Website creation support device

Abstract

To provide a website creation support device for creating a high-quality website in a short time by proposing a combination of parts optimized for each individual user. [Solution] The website creation support device 1 has a web part management database 1212 in which web part information which is the elements that make up a website is stored, a profile database 1211 in which profile information of a user 4 creating a website is stored, a communication unit 13 which enables communication of site-related information regarding the website between the user terminal 2 operated by the user 4, and a website creation support unit 11 which predicts the website desired by the user 4 based on the profile information and the site-related information, extracts selection candidates for web part information suitable for creating the website from the web part management database 1212, and presents this to the user 4 as web creation support information to support the creation of the website.

Description

[Technical field]

[0001] The present invention relates to a website creation support device. [Background technology]

[0002] Patent document 1 discloses a technique for a homepage design creation method that includes an answer receiving step of receiving answers to questions about products handled by the homepage to be created, a combination extraction step of extracting, based on the answers, a combination of homepage parts corresponding to each answer pattern from a storage means in which combinations of homepage parts are registered, and a design creation step of extracting parts from the storage means in which homepage parts are registered based on the combinations, and combining the extracted parts to create a homepage design.

[0003] According to the technology of Patent Document 1, a combination of pre-prepared parts is selected in response to answers to questions about the attributes of the products to be handled on the homepage that the user is trying to create, and the homepage design is constructed based on that combination, making it possible to create a homepage cheaply and easily. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2002-351783 A Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, the technology of Patent Document 1 makes it possible to select a combination of parts prepared in advance according to the answers to questions and to build a homepage design based on that combination, but since it is limited to the combination of parts registered for each answer pattern of questions about products, the combination of parts may not be optimal for creating the homepage desired by the user.If a combination of parts optimized for the user is not proposed, it becomes difficult to create a high-quality homepage in a short time, and improvement is desired.

[0006] The present invention has been made in consideration of the above circumstances, and its object is to provide a website creation support device that enables the creation of a high-quality website in a short amount of time by proposing combinations of parts that are optimized for each individual user. [Means for solving the problem]

[0007] As a result of intensive research into solving the above problems, the inventors have found that the above object can be achieved by predicting a website desired by a user based on profile information and site-related information and presenting the user with appropriate web part information selection candidates. The inventors have then completed the present invention. Specifically, the present invention provides the following.

[0008] The present invention includes a web part management database in which information on web parts, which are elements that configure a website, is stored; a profile database in which profile information of a user who creates the website is stored; a communication unit that enables communication of site-related information regarding the website between the user terminal operated by the user and the communication unit; The website creation support device has a website creation support unit that predicts the website desired by the user based on the profile information and the site-related information, extracts selection candidates for web part information suitable for creating the website from the web part management database, and presents them to the user as web creation support information to support the creation of the website.

[0009] According to the above configuration, the website desired by the user is predicted based on the profile information and site-related information, and appropriate web part information selection candidates are extracted from the web part management database and presented to the user. This reduces the burden on the user in selecting designs and functions, and suggestions optimized for each individual user are made, making it possible to create a high-quality website in a short time. Effect of the Invention

[0010] According to the present invention, a combination of parts optimized for each individual user is proposed, making it possible to create a high-quality website in a short period of time. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 is an explanatory diagram showing the flow of information in a website creation support device according to this embodiment. [Diagram 2] FIG. 2 is an explanatory diagram showing the contents of a data table of the web part management database of this embodiment. [Diagram 3] FIG. 3 is an explanatory diagram showing the contents of a data table of the profile database of this embodiment. [Figure 4] FIG. 4 is an explanatory diagram showing the flow of information in the website creation support device of this embodiment. [Diagram 5] FIG. 5 is an explanatory diagram showing the flow of information in the website creation support device of this embodiment. [Figure 6] FIG. 6 is an explanatory diagram showing the flow of information in the website creation support device of this embodiment. [Figure 7] FIG. 7 is an explanatory diagram showing the contents of the data table of the dropout rate database of this embodiment. [Figure 8] FIG. 8 is an explanatory diagram showing the flow of information in the website creation support device of this embodiment. [Figure 9] FIG. 9 is a flowchart of the website creation support program of this embodiment. [Figure 10] FIG. 10 is an explanatory diagram showing the process of creating a website in this embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] First, although the following disclosure, diagrams, and / or claims are described as being described alone or in combination with one or more other aspects, the subject matter of the immediate disclosure is not intended to be so limited. That is, the immediate disclosure, diagrams, and claims are intended to encompass the various aspects described herein, each alone or in one or more combinations with each other. For example, even if the immediate disclosure describes and illustrates a first embodiment, a second embodiment, and a third embodiment in such a manner that the first embodiment is described and illustrated in particular with respect to the second embodiment, or the second embodiment is described and illustrated only with respect to the third embodiment, the immediate disclosure and illustrations are not so limited and may include only the first embodiment, only the second embodiment, only the third embodiment, or one or more combinations of the first, second, and / or third embodiments, such as the first and second embodiments, the first and third embodiments, the second and third embodiments, or the first, second, and third embodiments.

[0013] The use of the phrase "or" in this document shall mean a "non-exclusive" arrangement unless expressly specified otherwise. For example, "item x is A or B" shall mean either: (1) item x is either A or B, but not both; or (2) item x is both A and B. In other words, the word "or" is not used to define an "exclusive" arrangement.

[0014] Additionally, as used herein, the phrases "comprising at least one of" and "comprising at least one of the following," when used in conjunction with a system or element, mean that the system or element includes one or more of the elements listed after the phrase. For example, if there are three types of elements, a first element, a second element, and a third element, the phrases "comprising at least one of" and "comprising at least one of the following" are to be interpreted as any of the following structural arrangements: a device including the first element, a device including the second element, a device including the third element, a device including the first and second elements, a device including the first and third elements, a device including the second and third elements, or a device including the first, second, and third elements.

[0015] A similar interpretation is intended when the phrase "used in at least one of the following" is used in this context. Additionally, "and / or" is used in this context as a verbal conjunction to indicate that one or more of the listed elements or conditions are included or occur. For example, a device that includes a first element, a second element, and / or a third element is to be interpreted as any of the following structural arrangements: a device that includes the first element, a device that includes the second element, a device that includes the third element, a device that includes the first element and the second element, a device that includes the first element and the third element, a device that includes the second element and the third element, or a device that includes the first element, the second element, and the third element.

[0016] In addition, the use of the phrase "and / or" in the text means a "non-exclusive" agreement, as stipulated in the Japanese Industrial Standards (JIS) "Format and preparation method of standard sheets JIS Z 8301."

[0017] Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings.

[0018] (Website Creation Support Device 1: Overview) As shown in FIG. 1, website creation support device 1 is configured to predict the website desired by user 4 based on profile information of user 4 creating a website and site-related information regarding the website transmitted and received between user terminal 2 operated by user 4, and to support the creation of a website by presenting user 4 with a selection of web part information suitable for website creation as web creation support information.

[0019] Here, a "website" refers to a set of web pages published on the Internet 5. These web pages are created using web technologies such as HTML, CSS, and JavaScript, and are accessible to users 4 through a browser. Websites are built for a variety of purposes, such as providing information, providing services, entertainment, and electronic commerce. A website is usually made up of multiple pages, each of which is connected to each other through links. This allows users 4 to easily find information through navigation. The content of a web page may be provided in various formats, such as text, images, video, and audio. A website is hosted on a computer system called a web server 3 and published on the Internet 5 through a domain name. Users 4 can reach the website pages by entering the domain name in the address bar of their browser or by accessing through a search engine. In addition, a website can improve its ranking on the pages that appear when a search is performed by entering a specific keyword or phrase in a search engine through SEO (search engine optimization). This increases traffic to the website and promotes the provision of targeted information and the use of services.

[0020] "Web part information" refers to elements that make up a website, that is, information about various elements that make up a website. This information includes detailed data about basic components that realize the design and functions of a website. Specifically, the web part information includes a wide range of elements, such as text, images, buttons, navigation menus, forms, videos, icons, and banners. Each web part is assigned metadata such as a part ID, part type, category, description, creation date, and update date, making it easy to identify and manage the parts. The web part information plays a role in improving the design and functionality of a website. For example, text parts include information such as font style, size, and color, which allows for consistency in the design. Image parts include information such as image size, resolution, and placement position, which allows adjustments to be made to enhance visual appeal. Interactive parts such as buttons and navigation menus include settings to improve the usability of the user interface. Furthermore, the web part information can be customized according to the needs of the user 4. This makes it possible to build a website that meets a specific purpose or theme. Based on the web part information, the user 4 can freely combine each web part, adjust the layout, and create his or her own website.

[0021] Web part information is information for improving the functionality of a website. For example, form parts incorporate interactive functions such as validating user input, setting the data transmission destination, and displaying user-friendly error messages. This makes it possible to provide an environment in which visitors to a website can comfortably use the website. In addition, video parts include information for effectively displaying media content, such as video playback settings, thumbnail designation, and adding captions. This makes it possible to easily integrate visually appealing content and attract the interest of visitors. In addition, web part information also contributes to SEO (search engine optimization). By setting appropriate metadata for each part, search engines can accurately understand the content of a website and can obtain a high ranking in search results. For example, by setting appropriate alternative text for an image part, it is possible to increase exposure in image searches. In addition, web part information includes display settings for different devices (desktop, tablet, smartphone), making it possible to provide an optimal layout according to the device. It is preferable that the web part information is integrated with a version control system, so that the update history of the part can be tracked and the part can be restored to a previous version if necessary.

[0022] "User 4" refers to an individual or a corporation that uses website creation support device 1 to build and manage a website. User 4 has various profile information according to the purpose of creating a website, and can receive support for creating a website customized based on that information.

[0023] Specifically, the user 4 first provides his / her profile information by logging in to the website creation support device 1. This includes his / her name, occupation, company name, and areas of interest. For example, if the user 4 is a lawyer, he / she is required to input detailed information such as that he / she is interested in business law and contracts. This profile information is recorded in the profile database 1211 and used by the website creation support unit 11. When the user 4 starts building a website, the website creation support unit 11 extracts optimal web part information from the web part management database 1212 based on the profile information and site-related information of the user 4. This allows the user 4 to be presented with a selection of web parts that are most suitable for his / her needs. For example, if the user 4 is a lawyer who provides legal services, a catchy slogan and profile parts that emphasize reliability and expertise are proposed.

[0024] Furthermore, the user 4 can easily operate the website through a user interface (UI) for adding specific designs and contents to each page of the website. The website can be constructed intuitively using drag-and-drop operations and text generation input forms. This makes it possible to create a high-quality website without knowledge of coding. When the user 4 publishes the website, the site can be deployed quickly and efficiently on the Internet 5 by utilizing the deployment and update functions. Furthermore, if the website needs to be updated or maintained even after publication, the user 4 can access the web part management database 1212 to add new parts or edit existing parts.

[0025] "Profile Information" refers to detailed information about an individual or legal entity that User 4 provides to build and manage a website. Profile Information includes User 4's name and occupation as basic identifying information. For example, if User 4 is a lawyer named "Yamada Taro", his / her name and occupation are recorded as part of the profile information. It also includes the name of the company or organization to which User 4 belongs, which provides a clear context for User 4's business or services. For example, if User 4 belongs to "Yamada Law Office", this information is recorded.

[0026] Furthermore, the profile information includes the user 4's areas of interest. This allows the website creation support device 1 to propose web parts and design templates that meet the specific needs and preferences of the user 4. For example, for a user 4 who is interested in business law and contracts, related legal content and contract templates are preferentially proposed. The profile information is recorded in the profile database 1211 of the website creation support device 1. The profile database 1211 is designed to centrally manage detailed information for each user 4 and to allow quick access as needed. This allows the user 4 to be provided with optimal support based on the individual profile information when building a website. For example, the profile information plays an important role in the selection and customization of web parts. Based on the user 4's occupation and areas of interest, appropriate catchphrases, service descriptions, profile parts, etc. are proposed. In addition, parts for emphasizing the user 4's expertise and experience are also provided. This allows the user 4 to build a website that effectively appeals to his or her strengths and characteristics.

[0027] "Site-related information" refers to various data necessary for creating and managing a website. This information includes detailed data on the components, design, and functions of a website. Specifically, site-related information includes information on the basic structure of a website. For example, detailed design information for each page of the website, such as the layout of each web page, the link structure between pages, the arrangement of navigation menus, service introduction pages, and inquiry pages.

[0028] As a concrete example, in the case of the Yamada Law Office website, this includes information such as the site name, site purpose, theme, color scheme, fonts, main menu items, homepage contents, service details page, lawyer introduction page, customer testimonials page, contact page, blog, FAQ, SNS integration, etc. This makes the entire site structure consistent and provides visitors with easy-to-understand navigation.

[0029] Site-related information also includes data about content. Specifically, it includes all content elements displayed on a web page, such as text, images, videos, buttons, and forms. Each content element includes detailed information such as its specific content, position, and style settings. For example, the image slider displayed on the top page, catchy copy, service overview, and customer testimonials are saved as site-related information.

[0030] In addition, site-related information also includes information about the design of the website. This includes color settings, font styles, layout patterns, theme settings, etc. This allows for consistency in the design of the entire website and strengthens the brand image of users 4. For example, Yamada Law Office uses a color scheme of navy blue, white, and gray, and sans-serif fonts (e.g., Arial, Helvetica) to achieve a professional, reliable, and simple design.

