects the "Edit" button included in the sales page or selects the "Page Edit" menu from the editing toolbar that is automatically displayed when accessing the sales page while logged in with administrator privileges.

[0132] Additionally, the editing interface providing unit (136) may generate the editing interface in the form of an HTML and JavaScript-based web page and transmit it to the user terminal (10). At this time, the editing interface may be displayed as an overlay on an existing sales page or provided through a separate dedicated editing URL.

[0133] Additionally, the editing interface providing unit (136) may create an editing interface configured to display a page display area (21) for displaying the sales page and a chatbot interface (22) for inputting editing content to be delivered to a chatbot model (138) together on one screen, and may be configured so that a user can input editing content through the chatbot interface (22) while selecting an editing target in the page display area (21).

[0134] Additionally, the above editing interface providing unit (136), upon receiving editing request information from the user terminal (10), obtains the URL or sales page information of the corresponding sales page from the site DB based on the editing request information, generates an editing interface including load command information for loading the sales page into a page display area based on the obtained sales page URL or sales page information, and can make the sales page display in the page display area on the user terminal (10) according to the load command information.

[0135] For example, as illustrated in FIG. 10, the editing interface may be configured as a split-screen structure in which a page display area (21) in which a sales page is displayed on the left or top and a chatbot interface (22) in which a chatbot model (138) can converse is displayed on the right or bottom.

[0136] At this time, the editing interface providing unit (136) can configure the page display area (21) to enable a selection function so that the user can directly select an editing target on the sales page displayed in the page display area (21) by means of a mouse click, touch, drag, etc., and to provide visual feedback such as highlighting, dotted borders, and color changes to the area selected by the user.

[0137] Additionally, the chatbot interface (22) is configured to include a conversation input window, a voice input button, a file upload button, and a message sending button for transmitting information about a selected area or edited content entered in the conversation input window to a service providing device (100), and can be configured so that a user can input edited content in natural language or give commands by voice.

[0138] In addition, the chatbot interface (22) includes a STT (Speech-to-Text) function (or STT module) that converts voice input into text, so that even elderly users who have difficulty using a keyboard can input editing information about the edited content entered by the user into the chatbot interface (22) using only voice input. At this time, the chatbot interface (22) may further include a function such as uploading a product image selected by the user from among the images stored in the user terminal (10) to the service providing device (100).

[0139] Additionally, the above editing interface may be configured to include an agent unit (23), and the agent unit (23) may be embedded in the editing interface in the form of a JavaScript module and executed by the user terminal (10) in the web browser of the user terminal (10).

[0140] Additionally, the agent unit (23) can identify a selection area (or a selection area corresponding to the editing target) containing an editing target selected by the user within a sales page displayed in the page display area (21), and generate selection area information for the selection area and automatically input it into the chatbot interface (22) or transmit it to the editing processing unit (137).

[0141] At this time, the agent unit (23) may perform at least one of the following: a first operation of verifying HTML element information corresponding to the selected area; a second operation of generating coordinate information of the selected area; a third operation of generating a captured image of a pre-set size area including the selected area; and a fourth operation of generating location description information regarding the relative positional relationship with one or more other contents located before, after, or above and below the selected area within the sales page, and may generate the selected area information including at least one of the HTML element information, the coordinate information, the captured image, and the location description information.

[0142] For example, as illustrated in FIG. 10, the agent unit (23) can detect the location where the click or touch occurred when a user clicks or touches a specific location in the page display area (21), and determine the selection area by identifying the content at that location.

[0143] That is, when a user clicks on any location in the sentence “There are about 40 pieces per bag (based on 30g)”, the agent unit (23) detects the clicked coordinates and can identify the text located at the coordinates or the entire sentence containing the text as a selection area.

[0144] At this time, the agent unit (23) can accurately identify the selected area by identifying the ID, class name, tag type, hierarchy, etc. of the selected HTML element through DOM (Document Object Model) analysis. For example, if a user clicks "40", the agent unit (23) can identify the HTML element information of the clicked location ( 40 You can identify the ) and determine the location or range of the corresponding HTML element as the selection area.