[0031] Site-related information also includes data related to SEO (search engine optimization). For example, meta tags, keywords, page titles, descriptions, etc. are stored to effectively communicate the content of the website to search engines. This can improve the website's search ranking and attract more visitors.

[0032] The site-related information is also intended to be used for the operation and maintenance of the website. For example, it includes information for collecting access analysis data and visitor behavior data and evaluating the performance of the website. This makes it possible to analyze visitor behavior patterns and identify improvements to the site. The site-related information is stored in the site-related database 1213. The site-related database 1213 is designed to centrally manage all information required for creating and managing the website and to enable the user 4 to quickly access the information he or she needs. This allows the website to be built and operated efficiently and effectively.

[0033] (Website Creation Support Device 1: Memory Unit 12) The website creation support device 1 has a storage unit 12. The storage unit 12 has a function for storing various data constituting a website and enabling rapid access when necessary. The main function of the storage unit 12 is to safely and efficiently store information including a web parts management database 1212, a profile database 1211, and various other databases, and to provide such data so that it can be accessed by the website creation support unit 11, etc. This makes it possible to provide the information required by the user 4 quickly and accurately.

[0034] More specifically, the memory unit 12 has a storage device for storing data. The storage device may be a solid-state drive (SSD) or a hard disk drive (HDD). These devices have the ability to quickly read and write large amounts of data, making it possible to efficiently manage large amounts of data such as web part information and profile information.

[0035] The storage unit 12 also has a RAM (random access memory) for temporarily storing data that requires fast access. The RAM temporarily stores data that is frequently accessed by the processor, improving the speed of data processing. The storage unit 12 has a database management system (DBMS). The database management system (DBMS) reads and writes data using SQL (Structured Query Language), and plays a role in maintaining the integrity and consistency of the data. This ensures that information stored in databases such as the web part management database 1212 and the profile database 1211 is kept accurate and up to date.

[0036] Furthermore, the storage unit 12 is provided with a backup system that periodically performs backups in preparation for data loss or damage. The backup system saves a copy of the data stored in the storage device in another location, enabling data recovery. This ensures data integrity even in the unlikely event of a failure. It is preferable that the storage unit 12 is provided with a security function to prevent unauthorized access to or leakage of data. Specifically, this includes data encryption, access control, and audit log recording. This protects confidential information and important data of the user 4, and realizes highly reliable data management.

[0037] (Website creation support device 1: memory unit 12: web parts management database 1212) The storage unit 12 configured as above has a web part management database 1212 in which information on web parts that are elements constituting a website is stored. The web part management database 1212 is a database for centrally managing information on various web parts that constitute a website. This database 1212 stores detailed information and metadata on the web parts. The web part management database 1212 includes a "customer information table," a "web part information table," and a "part usage status table." The customer information table records information on the user 4 who creates the website, including the customer ID, name, email address, account creation date, and part of profile information. This information enables customization and personalization for each user 4.

[0038] The web part information table stores detailed information about each web part. Specifically, metadata such as part ID, part type, content, version, creator, creation date, and tags are included. This clarifies the characteristics and use of each web part, and allows the appropriate web part to be selected quickly as needed. The part usage status table is for tracking how each web part was used by which user 4, and includes information such as customer ID, part ID, usage location ID, number of uses, and last use date, making it possible to evaluate the frequency of use and performance of the web part. By combining these tables, the web part management database 1212 aims to improve the efficiency and accuracy of the website creation process, as well as increase the reusability of web parts and support the creation of a lean website. In addition, it also makes it easy to suggest web parts based on the user 4's profile and to manage information required for updating and maintaining the site.

[0039] To explain the web parts management database 1212 in detail using a specific example, as shown in Fig. 2, the customer information table records information about the user 4 who creates a website. For example, Yamada Taro (customer ID 1) is a lawyer and belongs to Yamada Law Office. Tanaka Hanako (customer ID 2) is a designer and belongs to Tanaka Design Studio. The interests and concerns of each are also recorded, making it possible to customize and personalize based on the preferences and needs of user 4.

[0040] The Web Part Information Table stores detailed information about each Web Part. For example, Part ID WP001 is a catchphrase, "Yamada Law Office, trusted and proven," created by Yamada Taro. Part ID WP002 is a service description, "Legal consultation: We will solve your legal problems," also created by Yamada Taro. This information clarifies the characteristics and uses of each Web Part, allowing users to quickly select an appropriate Web Part as needed.

[0041] The part usage status table is used to track how each web part was used and by which user 4. For example, part ID WP001 used by Yamada Taro of customer ID 1 has been used three times on PAGE001, and the last date of use was January 15, 2023. Also, part ID WP004 used by Tanaka Hanako of customer ID 2 has been used once on PAGE004, and the last date of use was January 22, 2023. This makes it possible to evaluate the frequency of use and performance of parts.

[0042] (Website creation support device 1: memory unit 12: profile database 1211) As shown in FIG. 1, the storage unit 12 has a profile database 1211 in which profile information of a user 4 who creates a website is stored. The profile database 1211 is a database for centrally managing detailed profile information on the user 4 who creates a website. The profile database 1211 stores various data necessary for website creation, such as personal information, occupational information, and areas of interest of the user 4, and enables the provision of website creation support customized for each user 4. The profile database 1211 contains basic identification information of the user 4. Specifically, the name, occupation, company name, contact information, and the like of the user 4 are recorded. For example, for a lawyer named Yamada Taro, the name of Yamada Taro, the occupation of lawyer, information on the Yamada Law Office to which the user belongs, and the like are stored in the profile database 1211.

[0043] Furthermore, the profile database 1211 also includes information on the user 4's fields of interest. This includes the user 4's fields of expertise, hobbies, topics of interest, and the like. For example, the profile of a user 4 who is interested in business law and contracts includes information on business law and contracts. This information plays an important role when the website creation support unit 11 proposes optimal web parts and design templates to the user 4. The profile database 1211 functions in cooperation with each unit of the website creation support device 1. For example, the profile information received from the terminal of the user 4 through the communication unit 13 is recorded in the profile database 1211. The information in the profile database 1211 is also used when extracting appropriate web part information from the web part management database 1212. This enables customized support for each user 4, and realizes website creation according to individual needs.

[0044] To explain using a specific example, as shown in FIG. 3, the profile database 1211 includes customer ID, name, age, sex, occupation, company name, service content, interests, and past usage history. For example, Yamada Taro, customer ID 1, is a 35-year-old man who is a lawyer and belongs to Yamada Law Office. His service content is legal consultation, and he is interested in business law and contracts. He has also used the website creation support device three times in the past. Tanaka Hanako, customer ID 2, is a 28-year-old woman who is a designer and belongs to Tanaka Design Studio, where she provides graphic design. She is interested in UI / UX design and branding, and has used the support device once in the past.

[0045] The profile database 1211 includes information on the user 4's areas of interest in addition to basic identification information. In other words, the user 4's areas of expertise, hobbies, topics of interest, etc. are recorded. For example, Suzuki Ichiro (customer ID 3) is a 45-year-old man who works as an engineer at Suzuki Engineering and is in charge of system development. He is interested in AI and IoT, and has used the website creation support device 1 five times in the past.

[0046] (Website creation support device 1: User terminal 2) As shown in FIG. 1, a website creation support device 1 is capable of data communication with a user terminal 2. The user terminal 2 is a device such as a computer, smartphone, or tablet that is used by a user 4 to create and manage a website. The user terminal 2 functions as an interface for communicating with the website creation support device 1, and provides an environment in which the user 4 can intuitively operate the design and content of the website. The user terminal 2 establishes a connection with the website creation support device 1. This connection is made via the Internet 5, and can be accessed through a web browser or a dedicated application on the user terminal 2. This allows the user 4 to use various functions of the website creation support device 1.

[0047] The user terminal 2 has a user interface (UI) function. The UI is a screen for the user 4 to operate the design and contents of a website, and is designed to be intuitively operated through drag-and-drop operations and input forms. For example, the user 4 can input text to be posted on a web page using a text generation input form and upload images using an image input form. In addition, the user terminal 2 provides tools for selecting and customizing web parts. Based on the web part information from the web part management database 1212, the user 4 can select parts that suit his / her preferences and needs and add them to the page. This enables the user 4 to efficiently build a website that is optimal for his / her business or service.

[0048] The user terminal 2 also provides a real-time preview function. This allows the user 4 to immediately check the results of any changes made to the website. For example, when adjusting the layout or style of a web part, the user 4 can check in real time how the changes will be displayed. This makes it possible to improve the consistency and quality of the design. The user terminal 2 also has an authentication and authorization function for the user 4, which prevents unauthorized access and safely protects the data of the user 4. The communication between the user terminal 2 and the website creation support device 1 is encrypted, ensuring thorough data protection.

[0049] (Website creation support device 1: communication unit 13) The website creation support device 1 has a communication unit 13 that enables communication of site-related information regarding the website with a user terminal 2 operated by a user 4. The communication unit 13 has a communication function that establishes communication with the user terminal 2 operated by the user 4. The communication unit 13 is capable of communicating with the user terminal 2 via a data communication network such as the Internet 5. Here, examples of the "data communication network" include the Internet 5, an intranet, an extranet, a mobile communication network, a wide area network (WAN), a local area network (LAN), a metropolitan area network (MAN), and a satellite communication network.

[0050] The communication unit 13 is constructed with a highly reliable hardware configuration in order to reliably establish communication with the user terminal 2 and smoothly exchange information. Specifically, the communication unit 13 is provided with a network interface card (NIC) that ensures high-speed data transfer and stable connection and minimizes the risk of delays and data loss. The communication unit 13 also has routers and switches that are responsible for efficient routing and distribution of data packets and optimize network traffic management. Furthermore, the communication unit 13 is provided with a firewall and an intrusion detection system (IDS) that protect the network from external unauthorized access and internal security threats and enhance data protection for the user 4.

[0051] The data communication between the communication unit 13 and the user terminal 2 is not limited to a data communication network such as the Internet 5, and Bluetooth (registered trademark), a wireless communication standard for short-distance communication, may be used. The data communication may be a dedicated line to prevent information leakage. Furthermore, when data communication is performed between the communication unit 13 and the user terminal 2 via the Internet 5, it is preferable that the communication unit 13 and the user terminal 2 have a VPN (Virtual Private Network) function. When a VPN function is incorporated in the communication unit 13 and the user terminal 2, for example, data communication from the user terminal 2 passes through a VPN connection, so that the data communication is protected from external unauthorized access and secure communication is ensured.

[0052] The communication unit 13 configured as described above enables the user 4 to operate the design and contents of the website in real time. By connecting the communication unit 13 to the user terminal 2, the user 4 can transmit various information about the website and receive necessary data and support information. For example, data entered by the user 4 to build a website and information about selected web parts are transmitted to the website creation support unit 11 via the communication unit 13 and stored in the web part management database 1212. Similarly, selection candidates and design suggestions for web parts generated by the website creation support unit 11 are also transmitted to the user terminal 2 via the communication unit 13.

[0053] Furthermore, the communication unit 13 has a function for realizing real-time data synchronization. This allows the user 4 to immediately reflect changes to the design and contents of the website. For example, when the user 4 changes the layout, the changes are updated in real time via the communication unit 13 and displayed on the user terminal 2. This allows the user 4 to immediately check the results of the changes and quickly make any necessary corrections.

[0054] (Website Creation Support Device 1: Website Creation Support Unit 11) Website creation support device 1 has a control unit (computer) that is a website creation support unit 11. Website creation support unit 11 has a function of predicting the website desired by user 4 based on profile information and site-related information, extracting selection candidates of web part information suitable for website creation from web part management database 1212, and presenting it to user 4 as web creation support information, thereby supporting the creation of the website.

[0055] Here, "guessing the website desired by user 4" is performed through multiple steps. Specifically, profile information of user 4 is collected. This information includes name, occupation, company name, areas of interest, etc. For example, if user 4 is a lawyer and is interested in business law, this information is stored in the profile database 1211. Next, site-related information is collected. For example, details such as the purpose of the website user 4 wants to create, target audience, design preferences, and required features are included. After this information is collected, the profile information and site-related information are analyzed to guess the website desired by user 4. Large-scale language models (LLMs) and machine learning algorithms are used in this process.

[0056] "Selection candidates for web part information suitable for creating a website" is a suggested list of web parts that are most suitable for the characteristics and purpose of the website desired by User 4. These selection candidates are generated based on User 4's profile information and site-related information, and are elements for optimizing the design and functions of the website. For example, if User 4, a lawyer, desires a website focusing on business law, a catchy slogan emphasizing reliability, a description of legal consultation services, and profile parts that highlight expertise are suggested as selection candidates. Parts that display customer feedback and ratings are also selection candidates.

[0057] The website creation support unit 11 will be described in detail. The website creation support unit 11 has a support function for the user 4 to effectively and efficiently build and manage a website. The website creation support unit 11 selects optimal web parts and proposes a website design based on the profile information and site-related information of the user 4. Specifically, the website creation support unit 11 collects information such as the name, occupation, company name, and areas of interest of the user 4 as the profile information of the user 4, and collects detailed site-related information about the website that the user 4 wants to create. Based on the collected profile information and site-related information, the website creation support unit 11 predicts the website desired by the user 4 and generates a selection candidate of optimal web part information for it. A web part selection algorithm using a large-scale language model (LLM) or a machine learning model is used in this generation process. The details of the web part selection algorithm will be described later. This makes it possible to deeply understand the intentions and preferences of the user 4 and propose web parts and templates for providing optimal designs and functions.