[0145] At this time, the above-mentioned selection area can be determined as a point or an area depending on the user's input method.

[0146] In addition, the agent unit (23) may recognize meaningful content units (words, sentences, paragraphs, etc.) around the location as an expanded selection area, regardless of the accuracy of the click location.

[0147] As another example, as illustrated in FIG. 12, the agent unit (23) detects the selected image area when a user selects a specific image of a hand holding a strawberry chip in the product description area of ​​a sales page displayed in the page display area (21) as a selection area, and the HTML element information of the specific image (e.g., By identifying ), the entire specific image mentioned above can be identified as a selected area.

[0148] In addition, the agent unit (23) can detect or generate coordinate information of the selected area.

[0149] For example, if a user selects the area corresponding to "40 pieces" in the sentence "There are about 40 pieces per bag (based on 30g)" as a selection area, the agent unit (23) can extract absolute coordinates such as the X coordinate (250px), Y coordinate (450px), width (50px), and height (20px) of the selected "40 pieces" text, or calculate the relative position of the selected area relative to the entire sales page displayed in the page display area (21) and detect this as coordinate information.

[0150] Additionally, the agent unit (23) can detect coordinate information for the specific image, including location coordinates (e.g., X coordinate 100px, Y coordinate 300px), image size (width 400px, height 300px), image filename ("strawberry_with_hand.jpg"), DOM path, etc., when the specific image is selected as a selection area and identified.

[0151] Alternatively, the agent unit (23) may generate a capture image that captures an area of ​​a preset size including the selected area.

[0152] For example, the agent unit (23) can generate a captured image by utilizing the HTML5 Canvas API to capture an area including the selected "40" text and a surrounding 50-pixel range when text is selected as a selection area, and converting it into Base64 encoded image data.

[0153] Additionally, when the specific image is selected as a selection area, the agent unit (23) can identify the specific image as described above and capture the entire image area of ​​the specific image, or generate an image thumbnail of a preset size based on the specific image and generate it as a captured image.

[0154] Alternatively, the agent unit (23) may generate location description information regarding the relative positional relationship with other content located before, after, or above and below the selected area within the sales page.

[0155] For example, the agent unit (23) can generate location description information in the form of natural language, such as “40 texts located to the right of the text corresponding to ‘capacity selection’ and above the text corresponding to ‘size larger than a water bottle cap’”.

[0156] Alternatively, the agent unit (23) may generate location description information such as "an image located below the text corresponding to 'There are about 40 pieces per bag (based on 30g).'"

[0157] Additionally, the agent unit (23) may generate selection area information based on HTML element information, coordinate information, captured image, location description information, etc., and the selection area information may include at least one of HTML element information, coordinate information, captured image, and location description information.

[0158] At this time, the above-mentioned selection area information may be composed of information structured in JSON form.

[0159] Additionally, the agent unit (23) may automatically input this selection area information into the conversation input window of the chatbot interface (22) or directly transmit it to the editing processing unit (137) via WebSocket.

[0160] Additionally, the user terminal (10) can generate editing content information for the editing target included in the selection area that the user has entered through the chatbot interface (22), and can input and display the editing content information in a conversation input window that displays the user's conversation message included in the chatbot interface (22).

[0161] For example, as illustrated in FIG. 10, the user terminal (10) can generate editing content information for "change 40 to 30" entered by the user through the chatbot interface (22) via voice or text when the user inputs editing content such as "change 40 to 30" for the editing target "40" included in the selection area through the chatbot interface (22) and input it into the conversation input window of the chatbot interface (22).

[0162] At this time, the chatbot interface (22) includes a STT (Speech-to-Text) module that converts voice input into text, and when a user selects a voice input button included in the chatbot interface (22) to activate the voice input function of the user terminal (10) and inputs an editing request such as "Change 40 to 30" by voice, the user terminal (10) converts the user's voice into text through the STT module included in the chatbot interface (22), and then generates editing content information based on the converted text through the chatbot interface (22) and inputs it into the conversation input window included in the chatbot interface (22).