[0058] For example, when user 4 creates a website for a law firm, a catchy slogan emphasizing reliability, profile parts highlighting expertise, parts displaying customer feedback, etc. are suggested as selection candidates. In addition, effective combinations of parts are suggested based on user 4's past selections and success stories of other similar users. Website creation support unit 11 transmits information on the selected web parts as web creation support information to user terminal 2 via communication unit 13. User 4 can proceed with building a website based on this web creation support information. That is, user 4 can easily arrange and customize the suggested parts through drag-and-drop operations and input forms.

[0059] Furthermore, the website creation support unit 11 provides a real-time design preview function, which allows changes made by the user 4 to be reflected immediately, and allows the user 4 to proceed with the creation process while checking the appearance and functions of the website. This allows the user 4 to complete the website in a form that is closest to his or her vision.

[0060] (Website creation support device 1: Website creation support unit 11: Web part selection algorithm) Examples of web part selection algorithms include rule-based algorithms, machine learning algorithms based on machine learning models, recommendation algorithms, attribute-based selection algorithms, large-scale language algorithms based on large-scale language models, etc. By combining these algorithms, it is possible to select optimal web parts that meet the diverse needs of users 4.

[0061] In detail, the rule-based algorithm analyzes the profile information and site-related information of User 4 based on a set of predefined rules and conditions to infer the characteristics of the website desired by User 4. For example, it extracts website characteristics related to a specific profession or field of interest, and generates optimal web part information selection candidates based on them. This method is characterized by its simplicity, ease of understanding, and ability to reliably respond to specific conditions. For example, if User 4 is a "lawyer," it infers the structure of the website that emphasizes "reliability" and "professionalism," and selects related catchphrases and legal advice parts.

[0062] The machine learning algorithm learns from past data, finds patterns and trends, and makes predictions. The algorithm analyzes the profile information and site-related information of User 4 to infer the characteristics of the website desired by User 4. For example, it generates optimal web part information selection candidates based on the data of websites created in the past by User 4 and other similar users. This makes it possible to preferentially suggest web parts frequently used by User 4 and web parts highly rated by other similar users.

[0063] The recommendation algorithm predicts the characteristics of the website desired by User 4 based on User 4's preferences and behavioral history. This algorithm learns User 4's tendency to prefer specific designs and content, and generates web part information selection candidates based on that. For example, it can refer to designs that User 4 has used favorably in the past, or parts that other Users 4 have selected under similar conditions.

[0064] The attribute-based selection algorithm infers the characteristics of the website desired by User 4 based on the attribute information of each web part (occupation, interests, age, etc.). For example, it generates optimal web part information selection candidates taking into account attributes such as the design, function, frequency of use, and rating of the web parts. This makes it possible to provide the web parts most suitable for the specific requests and conditions of User 4.

[0065] The large-scale language algorithm uses a large-scale language model to analyze the profile information and site-related information of the user 4, and infers the characteristics of the website desired by the user 4. For example, if the user 4 inputs "I want to create a website for a reliable law firm," the large-scale language model extracts keywords such as "reliability" and "law firm" and infers the structure of the related website. Based on this, it generates optimal web part information selection candidates. Details of the website creation support unit 11 to which the large-scale language algorithm is applied will be described later.

[0066] (Website Creation Support Device 1: Effects) As described above, website creation support device 1 has web part management database 1212 which stores web part information which is the elements that make up a website, profile database 1211 which stores profile information of user 4 who creates the website, communication unit 13 which enables communication of site-related information regarding the website between user terminal 2 operated by user 4, and website creation support unit 11 which predicts the website desired by user 4 based on the profile information and the site-related information, extracts selection candidates for web part information suitable for creating the website from web part management database 1212, and presents this to user 4 as web creation support information, thereby supporting the creation of the website.

[0067] In the above-mentioned website creation support device 1, the web parts management database 1212 stores information on various web parts that constitute a website, and thus a wide variety of materials are available for realizing various designs and functions. The profile database 1211 also records profile information such as the age, sex, occupation, and interests of the user 4, and based on this, it is possible to accurately grasp the preferences and needs of the user 4. The communication unit 13 has a function of enabling two-way communication of site-related information with the user terminal 2 operated by the user 4, enabling real-time information exchange. This allows the user 4 to receive feedback at any time during the website creation process, and can respond quickly and flexibly.

[0068] The website creation support unit 11 has a function of predicting the website desired by the user 4 based on the profile information and the site-related information, and extracting appropriate web part information selection candidates from the web part management database 1212. This allows the user 4 to easily build a website that meets his / her desires. The website creation support unit 11 accurately reflects the preferences and needs of the user 4 and proposes an optimal combination of web parts, thereby realizing efficient and accurate website creation. With this configuration, the website creation support device 1 can automatically select optimal web parts based on the profile information and site-related information of the user 4 and present them to the user 4. As a result, the burden on the user 4 in selecting designs and functions is reduced, and suggestions optimized for each individual user 4 are made, so that a high-quality website can be created in a short time.

[0069] (Website Creation Support Device 1: Website Creation Support Unit 11: Example of Application of Large-Scale Language Model) In the website creation support device 1 configured as above, the website creation support unit 11 may be configured to include a large-scale language model, and to suggest web part selection candidates using this large-scale language model.

[0070] Specifically, as shown in FIG. 4, the website creation support unit 11 has a large-scale language model unit 1131 having a large-scale language model that combines one or more types of natural language processing to probabilistically predict how likely words or sentences given in a prompt are to occur in natural language, and a selection candidate creation prompt generation unit 1132 that, when site-related information is received from the user terminal 2 via the communication unit 13, includes the site-related information in the prompt of the large-scale language model unit 1131, predicts the website desired by the user 4 based on the web part information in the web part management database 1212 and the profile information in the profile database 1211, and causes the large-scale language model to generate selection candidates for web part information suitable for creating a website as web creation support information.

[0071] (Website creation support device 1: Website creation support unit 11: Large-scale language model unit 1131) The large-scale language model unit 1131 has a large-scale language model. Here, the "large-scale language model" is a type of probability model used in natural language processing, and is a model for probabilistically predicting how likely a given word or sentence is to occur as natural language. Specifically, the language model calculates the occurrence probability of a given word string or sentence, or compares the occurrence probabilities of multiple word strings or sentences, thereby making it possible to automatically generate the most likely word or sentence based on the context when predicting the next word or sentence, or to generate a sentence that satisfies a specific condition. That is, the large-scale language model is a language model that probabilistically predicts how likely a word or sentence given in a prompt is to occur as natural language by combining one or more types of processing such as morphological analysis, syntactic analysis, semantic analysis, context analysis, and intention analysis in natural language processing. The large-scale language model unit 1131 is configured to analyze a prompt, the contents of which instruct the generation of web creation support information such as selection candidates for web part information, using the large-scale language model, and to predict and generate web creation support information using the large-scale language model based on the contents of the analyzed prompt.

[0072] "Natural language processing" is a method for enabling a computer to understand text and voice data written in natural language and execute processing according to the purpose. Specifically, examples include morphological analysis, which breaks down natural language into "morphemes," the smallest units that make up the language, and assigns information such as parts of speech; syntactic analysis, which analyzes the grammatical structure of natural language to clarify the structure and meaning of a sentence; semantic analysis, which understands the meaning of words and sentences by analyzing the meaning of natural language, and performs logical judgment and inference; contextual analysis, which understands natural language while considering the context before and after a sentence; and intent analysis, which extracts the intention of a speaker or writer from a dialogue or text using natural language. As a result, "natural language processing" processes natural language by combining processes such as morphological analysis, syntactic analysis, semantic analysis, contextual analysis, and intent analysis, and enables the generation of web creation support information that supports the collaborative activities of this embodiment, machine translation, automatic summarization, question answering systems, and voice recognition.

[0073] A "prompt" is a word or sentence input to the large-scale language model unit 1131, and serves as a starting point for the large-scale language model to generate web creation support information. The generation of web creation support information by a large-scale language model based on a prompt is a major difference from normal (conventional) machine learning. In normal machine learning, a model learns from training data and predicts output data for input data. For example, a machine learning model for image recognition learns from training data of images of cats and dogs and classifies the input image as either a cat or a dog. On the other hand, when generating web creation support information by a large-scale language model, the model learns from training data as well as prompts that provide instructions and information regarding the output data that you want the model to generate.

[0074] In detail, in normal machine learning, the model can only generate the contents included in the teacher data, while the large-scale language model can generate new contents not included in the teacher data by prompts. For example, even if the teacher data only contains the contents of a project manager in the IT field, the prompts can also generate the contents of a project manager in the marketing field. In addition, in normal machine learning, the model can only generate variations of the contents included in the teacher data, while the large-scale language model can generate new variations of output data not included in the teacher data by prompts. For example, even if the teacher data only contains contents related to technical skills, the prompts can also generate contents related to soft skills. In addition, in normal machine learning, the model can only generate new contents by recursively combining the contents included in the teacher data, while the large-scale language model can generate creative contents not included in the teacher data by prompts. For example, even if the teacher data only contains the contents of the relationship between the IT engineer and the user 4, the prompts can generate a completely new perspective, for example, a new idea from the relationship between the IT engineer and the user 4.

[0075] In addition, it is preferable that the large-scale language model of the large-scale language model unit 1131 is specialized for creating web creation support information. That is, it is preferable that the large-scale language model is connected to a website database that stores data related to websites, and is specialized for creating web creation support information by constantly continuing to learn. The website database records a large amount of data on fields that are important in website creation, such as web design, user experience (UX), content creation, and SEO (search engine optimization). Specifically, the data includes success stories, best practices, the latest design patterns, and user feedback. By utilizing this data, the large-scale language model can make more accurate and appropriate suggestions.

[0076] (Website creation support device 1: Website creation support unit 11: Selection candidate creation prompt generation unit 1132) When the selection candidate creation prompt generation unit 1132 receives site-related information from the user terminal 2 via the communication unit 13, it includes the site-related information in a prompt of the large-scale language model unit 1131, and has the function of predicting the website desired by the user 4 based on the web part information in the web part management database 1212 and the profile information in the profile database 1211, and generating selection candidates of web part information suitable for creating a website in the large-scale language model as web creation support information.

[0077] The selection candidate creation prompt generation unit 1132 exists as an independent module from the large-scale language model unit 1131. The reason for this is that the selection candidate creation prompt generation unit 1132 has a role of generating appropriate prompts based on site-related information, which is different from the role of natural language processing that the large-scale language model unit 1131 has. For this reason, by treating the two as independent modules, it is possible to clearly distinguish the respective roles, and to facilitate system design and maintenance. Also, by implementing the selection candidate creation prompt generation unit 1132 as an independent module, it is possible to flexibly respond when the large-scale language model is updated.

[0078] The selection candidate creation prompt generation unit 1132 has an interface for accepting site-related information received from the user terminal 2, and prompt generation logic for generating a prompt with content instructing the creation of web creation support information based on the received site-related information. The prompt generation logic is for selecting a prompt template appropriate for the web creation support information. Conditional branching is used for this selection, and a prompt template that matches the web creation support information is selected. The selection candidate creation prompt generation unit 1132 also has a prompt transmission interface for transmitting the generated prompt to the large-scale language model unit 1131.

[0079] The prompt templates of the selection candidate creation prompt generating unit 1132 are prepared in advance. The prompt templates provide a document structure and a content framework that are the basis for effectively creating web creation support information in the selection candidate creation prompt generating unit 1132. These templates are customized and designed in advance according to the needs and aspects of the user 4. Specifically, the prompt templates predefine questions and information formats that match the interests and intentions of the user 4. For example, if the user 4 is a lawyer and is trying to create a website for his or her law firm, the prompt template includes items such as "a catchy copy that emphasizes reliability," "a detailed explanation of legal consultation services," and "an introduction of success cases." By using this prompt template, specific web part information required by the user 4 is effectively extracted.

[0080] Furthermore, the selection candidate creation prompt generation unit 1132 combines the profile information of the user 4 with the site-related information to grasp the overall picture of the website intended by the user 4 and generates a prompt according to that intention. For example, if the user 4 desires a "casual legal consultation website targeted at young people," the prompt template is customized to include contents such as "friendly design elements," "SNS linkage function," and "strengthened FAQ section." The prompt generation logic of the selection candidate creation prompt generation unit 1132 selects the prompt template that best suits the desires of the user 4 through complex conditional branching. At this time, more accurate prompts can be generated based on information learned from past data and similar cases. By sending this prompt to the large-scale language model unit 1131, specific web part information required for creating a website is generated.

[0081] The prompt transmission interface has a function of quickly and accurately transmitting the generated prompt to the large-scale language model unit 1131. The prompt transmission interface enables smooth transmission and reception of prompts, enabling real-time information processing. This enables user 4 to efficiently proceed with creating a website. As described above, the selection candidate creation prompt generation unit 1132 generates appropriate prompts based on the needs of user 4 and works in cooperation with the large-scale language model unit 1131 to provide web creation support information, thereby making it possible to effectively support user 4 in creating a website.

[0082] The selection candidate creation prompt generator 1132 may be constructed with a rule-based logic circuit or may be constructed using a machine learning model. In the rule-based case, the prompt is generated using explicit conditional branching and transition rules. On the other hand, in the case of using a machine learning model, a more flexible prompt can be generated using a model learned from data (information) in the web part management database 1212.

[0083] It is preferable that the selection candidate creation prompt generating unit 1132 is capable of receiving various information stored in the profile database 1211 and the web part management database 1212, and based on this information, the needs of the user 4 are continuously evaluated and the prompt template is updated. In this case, for example, when the user 4 adds new interests or concerns to the profile information, the prompt template is updated accordingly. This update is for reflecting the information input by the user 4 in real time, so that the selection candidate creation prompt generating unit 1132 can generate a prompt based on the latest information. For example, if the user 4 has recently shown an interest in "online seminars" or "webinars," the corresponding prompt template is automatically updated to include items such as "online seminar announcement," "webinar holding information," and "participant reviews."

[0084] In this way, the selection candidate creation prompt generation unit 1132 can continuously learn the profile information and past site-related information of the user 4 and flexibly update the prompt template based on the results. This allows the information provided by the user 4 when creating a website to be more accurately reflected, making it possible to select web part information that meets individual needs. Furthermore, the selection candidate creation prompt generation unit 1132 can optimize the prompt template based on the usage history and feedback of web parts selected by the user 4 in the past. For example, if a specific web part is highly rated, the template related to that part is strengthened and recommended to other users 4. In addition, for parts with low ratings, the cause is analyzed and the content of the template is modified as necessary.

[0085] In this way, the selection candidate creation prompt generator 1132 can improve the quality of website creation support information by continually improving the accuracy and applicability of the prompt templates, reflecting the trends and feedback of the user 4. This enables the user 4 to efficiently select web part information that best suits his or her needs, and quickly create a high-quality website.

[0086] According to the above configuration, as shown in FIG. 5, in the website creation support device 1, the large-scale language model unit 1131 and the selection candidate creation prompt generation unit 1132 learn using information from the memory unit 12 having each database 1211-1216, and the site-related information input through the communication unit 13 corresponds to an explanatory variable of machine learning, and in response to a prompt from the selection candidate creation prompt generation unit 1132, the large-scale language model unit 1131 generates web creation support information equivalent to a target variable of machine learning.

[0087] As a result, website creation support device 1 can integrate information from web part management database 1212 and profile database 1211 by including site-related information input by user 4 in the prompt of the large-scale language model, and generate selection candidates of web parts that are optimal for user 4. The selection candidates of web parts are transmitted to user terminal 2 via communication unit 13 as web creation support information, allowing user 4 to check and select the suggested web parts in real time. This enables user 4 to create a website intuitively and efficiently.

[0088] In this way, when a large-scale language model is used, the website creation support device 1 provides advanced functions that are difficult to achieve with normal machine learning algorithms. Specifically, the device utilizes the unique capabilities of large-scale language models, such as complex text generation using natural language processing and content generation based on context. In normal machine learning models, prediction, classification, clustering, etc., mainly using numerical data and categorical data, are common, and there are limitations in terms of processing large-scale text data and generating new text. On the other hand, large-scale language models are good at learning language patterns from large amounts of text data and generating new text based on a given prompt, so in situations where diverse and complex language is required, such as web creation support information, an approach using a large-scale language model is more appropriate, and it is possible to provide functions that are difficult to achieve with conventional machine learning methods alone.

[0089] (Website creation support device 1: Website creation support unit 11: Proposal information creation prompt generation unit 1133) As shown in FIG. 4, it is preferable that the website creation support unit 11 further has a proposed information creation prompt generation unit 1133 which, when selection candidates for web part information are generated, includes site-related information in a prompt of the large-scale language model unit 1131, and causes the large-scale language model to generate proposed information indicating a combination of selection candidates for web part information that will result in a website preferred by user 4 based on the profile information in the profile database 1211 as web creation support information.

[0090] The proposed information creation prompt generation unit 1133 has a function of including site-related information of the user 4 in the prompt of the large-scale language model unit 1131 after the selection candidates of the web part information are generated, and generating more detailed proposed information. Unlike the selection candidate creation prompt generation unit 1132, this proposed information creation prompt generation unit 1133 has a role of generating prompts for making suggestions regarding the structure and design of a specific website. In particular, it has a function of causing the large-scale language model to generate proposed information indicating how to combine the selection candidates of the web part information to create a more effective website.

[0091] More specifically, the suggested information creation prompt generation unit 1133 grasps the overall picture of websites preferred by user 4 based on the profile information and site-related information of user 4, and generates a prompt that matches that intention. For example, if user 4 desires a "casual legal advice site targeted at young people," the prompt generation unit generates suggested information including contents such as "friendly design elements," "SNS linkage function," and "strengthened FAQ section." This suggested information indicates how to combine selection candidates of web part information to create a more effective website.

[0092] Specifically, the suggested information creating prompt generating unit 1133 acquires user 4's profile information (e.g., occupation, interests, past site usage history, etc.) and site-related information (e.g., site purpose, target audience, required functions, etc.) in order to grasp the overall picture of the website desired by user 4. Based on this, the suggested information creating prompt generating unit 1133 extracts selection candidates for web part information that are most suitable for user 4's needs, and creates a prompt for analyzing how to combine them effectively.

[0093] When the large-scale language model is given prompts on how to combine the web part information options, it will generate suggested information such as, "On the homepage, place a catchy header that incorporates 'friendly design elements' and a concise introduction that immediately draws users in. Next, in the main content area, place a 'detailed explanation of legal consultation services' in a way that highlights specific cases and achievements. Also, display 'success stories' in the sidebar to make visitors feel trustworthy. Furthermore, include an 'enhanced FAQ section' at the bottom of the page to allow visitors to quickly obtain information. Finally, include 'social media integration features' on all pages to make it easy for visitors to share."

[0094] The proposal information creation prompt generation unit 1133 exists as an independent module from the large-scale language model unit 1131. The reason for this is that the proposal information creation prompt generation unit 1133 has a role of generating appropriate prompts based on site-related information, which is different from the role of natural language processing that the large-scale language model unit 1131 has. For this reason, by treating the two as independent modules, it is possible to clearly distinguish their respective roles and facilitate system design and maintenance. In addition, by implementing the proposal information creation prompt generation unit 1133 as an independent module, it is possible to flexibly respond when the large-scale language model is updated.

[0095] The proposed information creation prompt generation unit 1133 has an interface for accepting site-related information received from the user terminal 2, and prompt generation logic for generating a prompt with content instructing the creation of web creation support information based on the received site-related information. The prompt generation logic is for selecting a prompt template appropriate for the web creation support information. Conditional branching is used for this selection, and a prompt template that matches the web creation support information is selected. The proposed information creation prompt generation unit 1133 also has a prompt transmission interface for transmitting the generated prompt to the large-scale language model unit 1131.

[0096] The prompt templates for the proposed information creation prompt generation unit 1133 are prepared in advance. The prompt templates provide a document structure and a content framework that serve as a basis for effectively creating web creation support information in the proposed information creation prompt generation unit 1133. These templates are customized and designed in advance according to the needs and aspects of the user 4. In addition, the prompt transmission interface has a function of quickly and accurately transmitting the generated prompts to the large-scale language model unit 1131. The prompt transmission interface allows the prompts to be sent and received smoothly, enabling information processing in real time. This allows the user 4 to efficiently proceed with the creation of the website.

[0097] As described above, the proposed information creation prompt generation unit 1133 generates an appropriate prompt based on the needs of the user 4 and cooperates with the large-scale language model unit 1131 to provide web creation support information, thereby making it possible to effectively support the creation of a website by the user 4. In addition, the proposed information generation prompt generation unit 1133 generates proposed information indicating how to combine web part information selection candidates to create a more effective website, thereby supporting the construction of a high-quality website that meets the expectations of the user 4. In other words, by having the proposed information creation prompt generation unit 1133, the website creation support device 1 is able to provide the user 4 with more accurate design proposals, since it adds a function of proposing not only web part selection candidates but also their combinations. This significantly improves the convenience of the user 4 and further improves the efficiency of website creation.

[0098] (Website creation support device 1: Website creation support unit 11: Text information creation prompt generation unit 1134) It is preferable that the website creation support unit 11 has a text information creation prompt generation unit 1134 which, when selection candidates for web part information are generated, includes site-related information in a prompt of the large-scale language model unit 1131, and causes the large-scale language model to generate text information suitable for a website as web creation support information based on the profile information in the profile database 1211.

[0099] The text information creation prompt generator 1134 has an interface for accepting site-related information received from the user terminal 2, and a prompt generation logic for generating a prompt for generating text information suitable for a website based on the received site-related information. The prompt generation logic has a function for selecting an optimal prompt template for generating text information suitable for a website. This selection uses conditional branching to select a prompt template that matches the site-related information and profile information.

[0100] The prompt templates of the text information creation prompt generating unit 1134 are prepared in advance. The prompt templates provide a document structure and a content framework that serve as a basis for effectively creating text information suitable for a website in the text information creation prompt generating unit 1134. These templates are customized and designed in advance according to the needs and aspects of the user 4. Specifically, the prompt templates predefine questions and information formats that match the interests and intentions of the user 4. For example, if the user 4 is a lawyer and is trying to create a website for his or her law firm, the prompt template includes items such as a "catchphrase emphasizing reliability," "detailed explanation of legal consultation services," and "introduction of success cases." By using this prompt template, the text information creation prompt generating unit 1134 can effectively extract specific text information that matches the needs of the user 4.

[0101] Furthermore, the text information creation prompt generation unit 1134 combines the profile information of the user 4 with the site-related information to grasp the overall picture of the website intended by the user 4, and creates a prompt that generates text information according to the intention. For example, if the user 4 wants a "casual legal consultation website targeted at young people," the prompt template is customized to include contents such as "a friendly catchphrase," "an explanation of legal consultation services for young people," and "casual introduction of success cases." The prompt generation logic of the text information creation prompt generation unit 1134 selects the prompt template that is most suitable for the desires of the user 4 through complex conditional branching. At this time, a more accurate prompt can be generated based on information learned from past data and similar cases. By sending this prompt to the large-scale language model unit 1131, specific text information required for creating a website is generated.

[0102] For example, as catchphrases, "A law firm that promises your trust and peace of mind" and "Legal experts will support you" are generated, and as explanations of legal consultation services, "Our legal consultation services cover a wide range of fields, including contract drafting, business law, and labor issues. We provide advice and support tailored to the individual needs of our customers." and "At our firm, the initial consultation is free of charge. Please feel free to contact us." are generated. In addition, as introductions of success cases, "We quickly resolved Company A's contractual trouble and protected the company's trust." and "We resolved Mr. B's labor issue and restored a fair working environment." are generated. As enhancements to the FAQ section, "Q: Is the initial consultation free? A: Yes, the initial consultation is free." and "Q: How do I book a legal consultation? A: You can easily book by phone or on the website." are generated. As profiles, "Taro Yamada: With over 20 years of experience as a lawyer, he specializes in business law and contract drafting. He has solved many legal problems for companies and individuals." and "Hanako Tanaka: As an expert in UI / UX design, she has led many projects to success. She is also strong in branding." are generated. As features of the service, the following are generated: "We have earned the trust of our customers through our prompt response and careful explanations," and "We provide individually customized legal advice."

[0103] These pieces of text information are customized according to the intentions and needs of the user 4 and are used as website content. Furthermore, the text information generated by the large-scale language model unit 1131 becomes more accurate based on specific information and prompts provided by the user 4. This allows the user 4 to efficiently create a website that best suits his or her needs.

[0104] The prompt transmission interface has a function of quickly and accurately transmitting the generated prompts to the large-scale language model unit 1131. The prompt transmission interface enables smooth transmission and reception of prompts, enabling real-time information processing, thereby enabling user 4 to efficiently proceed with the creation of the website.

[0105] As described above, the text information creation prompt generation unit 1134 generates appropriate prompts based on the needs of user 4 and works in cooperation with the large-scale language model unit 1131 to provide web creation support information, thereby making it possible to effectively support user 4 in creating a website.

[0106] (Website Creation Support Device 1: Analysis and correction function for bounce rate) As shown in FIG. 6, it is preferable that the website creation support device 1 has a function of analyzing the bounce rate of each part of the website, identifying and indicating parts that need improvement, and automatically suggesting corrections based on the indications.

[0107] Specifically, the website creation support device 1 has a bounce rate data collection unit 1161 that collects bounce rates on the website on a web part basis, an evaluation information collection unit 1162 that collects evaluation information posted by visitors to the website, a bounce rate analysis unit 1163 that analyzes the collected bounce rates and identifies web parts with high bounce rates, and a revision suggestion creation support prompt generation unit 1164 that includes the web parts identified by the bounce rate analysis unit 1163 in a prompt of the large-scale language model unit 1131 and generates revision suggestions and explanatory text for the identified web parts in the large-scale language model as web creation support information based on the evaluation information collected by the evaluation information collection unit 1162.

[0108] (Website creation support device 1: Exit rate analysis and correction function: Exit rate data collection unit 1161) The exit rate data collection unit 1161 has a function for collecting detailed information on the exit rate of visitors at each part of the website. Specifically, it is possible to clearly identify which part of the website the visitor left and which web part caused the exit. The exit rate data collection unit 1161 uses a tracking technology that tracks how the visitor moves within the page and at what point he or she leaves the site.

[0109] The exit rate data collection unit 1161 monitors the access data for each web part in real time, and records and stores the data in the exit rate database 1214 in the storage unit 12. This makes it possible to analyze the behavioral patterns of visitors to specific web parts, and to obtain basic data for identifying web parts with high exit rates. The exit rate data collection unit 1161 automatically collects data and updates it periodically, making it possible to perform analysis based on the latest information.