[0163] As another example, as illustrated in FIG. 12, when a user selects the specific image to be edited and selection area information is generated, the user terminal (10) can generate editing content information for the corresponding editing request through the chatbot interface (22) and input it into the conversation input window when the user inputs an editing request content such as "Please delete the hand from the picture" in voice or text through the chatbot interface (22).

[0164] As another example, the agent unit (23) can identify the selected area even when the user selects an area containing multiple elements to change the layout.

[0165] For example, when a user drags and selects an entire section in which product images and product description text are arranged left and right, the agent unit (23) selects the container element of the corresponding section (e.g., It can identify the container and determine the layout structure (flexbox, grid, float, etc.) inside the container to include in the selection area information.

[0166] Specifically, the agent unit (23) can analyze the current CSS properties (display: flex, flex-direction: row, etc.) of the selected container and the arrangement relationship of child elements to generate layout description information such as "left and right arranged image-text sections" and generate selection area information including the same.

[0167] Afterwards, when a user inputs a layout change request such as "move the image to the right," "change the layout to top and bottom," or "reduce the image size" through the chatbot interface (22), the user terminal (10) can generate the request as editing content information and transmit it to the editing processing unit (137).

[0168] In addition, the agent unit (23) can automatically input the selected area information into the conversation input window of the chatbot interface (22) when the selected area information is generated.

[0169] Additionally, after the user inputs the editing content information into the conversation input window, the user terminal (10) can generate editing request information including the editing content information and the selection area information through the chatbot interface (22) and transmit it to the editing processing unit (137) when user input regarding the transmission of the user's editing content information and the selection area information entered into the conversation input window is input through the chatbot interface (22).

[0170] At this time, the user terminal (10) may transmit the selection area information and the editing content information individually to the editing processing unit (137) when the agent unit (23) directly transmits the selection area information.

[0171] In addition, in the configuration described above, the agent unit (23) included in the editing interface can visually distinguish (e.g., by color) editable elements. For example, the agent unit (23) can display a blue border when the user places the mouse pointer over a text element in the page display area (21), a green border when it places it over an image element, and an orange border when it places it over a button element, thereby allowing the user to intuitively recognize the editable area.

[0172] Meanwhile, the above editing processing unit (137) receives editing content information regarding the editing target included in the selection area through the chatbot interface (22) included in the editing interface in conjunction with the above editing interface providing unit (136) (S9), receives selection area information regarding the selection area identified by the agent unit (23) directly from the agent unit (23) or via the chatbot interface (22) (S9), updates the sales page by editing the editing target on the sales page through an artificial intelligence-based chatbot model (138) based on the sales page information regarding the sales page and the selection area information and the editing content information (S10, S11), and provides the updated sales page to the user terminal (10) (S12).

[0173] At this time, the chatbot model (138) may be configured as an artificial intelligence-based model included in any one of the plurality of components included in the service providing device (100), or may be configured as a separate component included in the control unit (130).

[0174] Alternatively, the chatbot model (138) may be included in an external server, and at least one of the plurality of components constituting the service providing device (100) may interact with the chatbot model (138) through an API provided by the external server.

[0175] In addition, the editing processing unit (137) can extract sales page information corresponding to the editing interface provided to the user terminal (10) from the site DB in conjunction with the editing interface providing unit (136).

[0176] Here, the sales page information is data constituting the sales page and may include HTML source code, CSS style information, JavaScript code, path information of image files included in the sales page, image data, metadata, etc. The sales page information refers to data structured in text form and converted into a format that can be input into the chatbot model (138), and may be a concept distinct from the sales page rendered and displayed in the user's web browser.

[0177] Additionally, the sales page information includes structural information and content information of the entire page that are necessary for the chatbot model (138) to accurately identify the editing target and perform editing operations, and the editing processing unit (137) can input the sales page data extracted from the site DB into the chatbot model (138) by configuring it into the sales page information.

[0178] As an example of the operation of the above-described editing processing unit (137), the editing processing unit (137) may receive editing request information from the user terminal (10) including selection area information entered by the agent unit (23) into the chatbot interface (22) and editing content information entered by the user through the chatbot interface (22).

[0179] At this time, the editing processing unit (137) may receive the selection area information and editing content information from the user terminal (10), respectively.