[0110] (Website creation support device 1: Analysis and correction function of the dropout rate: Evaluation information collection unit 1162) The evaluation information collecting unit 1162 has a function for collecting detailed evaluation information from visitors who visit the website. Specifically, it systematically records what evaluations and feedback the visitors have given to each web part and page. The evaluation information collecting unit 1162 collects information directly input by the visitors through interfaces such as evaluation forms, feedback buttons, and questionnaires. The evaluation information collecting unit 1162 collects and stores evaluation information in various formats such as text-based comments and evaluation scores. This makes it possible to specifically grasp the impressions and improvements that the visitors felt about a particular web part or content. This information is extremely important in evaluating the performance of the website and serves as basic data for quantitatively analyzing the high dropout rate and visitor satisfaction.

[0111] (Website Creation Support Device 1: Analysis and Correction Function of Exit Rate: Evaluation Database 1215) The evaluation information collected by the evaluation information collecting unit 1162 is stored in the evaluation database 1215. The evaluation database 1215 is a database for systematically managing and storing evaluation information collected from website visitors. The evaluation database 1215 records in detail the evaluations and feedback given by visitors to each web part and page. The evaluation database 1215 includes evaluation information in various formats, such as text-based comments, evaluation scores, and questionnaire response results, and this information serves as basic data for quantitatively analyzing the experiences and satisfaction levels of visitors.

[0112] The evaluation database 1215 includes a visitor ID, a page ID of the page that was evaluated, a web part ID of the web part that was evaluated, an evaluation score which is a numerical score of the evaluation given by the visitor to the web part or page, text comments which include detailed feedback and comments left by the visitor on the web part or page, and survey results which are the options and free-form comments given by the visitor when answering a survey.

[0113] The evaluation database 1215 is constructed using advanced database technology for efficiently storing, searching, and aggregating the above data. For example, by setting an index using a visitor ID or a page ID as a key, it is possible to quickly search for evaluation information related to a specific visitor or page. In addition, text mining technology is also incorporated for analyzing the evaluation scores and the contents of comments, making it possible to perform detailed analysis of the impressions and opinions of visitors.

[0114] The information recorded in the evaluation database 1215 is used in cooperation with other components of the website creation support device 1, such as the exit rate analysis unit 1163 and the revision proposal creation support prompt generation unit 1164. For example, the exit rate analysis unit 1163 uses the information in the evaluation database 1215 to understand visitors' dissatisfaction with and points to be improved regarding a specific web part, and identifies parts with high exit rates based on the results. The revision proposal creation support prompt generation unit 1164 uses the information in the evaluation database 1215 to suggest specific revision proposals and improvement measures.

[0115] (Website creation support device 1: Exit rate analysis and correction function: Exit rate analysis unit 1163) The exit rate analysis unit 1163 has a function for performing detailed analysis of the collected exit rate data and identifying which part of the website is causing visitors to exit. Specifically, the exit rate is calculated for each web part or page, and it is made clear which element is causing problems for visitors. The exit rate analysis unit 1163 uses statistical methods and machine learning algorithms based on the collected data to analyze patterns and trends of the exit rate.

[0116] To explain in detail how to analyze the patterns and trends of the exit rate using statistical methods based on the collected data, basic statistics are calculated. This includes basic statistical indicators such as the average exit rate, median, and standard deviation. These indicators are used to understand the overall exit rate trend and variation. Next, a time series analysis is performed. To analyze how the exit rate data fluctuates over time, a model that takes into account the effects of moving averages and seasonality is used. This reveals the fluctuation pattern of the exit rate over a specific period of time and evaluates the influence of seasons and events. In addition, a correlation analysis is performed. The correlation of the exit rate between different web parts and pages is investigated to understand how certain elements affect other elements. For example, if a specific page shows a high exit rate, analyzing the relationship between that page and other related pages and navigation elements can help identify the cause of the problem.

[0117] We also conduct segment analysis of the churn rate. We segment visitors based on demographic information and behavioral patterns, and analyze the churn rate for each segment. This makes it possible to clarify the tendency of a particular visitor group to churn under certain conditions, and to implement targeted improvement measures for that group. Finally, we conduct regression analysis. We identify multiple factors that affect the churn rate, and quantitatively evaluate the extent to which those factors affect the churn rate. This makes it possible to identify the most influential factors and prioritize improvement measures based on them.

[0118] In addition, the method of analyzing the patterns and trends of the bounce rate using a machine learning algorithm based on the collected data will be specifically explained. Various data related to the bounce rate obtained in the data collection stage (e.g., visitor demographic information, access time, page dwell time, click pattern, etc.) are preprocessed. Preprocessing includes data cleaning, completion of missing data, and detection and processing of outliers. Next, appropriate features are selected and input into the learning model of the machine learning algorithm. When selecting features, variables that may affect the bounce rate are identified so that those variables optimize the performance of the model. For example, page loading speed, type of content, and past behavioral history of visitors are taken into account.

[0119] Next, a model is built using a machine learning algorithm (e.g., decision tree, random forest, support vector machine, neural network, etc.). To train the model, the collected data is used as a training dataset so that the model learns the patterns and trends of the churn rate. After the model is trained, the performance of the model is evaluated using a validation dataset. Evaluation metrics include precision, recall, F1 score to balance precision and recall, and AUC-ROC (Area Under the Curve - Receiver Operating Characteristic) to evaluate the overall performance of the model. This allows us to confirm how accurate the model is in predicting actual data.

[0120] Furthermore, the model is tuned to improve its performance through optimizing hyperparameters. This process uses techniques such as cross-validation, grid search, and random search. Finally, the optimized model is used to analyze the patterns and trends of bounce rates in real time or batch processing. The results of this analysis help identify the cause of high bounce rates in specific parts or pages of a website. For example, if the slow loading speed of a particular page is the cause of high bounce rates, specific improvements can be suggested, such as optimizing that page.

[0121] In addition, the bounce rate analysis unit 1163 can utilize a variety of methods, such as heuristic analysis, visitor testing, heat map analysis, session replay, and feedback tools, in addition to statistical methods and machine learning algorithms, to analyze patterns and trends in bounce rates based on the collected data.

[0122] Here. "Heuristic analysis" is a method of solving problems based on experience and intuition. This method involves forming hypotheses based on visitor behavior patterns and website characteristics to identify the cause of the bounce rate. For example, if a particular page has a high bounce rate, you can intuitively evaluate problems with the design, content, and navigation of that page. "Visitor testing" is a method of identifying the cause of the bounce rate by having actual visitors use the website and observing their behavior and reactions. Visitor testing allows you to directly observe the points of bounce and the causes of visitor frustration. This allows you to gain a deeper understanding of the reasons why visitors leave.

[0123] "Heat map analysis" is a method to visually show which parts of a web page visitors pay attention to and where they click. By using heat maps, it is possible to understand which parts visitors leave and which elements are not attracting attention. This allows you to identify the elements that cause the dropout and take measures to improve. "Session replay" is a method to analyze the visitor's experience in detail by recording the visitor's behavior on a website and playing it back later. By using session replay, it is possible to specifically observe how the visitor operates the website and at what points the visitor leaves. This allows you to identify the visitor's behavior pattern and the cause of the dropout. "Feedback tool" is a method to collect feedback directly from visitors. Using pop-up windows and feedback forms, you can ask the visitor the reason for the dropout and what areas can be improved. By analyzing this feedback, you can understand the cause of the dropout and take appropriate measures to improve.

[0124] In order to improve the accuracy of the analysis, the churn rate analysis unit 1163 is preferably used in combination with the above-mentioned multiple analysis methods. For example, when a statistical method is combined with a machine learning algorithm, the statistical method is effective in quickly grasping basic patterns and trends of data. For example, by calculating basic statistics such as the average, median, and standard deviation, the overall trend of the data can be intuitively understood. At this stage, outliers in the data can be identified, and by grasping the basic distribution and range of variation, information useful for detailed analysis can be provided. On the other hand, machine learning algorithms are excellent at discovering complex patterns and nonlinear relationships that are likely to be overlooked by statistical methods. For example, by using algorithms such as decision trees, random forests, and neural networks, a predictive model can be constructed from the data, enabling more accurate predictions and classifications. This makes it possible to identify specific behavioral patterns and factors, rather than simply grasping trends, and to perform detailed analysis of the churn rate.

[0125] Combining statistical methods with machine learning algorithms increases the breadth and depth of analysis. For example, statistical methods can be used to grasp the basic overview of the data and discover specific patterns or anomalies, and then the data can be fed into machine learning algorithms for more advanced analysis. This process allows the complex factors behind the data to be revealed through specific predictions and classifications while quickly grasping the overall trend of the data. Furthermore, by evaluating the results of machine learning algorithms with statistical methods, it becomes possible to verify the accuracy and reliability of the model. For example, the performance of the model can be quantitatively evaluated by comparing the prediction results of the model with actual data and calculating indicators such as the average error and standard deviation. As a result, the combination of statistical methods and machine learning algorithms is excellent in that it quickly grasps the basic trends of the data while discovering complex patterns and nonlinear relationships, and achieving highly accurate predictions and classifications. This combination improves the accuracy of the churn rate analysis and makes it possible to take effective improvement measures.

[0126] The revision proposal creation support prompt generation unit 1164 has a role of including the web parts identified by the dropout rate analysis unit 1163 in the prompt of the large-scale language model unit 1131. Specifically, it has a function of generating revision proposals and explanations for the identified web parts using the large-scale language model based on the evaluation information collected by the evaluation information collection unit 1162. This revision proposal creation support prompt generation unit 1164 creates prompts to present improvement methods for web parts with high dropout rates.

[0127] The revision suggestion creation support prompt generation unit 1164 uses the exit rate analysis unit 1163 to identify web parts with high exit rates. Next, detailed information about the identified web parts is included in a prompt for the large-scale language model unit 1131. Furthermore, a prompt for generating specific revision suggestions is created based on the visitor evaluation information and feedback collected by the evaluation information collection unit 1162. This prompt includes instructions for the large-scale language model to generate appropriate revision suggestions and accompanying explanatory text.

[0128] When generating the prompt, the user 4's profile information and site-related information are also taken into consideration. This allows the generation of revision suggestions that are in line with the needs and expectations of the user 4. For example, if visitor feedback indicates that navigation on a particular page is complicated, the revision suggestion creation support prompt generation unit 1164 generates a prompt that includes specific suggestions for simplifying the navigation. This prompt is sent to the large-scale language model unit 1131, which generates specific revision suggestions and explanations. The generated revision suggestions and explanations are presented to the user 4 to help improve the website.

[0129] (Website Creation Support Device 1: Analysis and Correction Function of Exit Rate: Processing Operation) The website creation support device 1 configured as described above has the functions of collecting visitor behavior data on a website, identifying points with high dropout rates, and automatically generating improvement proposals. The specific processing operations of the website creation support device 1 will be explained below with reference to the dropout rate database 1214 in FIG. 7.

[0130] The exit rate data collection unit 1161 records the behavior of visitors on each page of the website. For example, actions such as a visitor clicking on the home page or scrolling through the about page are recorded together with a timestamp (see "Data table of exit rate data" in the figure). Data such as the visitor ID, page ID, action, and timestamp are recorded, which allows the visitor's exit behavior on each page to be identified.

[0131] The collected data is then sent to the exit rate analysis unit 1163, where it is analyzed using statistical methods and machine learning algorithms. This analysis calculates the exit rate of each web part (see the "Exit Rate Analysis Data Table" in the figure). For example, specific values ​​such as the exit rate of the homepage header being 0.05, the footer being 0.15, and the sidebar being 0.25 are shown. This makes it clear which web parts are causing problems for visitors.

[0132] A revision proposal creation assistance prompt generation unit 1164 operates on web parts with high bounce rates identified by the bounce rate analysis unit 1163. Information on web parts with high bounce rates is input as a prompt to the large-scale language model unit 1131, and a revision proposal and an explanatory text are generated taking into consideration visitor evaluation information collected by the evaluation information collection unit 1162. The revision proposal includes specific suggestions on how to improve the web part (see "Data table of revision proposal suggestion function" in the figure). For example, specific suggestions such as reducing clutter in the sidebar and adding contact information to the footer are presented.

[0133] The generated revision suggestions and explanatory text are then provided to the user terminal 2 as website creation support information. This allows the user 4 to revise the website based on the specific improvement suggestions. Furthermore, the analysis result display dashboard displays the total number of visits, bounce rate, average time on page, and conversion rate in real time, making it possible to continuously monitor the performance of the website.

[0134] (Website Creation Support Device 1: Exit Rate Analysis and Correction Function: Exit Rate Database 1214) The exit rate database 1214 records detailed data on the behavior of visitors to each page and web part on the website, and stores data for identifying locations with high exit rates. Specifically, it has a "data table of exit rate data," a "data table of exit rate analysis," a "data table of revision suggestion function," and a "data table of analysis result display dashboard."

[0135] The "data table of exit rate data" is collected by the exit rate data collection unit 1161 and records the detailed behavioral history of each visitor who visits the website. This data table consists of four main items: visitor ID, page ID, action, and timestamp. The visitor ID uniquely identifies each visitor, and the page ID indicates the specific web page the visitor visited. The action item records the specific action the visitor took on the page (e.g., click, scroll, hover, exit, etc.). The timestamp indicates the exact date and time when these actions occurred and is used to track the chronological order of the actions.

[0136] For example, the behavioral history of visitor ID1 shows that he clicked on the home page at 10:00 on January 1, 2023, and then performed a hover action on the same page at 10:15. Also, visitor ID2 is recorded as scrolling on the about page at 10:05, and then scrolling again on the same page at 10:20. As for visitor ID3, it shows that he clicked on the contact page at 10:10, and left that page at 10:25.

[0137] In this way, the data table of bounce rate data details the specific actions and timing of each visitor, allowing you to understand which parts of your website are being used by visitors and how they are being used. This data can be used to evaluate the performance of specific sections or content of your website and to analyze visitor behavior patterns.

[0138] The "Exit Rate Analysis Data Table" records the exit rate of each web part in detail. This data table quantifies the extent to which each part of the website affects visitor exit, and is very important for identifying areas for improvement on the website. Specifically, it contains the web part ID, page ID, exit rate, and analysis timestamp. The web part ID is an identifier that uniquely identifies each web part, and the page ID indicates the page on which the web part is located. The exit rate is an index that indicates the extent to which a specific web part contributes to visitor exit, and the higher the value, the greater the impact as a cause of exit. The analysis timestamp indicates the date and time when this exit rate was measured and analyzed.

[0139] For example, the header on the home page has a relatively low bounce rate of 0.05, which means that the header has little impact on visitor bounces. On the other hand, the banner on the contact page has a high bounce rate of 0.3, which means that visitors may have a negative reaction to the banner. Web parts with such high bounce rates are identified as areas in need of improvement.

[0140] The exit rate analysis unit 1163 analyzes these data in real time and continuously evaluates the performance of the website. This allows a quantitative evaluation of the impact of a specific web part on the exit of visitors and provides basic data for taking measures to improve the performance. For example, for web parts with high exit rates, it becomes possible to propose specific improvement measures such as reviewing the design or changing the content.

[0141] The "data table of the correction proposal function" records specific correction proposals for web parts with high bounce rates in detail. This data is the basis for clarifying the improvement points for each part of the website and making efficient improvements. Specifically, it includes the web part ID, the proposal content, the priority, and the proposal timestamp. The web part ID is an identifier for uniquely identifying each web part, which allows the correction proposal to be accurately associated with a specific web part. The proposal content describes the details of the specific correction proposal and clearly indicates what kind of improvement measures are required. The priority indicates the urgency or importance of the correction proposal and is classified as high, medium, or low. This makes it possible to effectively formulate an improvement plan for the entire website. The proposal timestamp indicates the date and time when the correction proposal was proposed, and is used for history management of the correction proposal and progress confirmation of the improvement.

[0142] For example, a low-priority suggestion for the homepage header, "Improve visibility," was recorded on January 3, 2023 at 10:00, with the aim of redesigning and positioning the header to help visitors find information more easily. Similarly, a high-priority suggestion for the sidebar, "Reduce clutter," was recorded, with the aim of organizing the sidebar elements to help visitors access the information they need without getting lost.

[0143] These specific correction suggestions are presented, which allows for efficient website improvement. By implementing the correction suggestions for each web part based on priority, it is possible to reduce the bounce rate. In addition, the proposal timestamp allows you to track which correction suggestions were presented and when, making it easy to manage the progress of improvement activities. Furthermore, the data from the correction suggestion function works in conjunction with the bounce rate analysis data. After web parts with high bounce rates are identified, the correction suggestions are quickly generated and presented, shortening the website improvement cycle. This is expected to continuously improve the performance of the entire website and increase visitor satisfaction.

[0144] The "Analysis Result Display Dashboard Data Table" contains information for detailed monitoring and evaluation of the overall performance of the website. This data table is an important source of information for website operators to understand the current status of their site and plan and take specific actions for improvement.

[0145] Total visits indicates the total number of visitors who visited a website within a specific period of time, and are used to evaluate the popularity and ability of a website to attract visitors. For example, a total visits of 1000 indicates that 1000 visitors visited the website within the relevant period. Exit rate indicates the percentage of visitors who left the site after viewing a particular page. For example, an exit rate of 0.2 means that 20% of visitors left the site after viewing a particular page. This data can be used to identify problematic pages or web parts and take measures to improve them.

[0146] Average time on page indicates the average time a visitor stays on a particular page and is used to evaluate the attractiveness of content and the ease of use of a page. For example, an average time on page of 2 minutes indicates that visitors stay on the page for an average of 2 minutes. This data is an indicator of how interested visitors are in the content. Conversion rate indicates the percentage of visitors who complete a targeted action on a website (e.g., purchase, registration, inquiry, etc.). This metric is important when evaluating the effectiveness of a website and business results. For example, a conversion rate of 5% means that 5% of visitors completed the targeted action. This data can help improve marketing strategies and conversion funnels.

[0147] The report timestamp indicates the date and time when each evaluation indicator was recorded, ensuring the freshness and real-time nature of the data. For example, if the report timestamp is 10:00 on January 4, 2023, it indicates that it is the latest performance data at that time. This allows website operators to understand the latest situation and respond quickly. In this way, the "Data Table on the Analysis Result Display Dashboard" provides information for evaluating the performance of the entire website from various angles.

[0148] (Website Creation Support Device 1: Analysis and Correction Function of Exit Rate: Effects) As described above, the website creation support device 1 has a bounce rate data collection unit 1161 that collects bounce rates on the website on a web part basis, an evaluation information collection unit 1162 that collects evaluation information posted by visitors to the website, a bounce rate analysis unit 1163 that analyzes the collected bounce rates and identifies web parts with high bounce rates, and a revision suggestion creation support prompt generation unit 1164 that includes the web parts identified by the bounce rate analysis unit 1163 in the prompt of the large-scale language model unit 1131 and causes the large-scale language model to generate revision suggestions and explanatory text for the identified web parts as web creation support information based on the evaluation information collected by the evaluation information collection unit 1162.

[0149] According to the above configuration, by providing the exit rate data collection unit 1161 that collects exit rate data on a web part basis, detailed behavioral data on each part of the website is collected. This behavioral data makes it possible to clearly understand which web parts are causing visitors to exit. In addition, through the evaluation information collection unit 1162, evaluation information from visitors who visit the website is collected. This evaluation information includes specific feedback and reviews from the visitors, and clarifies the problems and dissatisfaction points that the visitors are facing. This information plays an important role in analyzing and improving the exit rate. The exit rate analysis unit 1163 analyzes the collected exit rate data in detail and identifies web parts with high exit rates. This makes it possible to accurately identify which parts of the website are causing visitors to exit. Basic data is provided for taking specific improvement measures for the identified problems. Furthermore, the revision proposal creation support prompt generation unit 1164 includes the identified web parts with high exit rates in the prompts of the large-scale language model unit 1131, and generates revision proposals and explanatory text based on the evaluation information. By utilizing the large-scale language model, it becomes possible to analyze the feedback and reviews of visitors and automatically generate optimal revision proposals. The proposed revisions indicate specific improvements and provide guidelines for their implementation. In addition, by generating explanatory text, the background and purpose of the revisions can be clearly communicated. The proposed revisions and explanatory texts generated in this way are provided as web development support information, enabling quick and effective improvement of websites. This is expected to reduce the bounce rate.

[0150] (Website Creation Support Device 1: Automatic Adjustment Function for Web Parts) It is preferable that the website creation support device 1 has a function for automatically adjusting colors and layouts when combining web parts, thereby maintaining a consistent design.

[0151] Specifically, as shown in FIG. 8, the website creation support device 1 has a design consistency maintenance unit 1171 that adjusts the color and layout of the website based on the theme color, brand guidelines, and existing design patterns contained in the site-related information when web parts are combined, a style adjustment unit 1172 that adjusts the style of each web part based on the adjustment results of the design consistency maintenance unit 1171, and a design information communication control unit 1173 that transmits the website design information adjusted by the style adjustment unit 1172 to the user terminal 2 via the communication unit 13.

[0152] (Website creation support device 1: Automatic adjustment function for web parts: Design consistency maintenance unit 1171) The design consistency maintenance unit 1171 has an important function of automatically adjusting colors and layouts to maintain the consistency of the design of the entire site when web parts are combined. That is, the design consistency maintenance unit 1171 has a function of maintaining the overall visual harmony of the website based on elements such as theme colors, brand guidelines, and existing design patterns contained in the site-related information.

[0153] Specifically, the design consistency maintenance unit 1171 analyzes site-related information and extracts basic color tones, font styles, and layout patterns based on theme colors and brand guidelines. These pieces of information are applied so that the web parts have an appropriately unified design. For example, if the brand guidelines specify a design based on "blue," the design consistency maintenance unit 1171 adjusts the color scheme of all web parts to that theme color. Adjustments are also made based on existing design patterns. Existing design patterns refer to designs that have been successful in the past and standard design templates, and the layout of the website is automatically optimized based on these. For example, the placement of the navigation bar, the style of buttons, and the placement of text blocks are adjusted to be consistent.

[0154] Furthermore, the design consistency maintenance unit 1171 also has a dynamic adjustment function. The dynamic adjustment function is a function that immediately responds when the user 4 adds or changes the placement of a web part, and performs processing to automatically maintain the consistency of the design. For example, when a new web part is added, the design consistency maintenance unit 1171 automatically adjusts the color, font style, and placement of the web part to match the existing theme color and brand guidelines. This allows the newly added part to visually harmonize with other parts, and maintains a unified design. The dynamic adjustment function is also applied when the placement of an existing web part is changed. For example, when the user 4 moves a part to a different position, the design elements around the part are automatically adjusted, and a consistent layout is maintained as a whole. This includes adjusting margins, rearranging text, changing image sizes, and the like.

[0155] Furthermore, the dynamic adjustment function provides visual feedback in real time every time the user 4 adds or changes the position of a web part. This allows the user 4 to instantly see how the changes are reflected and intuitively adjust the design. For example, when moving a part by dragging and dropping it, the user 4 can preview in real time how the new position of the part will look. In this way, the dynamic adjustment function allows the user 4 to quickly build a website that has a unified look as a whole without having to worry about adjusting the design of each part. This allows efficient website creation while maintaining the quality of the design.

[0156] (Website creation support device 1: Automatic adjustment function of web parts: Style adjustment unit 1172) The style adjustment unit 1172 has a function of automatically adjusting the style of each web part based on the adjustment result of the design consistency maintenance unit 1171. This maintains the consistency and visual harmony of the design of the entire website.

[0157] Specifically, the style adjustment unit 1172 receives the adjustment results generated by the design consistency maintenance unit 1171 and performs style setting for each web part based on the adjustment results. This process includes adjustment of design elements such as color, font, margin, padding, layout, size, etc. For example, if the theme color of the entire website is blue, the newly added web part will also be automatically set to match the theme color.

[0158] The style adjustment unit 1172 maintains visual unity by making each web part follow consistent design guidelines. Also, when the user 4 customizes the style of the website, adjustments are made so as not to disrupt the overall design balance. For example, even if the user 4 changes the font size of a specific part, other style elements are automatically adjusted to maintain overall harmony, taking into account the balance with other parts.

[0159] Furthermore, the style adjustment unit 1172 has a function of reflecting style changes in real time. This means that every time the user 4 makes a change to the design, the change is immediately applied to the entire website, providing visual feedback. For example, if the style of a button is changed, the change is immediately applied to the buttons on all pages, allowing the user 4 to make adjustments while checking a consistent design.

[0160] In this way, the style adjustment unit 1172 automatically adjusts the style of each web part based on the adjustment results of the design consistency maintenance unit 1171, thereby maintaining consistency in the design of the entire website and providing a function to assist the user 4 in creating an efficient, high-quality website.

[0161] (Website creation support device 1: Automatic adjustment function of web parts: Design information communication control unit 1173) The design information communication control unit 1173 has a function of quickly and accurately transmitting the website design information adjusted by the style adjustment unit 1172 to the user terminal 2. This enables the user 4 to check the design changes in real time and make further adjustments as necessary.

[0162] Specifically, the design information communication control unit 1173 has a function of receiving the design information generated by the style adjustment unit 1172 and transmitting it in an appropriate format to the user terminal 2. The design information includes style settings such as color, font, layout, size, etc. of each web part. This information is immediately reflected in the web browser or design tool used by the user 4 and displayed as visual feedback.

[0163] The design information communication control unit 1173 uses a high-speed, reliable communication protocol to transmit design information with low latency. This allows the user 4 to receive quick feedback as design changes are reflected in real time. For example, if the color of a button is changed, the change is immediately reflected on all related pages, and the user 4 can make further adjustments while checking the changed design.

[0164] The design information communication control unit 1173 also has an error check function for ensuring communication stability and data consistency when transmitting design information. This prevents data corruption and transmission errors, and ensures that accurate design information is always delivered to the user terminal 2. The error check function checks the consistency of data before and after transmitting the design information, and retransmits the data if a problem occurs. Furthermore, the design information communication control unit 1173 also has a function for receiving feedback from the user 4. When the user 4 reports a change request or problem regarding the design, the information is collected, and feedback is provided to the style adjustment unit 1172 and the design consistency maintenance unit 1171 as necessary. In this way, the design information communication control unit 1173 provides an important function for supporting real-time design confirmation and adjustment by quickly and accurately transmitting the design information adjusted by the style adjustment unit 1172 to the user terminal 2, thereby enabling the user 4 to efficiently create a high-quality website.

[0165] The website creation support device 1 configured as described above assists the user 4 in quickly achieving a unified design through an automatic adjustment function based on theme colors, brand guidelines, and existing design patterns, and further reduces the burden on the user 4 through detailed style optimization and real-time design confirmation functions, enabling efficient and high-quality website creation.

[0166] (Website Creation Support Device 1: Other Configurations) As shown in Fig. 1, website creation support device 1 has website creation support unit 11, which is a computer, a storage unit 12, and a communication unit 13. Some or all of the units included in website creation support unit 11 may be configured as either hardware or software. Input to website creation support unit 11 is performed through input unit 22, such as a keyboard, of user terminal 2 operated by user 4. In addition, web creation support information created by website creation support unit 11 is displayed on display unit 21, such as a display of user terminal 2.