[0180] Additionally, the editing processing unit (137) can apply the received selection area information, editing content information, and sales page information for the sales page to the chatbot model (138) to generate editing command information so that editing is performed according to the editing content information on a specific editing target within the selection area of ​​the sales page through the chatbot model (138), and update the sales page by editing the editing target within the selection area according to the editing command information, or obtain the updated sales page through the chatbot model (138) in which editing is performed according to the editing content information on the editing target by applying the received selection area information, editing content information, and sales page information for the sales page to the chatbot model (138).

[0181] For example, as illustrated in FIGS. 10 and 11, when a user selects "40" from the phrase "There are about 40 pieces per bag (based on 30g)" in the page display area (21) and inputs an edit content "Change 40 to 30" through the chatbot interface (22), the edit processing unit (137) receives selection area information (HTML element information, coordinate information, etc.) for the "40" text from the user terminal (10) and receives an edit content information such as "Change 40 to 30" from the user terminal (10).

[0182] Afterwards, the editing processing unit (137) can input the received selection area information and editing content information and the sales page information into the chatbot model (138) and perform editing processing on the sales page through the chatbot model (138).

[0183] As an example of this, the chatbot model (138) can identify the user's intent to "change 40 to 30" through natural language processing, identify text that is an edit target corresponding to "40" in the selected area on the sales page according to the sales page information based on the selected area information, and generate editing command information to change the identified edit target to text corresponding to "30".

[0184] At this time, the above editing command information may include a command to modify HTML code.

[0185] Accordingly, the editing processing unit (137) [in accordance with] the editing command information output by the chatbot model (138) in the HTML code of the sales page 40 cast 30 By directly modifying it, it can be immediately changed to "There are about 30 pieces per bag (based on 30g)" as shown in Fig. 11.

[0186] At this time, the HTML code of the sales page may be included in the sales page information, and the editing processing unit (137) may update the sales page by modifying the HTML code included in the sales page information.

[0187] Alternatively, the editing processing unit (137) may input sales page information including the entire HTML code of the sales page along with the selection area information and editing content information into the chatbot model (138), and the chatbot model (138) may directly analyze the input HTML code of the sales page to find text ("40") corresponding to the editing target in the selection area, generate a complete HTML code that changes it to other text ("30") requested by the user, and output an updated sales page.

[0188] In addition, the editing processing unit (137) can store the updated sales page obtained directly from the chatbot model (138) as is in the site DB and transmit it to the user terminal (10).

[0189] At this time, the chatbot model (138) can receive input information including the selection area information, editing content information, and sales page information, and output sales page information for the updated sales page, and the editing processing unit (137) can store the updated sales page in the site DB by storing the sales page information output through the chatbot model (138) in the site DB.

[0190] In this way, the present invention supports the editing of the editing target by having the agent unit (23) identify information about the selection area selected by the user to specify the editing target and provide it to the editing processing unit (137), so that even if the sales page contains multiple content or objects identical to the editing target (for example, if text corresponding to "40" is distributed in multiple areas other than the selection area on the sales page), the editing processing unit (137) can accurately identify the editing target specified by the user based on the selection area information.

[0191] As another example, as illustrated in FIGS. 12 and 13, when a user selects an image of a hand holding a strawberry chip in the page display area (21) of an editing interface and inputs an edit content such as "Please delete the hand from the picture" through the chatbot interface (22), the editing processing unit (137) can receive selection area information for the image from the user terminal (10) and receive edit content information such as "Please delete the hand from the picture."

[0192] Accordingly, when the editing processing unit (137) inputs the sales page information, the selection area information, and the editing content information into the chatbot model (138), the chatbot model (138) recognizes a specific image to be edited on the sales page based on the selection area information, identifies that a "hand removal" operation is required based on the editing content information, and then generates and outputs editing command information for image editing of the specific image.

[0193] At this time, the chatbot model (138) can utilize its own multimodal AI function or link with an image generation model pre-set in the editing processing unit (137) to analyze a specific image ("strawberry_with_hand.jpg") to be edited, mask the hand area in the specific image, and then apply an inpainting algorithm to generate a new image ("strawberry_without_hand.jpg") with the hand removed, and generate editing command information including the new image.

[0194] Accordingly, the editing processing unit (137), according to the editing command information generated by the chatbot model (138), in the HTML code of the sales page, an existing specific image ( ) the above new image( It can be replaced with ).

[0195] At this time, the editing processing unit (137) can store a new image in the site DB based on the editing command information and replace the image path of the specific image with the image path of the new image in the HTML code of the sales page, thereby replacing the specific image to be edited on the sales page with the new image.

[0196] Alternatively, the chatbot model (138) may obtain the specific image (or image file of the specific image) from the site DB based on the image path of the specific image included in the sales page according to the sales page information, directly edit the specific image to create a new image with the hand removed, and generate and output a completely updated sales page by automatically updating the image path of the new image in the HTML code of the sales page.

[0197] Accordingly, the editing processing unit (137) can store the new image received from the chatbot model (138) in the site DB by referring to the image path according to the updated sales page, and apply the updated sales page as is so that it can be immediately changed to a new image with the hand removed as shown in FIG. 13.

[0198] Additionally, the chatbot model (138) may receive input information including the selection area information, editing content information, and sales page information, and output editing command information or output sales page information for the updated sales page, and may output response message information to be transmitted to the chatbot interface (22) of the user terminal (10), and the editing processing unit (137) may transmit the response message information to the user terminal (10) in conjunction with the editing interface providing unit (136).

[0199] Accordingly, as illustrated in FIGS. 10 and 12, the user terminal (10) can display response message information such as "I will change the text as requested" or "I will change the picture with the hand removed" through the chatbot interface (22) included in the editing interface upon receiving the response message information.

[0200] In addition, if the above-mentioned editing target is a layout, the above-mentioned editing processing unit (137) can change the layout of the sales page, such as the placement, size, and alignment of page components.

[0201] For example, if a user selects a section where product images and text are arranged horizontally and inputs an edit request saying "change to vertical arrangement," the editing processing unit (137) can receive selection area information (HTML element information, current layout attributes, etc.) for a container element and an edit request information saying "change to vertical arrangement" from the user terminal (10).

[0202] Additionally, when the editing processing unit (137) inputs the selection area information, editing content information, and sales page information into the chatbot model (138), the chatbot model (138) can analyze the current layout structure and generate editing command information including a CSS modification command to implement the requested layout change.

[0203] In addition, the above editing processing unit (137) may, according to the above editing command information, a style attribute within the HTML code of the sales page or <style> 태그 내의 CSS 규칙을 수정하여, 사용자가 요청한 레이아웃 변경을 즉시 반영할 수 있다.

[0204] 또는, 상기 챗봇 모델(138)은 레이아웃이 변경된 완전한 HTML 코드를 포함하는 갱신된 판매 페이지를 직접 생성하여 출력할 수도 있으며, 이 경우 상기 편집 처리부(137)는 갱신된 판매 페이지를 그대로 적용할 수 있다.

[0205] 또한, 상기 편집 처리부(137)는, 상기 편집 인터페이스 제공부(136)와 연동하여, 편집 작업이 완료될 때마다 갱신된 판매 페이지를 상기 사이트 DB에 저장하고, 갱신된 판매 페이지의 HTML 코드를 상기 사용자 단말(10)에 전송하여 편집 인터페이스의 페이지 표시 영역(21)에 즉시 반영되도록 함으로써, 지속적인 수정 작업이 가능하도록 지원할 수 있다.

[0206] 또한, 상기 편집 처리부(137)는, 상기 편집 인터페이스에 포함된 편집 완료 버튼의 선택에 따른 편집 완료 정보를 상기 사용자 단말(10)로부터 수신하면, 편집 세션을 종료하고, 갱신된 판매 페이지의 최종 버전을 확정하여 상기 사이트 DB에 저장하며, 해당 판매 페이지를 공개 상태로 전환하여 구매자들이 즉시 접근 가능하도록 할 수 있다.

[0207] 또한, 상기 편집 처리부(137)는, 편집 이력을 관리하는 기능을 포함할 수 있다.