[0167] The website creation support device 1 has an input device and a display device, which are not shown. Examples of the input device include a keyboard, a mouse, a touch panel, and a voice input device. Examples of the display device include a liquid crystal display device. This allows the website creation support device 1 to be configured with an information processing device such as a general personal computer, a laptop computer, a smartphone, or a tablet terminal. The website creation support device 1 may lack at least one of the input device and the display device. In this case, since the input device and the display device are provided in an external terminal, which is not shown, the website creation support device 1 can be used as a website creation support server.

[0168] In this embodiment, the case where the functions of the website creation support device 1 are installed in an information processing device is described, but the present invention is not limited to this, and the functions of the website creation support device 1 may be installed in cloud computing. In this case, cloud computing adds computer resources as needed, making it possible to process large amounts of information, enabling faster and more efficient processing, and also making it possible to easily expand processing capacity to accommodate a significant increase in the number of users 4.

[0169] 5, the website creation support device 1 includes a profile database 1211, a web part management database 1212, a site-related database 1213, a dropout rate database 1214, and an evaluation database 1215 as databases, and may further include other databases 1216 such as a visitor behavior database that records detailed data on the behavior of visitors on the site, a conversion database that records data on how website visitors have made conversions (e.g., purchases, form submissions, subscription registrations), and an error / bug database that manages records of technical errors and bugs that occur on the site. By combining these databases, the website creation support device 1 can collect and analyze data from various perspectives, such as visitor behavior, evaluations, and technical problems, and comprehensively improve the performance of the site.

[0170] (Website creation support program) The above describes the case where website creation support unit 11 of website creation support device 1 includes functional blocks, but next, a description will be given of the case where these functional blocks are executed by the processing steps of software (website creation support program). Note that each processing step (S1 to S14) in the website creation support program is executed in parallel by multitasking control.

[0171] As shown in FIG. 9, the customer management process (S1) has a function of efficiently and safely managing all data related to the customer, i.e., the user 4. This process is mainly performed by utilizing the profile database 1211, and includes a series of operations for registering, updating, deleting, and inquiring about the information of the user 4. Specifically, a wide range of data is handled, from basic information of the user 4 (such as name, address, and contact information) to detailed profile information (such as occupation, interests, and past website usage history). In the customer management process (S1), a process for registering a new customer is executed. This process includes a step of inputting basic information of the customer and generating a customer ID. Next, when information of an existing customer needs to be updated, an update process is executed to reflect the latest information. This update process is executed when the customer changes his / her own information or when the administrator corrects the data as necessary.

[0172] Furthermore, in the customer management process (S1), customer authentication and authorization processes are also executed. This allows the access rights of each customer, who is the user 4, to be managed so that only users 4 with appropriate authority can access specific data and functions. In addition, strong security protocols are implemented to prevent unauthorized access. The customer data collected and managed in this way is used to support website creation. For example, optimal web parts are selected and designs are proposed based on the profile information of the user 4. Also, personalized support is provided by referring to the past usage history of the user 4.

[0173] For example, in the profile database 1211 of FIG. 3, Yamada Taro, customer ID 1, is a 35-year-old man who is a lawyer and belongs to Yamada Law Office. He is interested in business law and contracts, and has used the service three times. By managing the information of customers such as Yamada Taro, it is possible to provide him with the appropriate legal consultation services he needs. Furthermore, Tanaka Hanako, customer ID 2, is a 28-year-old woman who is a designer and belongs to Tanaka Design Studio. She provides graphic design and is interested in UI / UX design and branding. She has used the service once, and based on this information, it is possible to provide optimal design services to Tanaka Hanako. Furthermore, Suzuki Ichiro, customer ID 3, is a 45-year-old man who works as an engineer at Suzuki Engineering. He is in charge of system development and is interested in AI and IoT. He has used the service five times in the past, and based on this data, optimal technical support and new development projects are proposed to Suzuki Ichiro.

[0174] In this way, the customer management process (S1) manages detailed customer profile information, making it possible to provide services that are individually customized for each customer.

[0175] The user interface process (S2) is a process for designing and managing the interface used by the user 4 when operating the website creation support device 1. This process is designed to allow the user 4 to operate it intuitively and efficiently. Specifically, the user interface process (S2) arranges and adjusts interface elements such as various input forms, buttons, menus, and navigation bars so that the user 4 can smoothly proceed through each step of website creation.

[0176] Specifically, interface elements such as various input forms, buttons, menus, and navigation bars are arranged and adjusted so that user 4 can smoothly proceed through each step of creating a website. For example, as shown in FIG. 10, a main category list in the leftmost area of ​​display unit 21 provides a menu for user 4 to select the type of web part he or she wants to create. Here, categories such as text, image, gallery, button, and form are displayed. User 4 can proceed to the next step by selecting the required category from this menu.

[0177] If User 4 selects the "Text" category, a "Text Category" menu is then displayed. This menu contains specific text parts such as catch phrases, headings, paragraphs, lists, and quotes. User 4 can then select a more specific text part from here. Furthermore, if User 4 selects "Catch Phrase" from the text category, a "Catch Phrase List" is displayed, providing specific catch phrase options. User 4 can then choose the most suitable catch phrase and add it to the website by dragging and dropping it into the editing window.

[0178] The contents of the menu list in the user interface process (S2) are displayed with priority from among the selection candidates of web part information suitable for creating a website, which are generated by the website creation support unit 11. Specifically, the large-scale language model unit 1131 or the like proposes optimal web parts based on the profile information and site-related information of the user 4, and these are reflected in the menu list. This mechanism allows the user 4 to quickly find the parts that are most suitable for his or her needs and purposes. For example, if the user 4 is a lawyer and is trying to create a website for his or her law firm, web parts such as a "catchphrase that emphasizes reliability" and a "detailed explanation of legal consultation services" are displayed with priority in the menu list. In this way, the user interface process (S2) supports the process of creating a website by efficiently providing web parts that match the purpose of the user 4.

[0179] In this way, the user interface process (S2) employs a hierarchical menu structure, enabling user 4 to easily find and select the desired web part. Also, by utilizing the drag-and-drop function, intuitive operation is possible, allowing user 4 to easily add and edit the components of the website. Furthermore, the user interface process (S2) is equipped with a real-time feedback function. When user 4 adds or edits a web part, the results are displayed instantly, allowing user 4 to proceed with the creation while checking the progress of their work. This feedback function allows user 4 to find the optimal design through trial and error.

[0180] The user input process (S3) is a process for the user 4 to input the information required for each step of creating a website. This process is responsible for accurately receiving the data input by the user 4 and processing and saving it appropriately. Specifically, in the user input process (S3), the user 4 accesses an input form through a user interface and inputs the required information. For example, in the case of a text part, the user 4 can input contents such as a catchy copy, a heading, and a paragraph. In the case of an image part, an image file is uploaded and a description is input. At this time, the user 4 can instantly check the input data through a real-time preview function displayed on the input form.

[0181] The data entered by the user 4 is sent to the back-end system, where it is converted into an appropriate format. Here, the "back-end system" refers to the entire system that operates behind the scenes of a website or application. For example, it includes database management, authentication and authorization execution, etc. The back-end system receives requests from the user interface (front-end), processes the data accordingly, and returns an appropriate response. The conversion process includes text formatting, image resizing, data validation, etc. Data validation verifies that the input is accurate and appropriate, and checks for incorrect data or formatting errors. For example, if a required field is not filled in or the file format is not supported, an error message is displayed to prompt the user 4 to re-enter the data. The converted and validated data is stored in the appropriate database in the storage unit 12. This storage process makes the entered data available for other processing steps and is reflected in the structure and design of the website.

[0182] The web part management process (S4) manages the web parts that make up each part of a website. Specifically, information on all available web parts is managed using a web part management database 1212. Detailed information on each web part, such as the part ID, name, type, function, and creation date, is recorded in the web part management database 1212. By referring to the web part management database 1212, user 4 can easily find the web part he or she needs.

[0183] The web part management process (S4) also utilizes a web part selection algorithm to select the web part that best suits the needs of user 4 and the requirements of the site. The web part selection algorithm analyzes the input information and profile information of user 4 and suggests the most suitable web part based on the analysis. Furthermore, the web part management process (S4) provides a web part management function specific to the customer (user 4). For example, the web parts that a particular user 4 has used in the past or frequently used web parts can be tracked to promote reuse, thereby improving work efficiency.

[0184] Furthermore, the web part management process (S4) recommends the reuse of existing web parts by using a web part reuse algorithm. The web part reuse algorithm analyzes performance data and feedback of web parts that have been used in the past, and re-proposes optimal parts based on that data. This allows the user 4 to efficiently reuse high-quality web parts.

[0185] In addition, the web part management process (S4) may use a large-scale language model to automatically generate and customize web parts. In this case, natural language processing is used to generate web parts in response to the requirements of user 4, and existing parts are improved and new parts are proposed to support more accurate site construction.

[0186] The design generation process (S5) is a step for automatically generating the appearance and layout of the website. The purpose of this process is to integrate various design elements and provide a consistent design in order to realize the website design desired by the user 4.

[0187] Specifically, the design generation process (S5) plans an overall design composition based on the web part information provided by the web part management process (S4). This includes all visual elements, such as the website layout, color scheme, font style, and image placement. These elements are assembled to create a consistent design according to the requirements and brand guidelines of the user 4. The design generation process (S5) also generates a design that is modified and adjusted to meet specific needs and requirements based on the input information of the user 4. For example, it uses machine learning algorithms and natural language processing technology to select and place design elements that match the design style and tone desired by the user 4. It also uses image generation AI and text generation engines to generate unique visual and text content and incorporate them into the design.

[0188] The design generation process (S5) makes adjustments in line with the theme color and brand guidelines. This adjusts the overall color scheme and font style to fit User 4's brand image and design policy, maintaining consistency in the design. This achieves a professional, unified website design. The design generation process (S5) also provides a real-time feedback function. Every time User 4 adds or changes a design element, the preview is updated instantly. This feedback function allows User 4 to pursue the optimal design while checking in real time how the design changes are reflected.

[0189] The automatic adjustment process (S6) is a step for dynamically optimizing the design elements and layout of the website. This process is performed to maintain the consistency and visual harmony of the overall design when the web parts are combined.

[0190] Specifically, the automatic adjustment process (S6) automatically adjusts the colors and layout of the website based on the theme colors, brand guidelines, and existing design patterns contained in the site-related information. This eliminates the need for the user 4 to adjust design elements individually, and makes it easy to achieve a consistent appearance. For example, the hue, font size, and position of a newly added web part are dynamically adjusted so that it harmonizes with the existing design.

[0191] The automatic adjustment process (S6) provides a function for maintaining design consistency in real time, even when User 4 adds a new web part or changes the layout of an existing web part. This allows User 4 to freely add and edit content while maintaining a sense of unity across the site, without having to worry about adjusting the design. For example, if User 4 adds an "Inquiries" section, the design of that section is automatically adjusted to match the other sections.

[0192] Furthermore, the automatic adjustment process (S6) also performs adjustments to realize a responsive design that takes into consideration the display on different devices and screen sizes used by the user 4. This makes it possible to provide consistent display content on a variety of devices, including smartphones, tablets, and desktop computers. For example, when viewing on a smartphone, the layout is converted to portrait orientation and automatically adjusted so that important information is positioned in an easy-to-see position.

[0193] The deployment and update process (S7) is the step of executing a series of procedures necessary to publish and maintain the generated and adjusted website. The purpose of this process is to ensure that the website works properly on the Internet 5 and that the contents are properly updated whenever the user 4 makes changes or updates.

[0194] Specifically, the deployment and update process (S7) uploads the files and data of the completed website to the specified web server 3 and publishes it. This includes uploading static files such as HTML files, CSS files, JavaScript files, and image files, as well as setting up a database. When the user 4 clicks the publish button for the website, the deployment engine places these files and data in the appropriate folders, and the website becomes viewable at the specified domain.

[0195] The deployment and update process (S7) also automatically manages updates to existing websites. When user 4 changes the content of the website or adds new content, the deployment engine detects the change, regenerates the necessary files, and uploads them to the web server 3. This automatic update function allows user 4 to always publish the latest content, and saves the trouble of manually uploading files. The deployment and update process (S7) also performs version management of the changes. This is useful if user 4 wants to revert to a previous version or if a specific update causes a problem. The deployment engine records the change history of each version and can revert to a previous state if necessary. This function improves the stability and reliability of the website.

[0196] Furthermore, the deployment and update process (S7) monitors the performance and security of the published website. This includes the site's load time, error logs, and detection of security threats. If the monitoring system detects an anomaly, a notification is sent to User 4, enabling a quick response. This continuous monitoring makes it possible to maintain the performance and security of the website and provide a high quality service to User 4.

[0197] The customer-specific update notification process (S8) is a step for notifying each customer of updates and changes to the website in a timely manner. The purpose of this process is to ensure that updates made by the user 4 are reflected and to promptly inform the user 4 of important changes.

[0198] Specifically, the customer-specific update notification process (S8) notifies the user 4 that the changes made to the website have been completed when the user 4 makes changes to the website. This notification applies to all updates, such as adding new content to the website, editing existing content, and changing the design. When the user 4 saves the change data, the content is automatically confirmed and a notification is sent that the changes have been successfully applied. The update notification is changed and adjusted to meet the specific needs and requirements based on the settings of the user 4. For example, the user 4 can select the frequency and method of receiving notifications. Multiple methods are provided, such as email notifications, SMS notifications, and push notifications, and can be set according to the user 4's preferences. This allows the user 4 to receive important information in the way that suits him or her best.