[0208] 일례로, 상기 편집 처리부(137)는, 모든 편집 작업 각각에 대해 타임스탬프, 편집 전 내용, 편집 후 내용, 편집자 정보를 포함하는 편집 이력 정보를 상기 사이트 DB에 저장할 수 있으며, 사용자 단말(10)로부터 판매 페이지를 이전 상태로 복원하기 위한 복원 요청 정보를 수신하면, 해당 복원 요청 정보에 대응하는 편집 이력 정보를 기초로 판매 페이지를 이전 버전(이전 상태)으로 즉시 복원할 수 있다.

[0209] 또한, 상기 편집 인터페이스 제공부(136)는, 상기 사용자 단말(10)로부터 상기 랜딩 페이지에 대한 편집 요청 정보를 수신할 수도 있다.

[0210] 일례로, 상기 편집 인터페이스 제공부(136)는, 사용자가 상기 랜딩 페이지에 접속한 상태에서 랜딩 페이지 편집 버튼을 선택하거나 관리자 메뉴에서 "랜딩 페이지 편집"을 선택하여 랜딩 페이지에 대한 편집 요청 정보를 전송하면, 상기 판매 페이지와 동일한 방식으로 상기 랜딩 페이지를 편집할 수 있는 편집 인터페이스를 생성하여 제공할 수 있다.

[0211] 이때, 상기 편집 인터페이스 제공부(136)는, 상기 랜딩 페이지를 페이지 표시 영역에 표시하고, 챗봇 인터페이스와 에이전트부를 포함하는 편집 인터페이스를 생성하여, 사용자가 상기 랜딩 페이지의 판매자 소개 영역, 상품 요약 정보 배열, 배경 이미지 등을 선택하고 상기 챗봇 인터페이스를 통해 편집할 수 있도록 지원할 수 있다.

[0212] 또한, 상기 편집 처리부(137)는, 상기 랜딩 페이지에 대한 편집 내용 정보와 선택 영역 정보를 수신하면, 상기 판매 페이지를 편집하여 갱신한 방식과 동일한 방식으로 상기 챗봇 모델(138)을 통해 상기 랜딩 페이지를 편집하여 갱신하고, 갱신된 랜딩 페이지를 상기 사이트 DB에 저장하여 상기 사용자 단말(10)에 제공할 수 있다.

[0213] 한편, 상술한 실시예에서 설명한 본 발명의 챗봇 모델(138)은 단일 모델로 구성될 수도 있으나, 복수의 전문화된 인공지능 모델들이 유기적으로 결합된 하이브리드 모델로 구현될 수도 있다.

[0214] 일례로, 상기 챗봇 모델(138)은, GPT(Generative Pre-trained Transformer) 계열, Gemini, LLaMA(Large Language Model Meta AI) 등과 같은 대규모 언어 모델(Large Language Model, LLM), 멀티모달(Multimodal) AI 모델 또는 이들의 조합 등이 사용될 수 있다.

[0215] 이러한 LLM은 사용자가 입력한 자연어 형태의 편집 내용 정보를 이해하고, 에이전트부(23)로부터 전달받은 선택 영역 정보의 문맥을 파악하여 사용자의 복합적인 편집 의도를 해석하는 데 탁월한 성능을 보인다.

[0216] 또한, 편집 대상이 이미지인 경우, 상기 챗봇 모델(138)은 사용자의 자연어 명령("그림에서 손을 삭제해줘")을 해석하여 Stable Diffusion, Midjourney, DALL-E와 같은 이미지 생성 모델(Image Generation Model)에 전달할 프롬프트(Prompt)를 생성하고, 해당 이미지 생성 모델을 호출하여 신규 이미지를 생성하도록 구성될 수 있다. 이 경우, 챗봇 모델(138)은 사용자의 언어적 명령을 시각적 결과물로 변환하는 지능적인 오케스트레이터(Orchestrator) 역할을 수행할 수 있다.

[0217] 나아가, 상기 챗봇 모델(138)은 검색 증강 생성(Retrieval-Augmented Generation, RAG) 기술을 적용할 수도 있다. 이는 미리 저장된 디자인 템플릿, 성공적인 마케팅 문구, 상품 카테고리별 정보 등을 포함하는 내부 데이터베이스로부터 가장 적절한 정보를 실시간으로 검색하고, 이를 기반으로 사용자에게 편집 스타일을 제안하거나 판매 페이지의 콘텐츠를 더욱 풍부하게 생성하는 데 활용될 수 있다.

[0218] 또한, 본 발명에서 설명한 인공지능 또는 딥러닝 기반의 모델이나 모듈은, 딥러닝 알고리즘 또는 하나 이상의 신경망 모델로 구성될 수 있으며, 해당 신경망 모델(또는 신경망)은 입력층(Input Layer), 하나 이상의 은닉층(Hidden Layers) 및 출력층(Output Layer)으로 구성될 수 있다. 또한, 상기 신경망 모델에는 DNN(Deep Neural Network), RNN(Recurrent Neural Network), LSTM(Long Short-Term Memory), GRU(Gated Recurrent Unit), CNN(Convolutional Neural Network), Transformer, Attention 메커니즘, 그래프 신경망(Graph Neural Network), LLM 등과 같은 다양한 종류의 신경망이 적용될 수 있다.

[0220] 본 발명의 실시예들에서 설명된 구성요소는, 예를 들어, 메모리, 프로세서, 콘트롤러, ALU(arithmetic logic unit), 디지털 신호 프로세서(digital signal processor), 마이크로컴퓨터, FPGA(Field Programmable Gate Array), PLU(programmable logic unit), 마이크로프로세서, 논리게이트(Logic gate) 등의 하드웨어, 명령어 세트를 포함하는 소프트웨어 내지 이들의 조합 또는 명령(instruction)을 실행하고 응답할 수 있는 다른 어떠한 장치와 같이, 하나 이상의 범용 컴퓨터 또는 특수 목적 컴퓨터를 이용하여 구현될 수 있다.

[0221] 본 명세서에 기술된 다양한 장치 및 구성부는 하드웨어 회로(예를 들어, CMOS 기반 로직 회로), 펌웨어, 소프트웨어 또는 이들의 조합에 의해 구현될 수 있다. 예를 들어, 다양한 전기적 구조의 형태로 트랜지스터, 로직게이트 및 전자회로를 활용하여 구현될 수 있다.

[0222] 전술된 내용은 본 발명이 속하는 기술 분야에서 통상의 지식을 가진 자라면 본 발명의 본질적인 특성에서 벗어나지 않는 범위에서 수정 및 변형이 가능할 것이다. 따라서, 본 발명에 개시된 실시예들은 본 발명의 기술 사상을 한정하기 위한 것이 아니라 설명하기 위한 것이고, 이러한 실시예에 의하여 본 발명의 기술 사상의 범위가 한정되는 것은 아니다. 본 발명의 보호 범위는 아래의 청구범위에 의하여 해석되어야 하며, 그와 동등한 범위 내에 있는 모든 기술 사상은 본 발명의 권리범위에 포함되는 것으로 해석되어야 할 것이다.부호의 설명

[0223] 10: 사용자 단말 100: 서비스 제공 장치110: 통신부 120: 저장부130: 제어부 131: 서비스 관리부132: 사용자 분석부 133: 상품 분류부134: 트렌드 분석부 135: 페이지 생성부136: 편집 인터페이스 제공부137: 편집 처리부 138: 챗봇 모델21: 페이지 표시 영역 22: 챗봇 인터페이스23: 에이전트부< / style>