[0199] The customer-specific update notification process (S8) also notifies users about important system updates and the application of security patches. These notifications include system maintenance, the addition of new functions, and the correction of security vulnerabilities. The stability and security of the system are maintained by notifying users 4 in advance and encouraging them to take necessary measures so that the operation of the website is not disrupted. Furthermore, the above notifications also include feedback from users 4 about the performance of the website. For example, if there is a change in an important performance indicator, such as a sudden increase in the number of visitors or an abnormal increase in the dropout rate, the information is notified to users 4. The customer-specific update notification process (S8) also has a function for managing the history of notification contents. Users 4 can check past notifications and recheck the update history and important messages. This history management function allows users 4 to refer to necessary information at any time, which can be useful for the operation of the website.

[0200] The customer-specific update reflection process (S9) is a step for appropriately reflecting updates and changes made to the website of each user 4. The purpose of this process is to quickly and accurately apply the requests and changes of the user 4 to the website.

[0201] Specifically, the Customer-Specific Update Reflection Process (S9) starts the process of reflecting the updates and changes made by the user 4 to the website. This includes any kind of change, such as adding or editing text or images, adjusting the design, adding features, etc. Once the user 4 confirms the changes, the content is analyzed and the procedure for applying the updates is automatically determined. The updates are then reflected on each web page. This includes placing the changed content in the appropriate position and applying the updates while maintaining consistency with existing content. For example, if new text is added, the text is adjusted so that it is balanced with other content. Also, if there are design changes, the overall layout and style are updated to maintain consistency.

[0202] The customer-specific update reflection process (S9) also includes a verification step to confirm that the updates have been reflected correctly. This verification step checks whether the updates are displayed as intended and whether the functions are working correctly. For example, if a new function is added, it is tested to see if the function works properly and that there are no defects. It also includes a function to clear the cache after the updates have been reflected. This allows users 4 and visitors to immediately check the latest information. Clearing the cache is extremely important, especially when large-scale changes have been made, as it prevents old data from being displayed.

[0203] The customer-specific update reflection process (S9) also has a function for managing the update history, which allows the user 4 to check the updates and change history that have been made in the past. This history management function is useful when it is necessary to return to a previous state or when past changes are required as a reference. Furthermore, the customer-specific update reflection process (S9) also includes a function for notifying the user 4 when an update is complete.

[0204] The text management process (S10) is a step for creating, editing, storing, and managing text content used on a website. The purpose of this process is to help the user 4 efficiently manage the text content of the website and place it in an appropriate location.

[0205] Specifically, the text management process (S10) collects text data entered by the user 4 and edits or corrects it as necessary. This includes creating text elements such as headings, paragraphs, lists, and citations. The user 4 can easily enter text and adjust the format using a text editor. The text management process (S10) also has a version management function that automatically saves the entered text and can revert to a previous version if necessary.

[0206] Furthermore, the text management process (S10) also provides functions for maintaining the consistency and quality of the text content of the entire website. For example, it performs spell checking, grammar checking, and application of style guides to ensure the quality of the text. It also supports optimal keyword placement and metadata editing from the perspective of SEO (search engine optimization). This makes it possible to improve the website's ranking in search engines and increase visitor access. The text management process (S10) works in conjunction with a text database. This database stores all text content, allowing User 4 to quickly search and access the text he or she needs.

[0207] The exit rate data collection process (S11) is a step for collecting and analyzing the details of the visitor exit rate for each part of the website. This process plays an important role in evaluating the performance of the website and identifying areas that need improvement.

[0208] Specifically, the exit rate data collection process (S11) tracks the behavioral data of visitors when they visit a website in real time. This includes the timing of page transitions, the time spent on each page, where specific actions (clicks, scrolls, hover, etc.) were performed, and the point at which the visitor left the site. This data is automatically collected and stored in the exit rate database 1214.

[0209] The exit rate data collection process (S11) uses tracking technology to record visitor behavior in detail. Specifically, a unique visitor ID is assigned to each visitor, and it is tracked which pages the visitor visited on the site and what operations he or she performed. In addition, the page ID and timestamp of the action are also recorded, making it possible to identify the specific timing and location where the exit occurred. For example, if visitor ID1 clicks on a home page and then leaves the site after staying there for a certain period of time, the detailed behavioral history is saved as data.

[0210] Furthermore, the exit rate data collection process (S11) also has a function of monitoring the access data for each web part in real time. This provides basic data for analyzing the extent to which a specific web part affects the exit of visitors. The collected data is used to identify web parts with high exit rates.

[0211] The exit rate analysis process (S12) is a step in which the patterns and trends of exit rates in each part of the website are analyzed based on the collected visitor behavior data. This process plays an important role in improving website performance and increasing visitor satisfaction.

[0212] Specifically, the exit rate analysis process (S12) uses the data collected in the exit rate data collection process (S11) to calculate the exit rate of each web part and analyzes the data in detail. The exit rate is an index that indicates the rate at which visitors leave the site at a specific web part or page, and this index can be used to identify which parts need improvement.

[0213] In the exit rate analysis process (S12), a combination of statistical methods and machine learning algorithms is used. Statistical methods are effective in quickly grasping basic patterns and trends in data. For example, basic statistics such as the average, median, and standard deviation are calculated to grasp the overall trend of the data. This makes it possible to quickly identify pages and web parts with high exit rates. On the other hand, machine learning algorithms are excellent at discovering more complex patterns and nonlinear relationships. For example, by using algorithms such as decision trees, random forests, and neural networks, a predictive model can be constructed from the data, enabling more accurate predictions and classifications. This makes it possible to obtain analysis results that identify the cause of high exit rates and propose improvement measures.

[0214] Furthermore, the exit rate analysis process (S12) also utilizes a variety of methods, such as heuristic analysis, customer testing, heat map analysis, session replay, and feedback tools. Combining these methods enables a more comprehensive analysis and improves the accuracy of the analysis. For example, heat map analysis can visually show which parts visitors are paying attention to, and session replay can reproduce the specific actions of visitors and analyze them in detail.

[0215] The proposed correction process (S13) is a step of presenting specific proposed corrections to the user 4 for the identified problems based on the results of the exit rate analysis process (S12). The purpose of this process is to support the user 4 in quickly implementing effective improvement measures to improve the performance of the website.

[0216] Specifically, the modification proposal presentation process (S13) generates improvement proposals for the web parts and pages with high bounce rates identified in the bounce rate analysis process (S12) and displays them to the user 4. This includes a wide range of proposals, such as changing the design or content of the web parts, and adding or removing functions. The modification proposals are provided with detailed explanations so that the user 4 can easily understand and put them into practice.

[0217] The revision proposal presentation process (S13) has a function of presenting the revision proposal generated by the revision proposal creation support prompt generation unit 1164 in a visually easy-to-understand manner to the user 4. Specifically, the contents of the revision proposal are displayed as a pop-up window, a notification banner, or an alert on the dashboard. This enables the user 4 to quickly understand which parts of the website need improvement and take appropriate action.

[0218] The proposed revision process (S13) can change the display method according to the priority and urgency of the proposed revision. For example, a proposed revision with a high urgency requires immediate action and is therefore displayed prominently in the center of the screen. On the other hand, a proposed revision with a low priority is displayed in the notification center or part of the dashboard so that the user 4 can check it later. In this way, by adopting a display method according to the importance of the proposed revision, the burden on the user 4 is reduced and effective action is promoted. Furthermore, the proposed revision process (S13) has a function of tracking the implementation status of the proposed revision. When the user 4 takes action based on the proposed revision, the result is recorded and used for later analysis. This makes it possible to evaluate whether the proposed revision actually contributed to improving the performance of the website. This feedback loop allows the effectiveness of the proposed revision to be continuously verified and data is accumulated for making more accurate proposals.

[0219] The analysis result display process (S14) is a step for visually displaying the performance of the website and the behavioral data of the user 4. The purpose of this process is to analyze the collected data and enable the user 4 to intuitively grasp the current state of the website.

[0220] Specifically, the analysis result display process (S14) compiles key evaluation indicators such as the exit rate, number of visitors, duration of stay, and conversion rate, and displays them in the form of a dashboard or report. This allows the user 4 to monitor the performance of the website in real time and quickly take necessary improvement measures. The analysis result display process (S14) acquires necessary data from various databases, and generates graphs and charts based on the data. For example, the transition of the exit rate for each page is displayed in a line graph based on the data acquired from the exit rate database 1214. In addition, data on the conversion rate and number of visitors are visually represented in bar graphs and pie charts. This allows the user 4 to understand the trends and fluctuations of the data at a glance.

[0221] Furthermore, the analysis result display process (S14) has a function that allows the user 4 to view detailed data through a user interface. The user 4 can display detailed analysis results limited to a specific period or a specific page. This function allows the user 4 to obtain information to more specifically identify the cause of a problem and take appropriate measures. The analysis result display process (S14) provides real-time data updates and good operability, enabling the user 4 to make quick decisions based on data. For example, if the number of visitors or conversion rate fluctuates suddenly, the user 4 can immediately analyze the cause and take countermeasures. Such real-time data display and operability make it possible to constantly optimize the performance of the website.

[0222] The analysis result display process (S14) can also perform comparison with past data and trend analysis. This allows the user 4 to understand how the performance of the website has changed over time and obtain basic information for considering long-term improvement measures. This trend analysis is also useful for evaluating seasonal fluctuations and the effectiveness of specific campaigns.

[0223] The above website creation assistance program can be installed in an information processing device such as a personal computer or a tablet terminal to cause the information processing device to function as website creation assistance device 1. The program may be distributed in a state recorded on a computer-readable recording medium such as a CD-ROM or a USB memory, or may be distributed via a two-way communication network or communication line such as the Internet or a one-way communication network such as a television broadcast. In other words, the initial setting storage unit 16 may be a computer-readable recording medium such as a CD-ROM or a USB memory.

[0224] (Website Creation Support Method) The website creation support device 1 is configured to cause a computer (website creation support unit 11) to execute a website creation support method. Specifically, it is a method for causing a computer (website creation support unit 11) to execute each of the above processing steps.

[0225] It should be noted that within the scope of the concept of the present invention, a person skilled in the art may come up with various modifications and alterations. Therefore, it is understood that such modifications and alterations fall within the scope of the present invention. For example, those in which a person skilled in the art appropriately adds, deletes, or modifies the design of the above-described embodiment, or adds, omits, or modifies the conditions of a process, may also be included in the scope of the present invention as long as they include the gist of the present invention. [Explanation of symbols]

[0226] 1. Website creation support device 2. User terminal 3. Web Server 4 User 5. Internet 11. Website Creation Support Department 12 Storage section 13. Communications Department 1211 Profile Database 1212 Web Part Management Database 1213 Site Related Database 1214 Dropout Rate Database 1215 Rating Database 1131 Large-Scale Language Modeling Department 1132 Selection candidate creation prompt generation unit 1133 Proposal information creation prompt generation unit 1134 Text information creation prompt generation unit 1161 Dropout Rate Data Collection Department 1162 Evaluation Information Collection Department 1163 Dropout Rate Analysis Department 1164 Revision Proposal Creation Support Prompt Generation Unit 1171 Design Integrity Maintenance Department 1172 Style Adjustment Section 1173 Design Information and Communication Control Section

Claims

1. a web part management database in which information on web parts which are elements constituting a website is stored; a profile database in which profile information of a user who creates the website is stored; a communication unit that enables communication of site-related information regarding the website between the user terminal operated by the user and the communication unit; a website creation support unit that predicts the website desired by the user based on the profile information and the site-related information, extracts from the web part management database selection candidates of the web part information suitable for creating the website, and presents the selection candidates to the user as web creation support information, thereby supporting the creation of the website; The website creation support unit includes: a large-scale language model unit having a large-scale language model that probabilistically predicts how likely a word or sentence given in a prompt is to occur in natural language by combining one or more types of natural language processing; a selection candidate creation prompt generation unit which, when receiving the site-related information from the user terminal through the communication unit, predicts the website desired by the user based on the web part information in the web part management database and the profile information in the profile database, and generates a prompt for generating selection candidates for the web part information suitable for creation of the predicted website, thereby causing the large-scale language model unit to generate the selection candidates as the web creation support information; a dropout rate data collection unit that collects dropout rates on the website for each of the web parts; a rating information collection unit for collecting rating information provided by visitors to the website; a dropout rate analysis unit that analyzes the collected dropout rates and identifies the web parts with high dropout rates; a revision proposal creation support prompt generation unit that generates a prompt for generating a revision proposal and an explanatory text for the web part based on the evaluation information collected by the evaluation information collection unit, the prompt generating unit causing the large-scale language model unit to generate the revision proposal and the explanatory text as the web creation support information, the revision proposal and the explanatory text being based on the evaluation information collected by the evaluation information collection unit, the revision proposal and the explanatory text being generated as the web creation support information. Website creation support device.

2. a design consistency maintaining unit that adjusts the color and layout of the website based on a theme color, a brand guideline, and an existing design pattern included in the site-related information when the web parts are combined; a style adjustment unit that adjusts the style of each web part based on the adjustment result of the design consistency maintenance unit; a design information communication control unit that transmits the design information of the website adjusted by the style adjustment unit to the user terminal through the communication unit.

2. The website creation support device according to claim 1.

Images