Claims

Claim 1 A service providing device for supporting the creation of a personalized web page for product sales, comprising: a user analysis unit that generates personalized information by analyzing a user's SNS account page; a product classification unit that automatically classifies the type of the product to be sold into one of a plurality of pre-set product types to create a sales page, which is a web page for selling the product to be sold, based on product registration information of the product to be sold that the user intends to sell; a trend analysis unit that generates trend information by analyzing one or more other sales pages in which product types belonging to the same category as the specific product type classified by the product classification unit are set; and a page generation unit that generates a personalized sales page based on the personal information, product registration information, trend information, and the specific product type, wherein the page layout is determined based on the trend information, and sales options corresponding to the specific product type are automatically set, and the page includes product description content generated based on the personalized information and product registration information. Claim 2 A service providing device for supporting the creation of a personalized web page for product sales, wherein, in claim 1, the user analysis unit extracts at least one of a user profile, post content, interest tags, and follower information from the SNS account page, and generates personalized information including at least one of the user's field of interest, target customer base, and post content based on the extracted information. Claim 3 A service providing device for supporting the creation of a personalized web page for product sales, wherein, in claim 1, the trend analysis unit analyzes a page layout pattern for one or more popular sales pages satisfying a preset selection condition among one or more other sales pages in which a product type belonging to the same category as the specific product type is set, and generates trend information regarding the page layout pattern. Claim 4 A service providing device for supporting the creation of a personalized web page for product sales according to claim 1, wherein the page generating unit adds at least one of quantity, size, color, capacity, packaging method, sales unit, delivery method, payment method, availability period, license type, course schedule, course period, number of students, reservation date, number of guests, room type, schedule, and calendar to the sales options according to the specific product type. Claim 5 A service providing device for supporting the creation of a personalized web page for product sales according to claim 1, wherein the page generating unit generates a landing page to introduce a user's product for sale based on the personalized information, and when the sales page is generated, generates product summary information based on the sales page and automatically adds the product summary information, which has link information connected to the sales page set, to the landing page. Claim 6 A service providing device for supporting the creation of a personalized web page for product sales according to claim 1, further comprising: an editing interface providing unit that, upon receiving editing request information for an editing request for the sales page, creates and provides an editing interface including the sales page and a chatbot interface, and an agent unit that identifies a selection area selected by a user on the sales page; and an editing processing unit that receives editing content information regarding editing content of an editing target included in the selection area through a chatbot interface included in the editing interface, receives selection area information regarding a selection area identified by the agent unit, and updates the sales page by editing the editing target on the sales page through an artificial intelligence-based chatbot model based on sales page information regarding the sales page, the selection area information, and the editing content information. Claim 7 A service providing device for supporting the creation of a personalized web page for product sales, characterized in that, in claim 6, the editing interface is configured to display a page display area for displaying the sales page and the chatbot interface together on a single screen so that a user can input editing content through the chatbot interface while selecting an editing target in the page display area, and the agent unit is configured to identify a selection area containing the editing target selected by the user through the editing interface, generate selection area information for the selection area, and automatically input it into the chatbot interface or transmit it to the editing processing unit. Claim 8 A service providing device for supporting the creation of a personalized web page for product sales according to claim 6, wherein the editing processing unit receives selection area information generated by the agent unit and editing content information entered by a user through the chatbot interface, applies the received selection area information, editing content information, and sales page information to the chatbot model to generate editing command information so that editing according to the editing content information is performed on a specific editing target within the selection area of ​​the sales page through the chatbot model, and updates the sales page by editing the editing target within the selection area according to the editing command information, or obtains the updated sales page through the chatbot model in which editing according to the editing content information is performed on the editing target by applying the received selection area information, editing content information, and sales page information to the chatbot model. Claim 9 A service providing device for supporting the creation of a personalized web page for product sales, characterized in that, in claim 6, the chatbot interface includes a voice recognition function that converts voice input into text, thereby enabling a user to input the editing content information by voice. Claim 10 A method of providing a service by a service providing device, comprising: a step of generating personalized information by analyzing a user’s SNS account page; a step of automatically classifying the type of the product to be sold into one of a plurality of pre-set product types to create a sales page, which is a web page for selling the product to be sold, based on product registration information of the product to be sold that the user intends to sell; a step of generating trend information by analyzing one or more other sales pages in which product types belonging to the same category as the specific product type classified for the product to be sold are set; and a step of generating a personalized sales page based on the personalized information, product registration information, trend information, and the specific product type, wherein the personalized sales page includes product description content generated based on the personalized information and product registration information, the page layout is determined based on the trend information, and sales options corresponding to the specific product type are automatically set.

Images