Independent station building and decoration system and method based on Agent RAG and universal language model

Abstract

The invention discloses an independent station building and decoration system based on Agent RAG and a general language model. Comprising a basic information input module, a competing product site generation module, a competing product report generation module, a website architecture generation module, a website keyword generation module, a website topic generation module, a page template planning module and a website data generation module which are connected with an Agent RAG and a universal language model. The invention further discloses an independent station building decoration method based on the Agent RAG and the universal language model, material content generation of the website can be rapidly achieved through the Agent RAG and the universal language model, and the generated content materials can be manually modified and adjusted. The efficiency can be improved, and the SEO effect of the independent station can be ensured.

Description

Technical Field

[0001] This application relates to the field of website construction and decoration, particularly an independent website construction and decoration system and method based on Agentic RAG and general language models. Background Art

[0002] With the gradual increase in the quantity and total amount of global cross-border trade, in order to better display, promote, and publicize their products to foreign customers, cross-border independent websites are an important choice for Chinese cross-border merchants. However, the construction and SEO of cross-border independent websites are traditional manual offline processes, with a complex overall process, long cycle, and low efficiency. Website builders manually search for competitor sites, analyze and summarize competitor sites to formulate website construction plans, manually design website architectures, plan and expand keywords based on customer information, plan website page templates, use search engines to plan website topics, place keywords on corresponding pages, and plan the entire process relying solely on the personal experience of website builders, without strict specifications, resulting in uneven construction effects and SEO levels. From the supply side to the demand side of the entire industry, there is an urgent need to ensure the construction and SEO optimization effects, improve the construction efficiency, and shorten the construction cycle, which is also the goal pursued by website construction service providers. Summary of the Invention

[0003] To solve the technical problems existing in the prior art, this application proposes an independent website construction and decoration system and method based on Agentic RAG and general language models, which can quickly generate website material content through Agentic RAG and general language models, improve efficiency, and at the same time ensure the SEO effect of independent websites.

[0004] To solve the above-mentioned existing technical problems, the purpose of this application is achieved by adopting the following technical solutions: An independent website construction and decoration system based on Agentic RAG and general language models includes a basic information entry module, a competitor site generation module, a competitor report generation module, a website architecture generation module, a website keyword generation module, a website topic generation module, a page template planning module, and a website data generation module connected to Agentic RAG and general language models; the basic information entry module is used for users to fill in the basic information of the website; the competitor site generation module is used to mine and generate competitor sites; the competitor report generation module is used to analyze competitor sites and obtain a competitor report on the advantages and disadvantages of the website; the website architecture generation module is used to generate the architecture of the website; the website keyword generation module is used to plan and expand keywords for the pages within the website; the website topic generation module is used to plan relevant topics for the pages within the website; the page template planning module is used to plan templates for each page type within the website; the website data generation module is used to generate the component content of each page, finally generate the page data of the independent website, and present the effect of independent website construction and decoration on the page.

[0005] Preferably, the basic information input module includes a website basic information input window and a website basic information database.

[0006] Preferably, the competing product site generation module includes an external Agentic RAG and a general language model docking port, a competing product site management database, and a competing product site modification and editing window.

[0007] Preferably, the competing product report generation module includes an external Agentic RAG and a general language model docking port, a competing product report management database, and a competing product site report modification and editing window.

[0008] Preferably, the website architecture generation module includes an external Agentic RAG and a general language model docking port, a website architecture management database, and a website architecture modification and editing window.

[0009] Preferably, the website keyword generation module includes an external Agentic RAG and a general language model docking port, a website keyword management database, and a website keyword modification and editing window.

[0010] Preferably, the website topic generation module includes an external Agentic RAG and a general language model docking port, a website topic management database, and a website topic modification and editing window.

[0011] An independent website building and decoration method based on Agentic RAG and a general language model, using an independent website building and decoration system based on Agentic RAG and a general language model, includes a basic information input module, a competing product site generation module, a competing product report generation module, a website architecture generation module, a website keyword generation module, a website topic generation module, a page template planning module, and a website data generation module connected to Agentic RAG and the general language model; includes the following steps: Step 1: The user inputs the basic information of the website through the basic information input module; Step 2: The competing product site generation module uses Agentic RAG and the general language model to mine competing product sites according to the basic information of the website input in Step 1; Step 3: The competing product report generation module uses Agentic RAG and the general language model to analyze the competing product sites based on the SEO optimization standards and rules, and generates a competing product report for the competing product sites; Step 4: Generate Website Architecture Module Utilize Agentic RAG and general language models to plan the page architecture of the website, including page titles and content for the header and footer, based on the competitor report of the competitor site in Step 3 and the basic profile information entered by the user in Step 1. Step 5: Generate Website Keyword Module Utilize Agentic RAG and general language models to mine keywords for each page based on the page titles and content generated in Step 4, and compare the relevance of the keywords to the corresponding pages. Finally, retain the keywords relevant to the page content. Step 6: Generate Website Topic Module Utilize Agentic RAG and general language models to plan page topics and content for each page in Step 4 based on the keywords generated for each page in Step 5. Step 7: Page Template Planning Module Utilize Agentic RAG and general language models to plan the templates for each page type of the website based on the competitor report generated in Step 3, and generate page templates. Step 8: Generate Website Data Module Utilize Agentic RAG and general language models to plan the component content of each template based on the title, content, and topic information of each page, and finally generate the page data for the independent website and present the effect of independent website construction and decoration on the page.

[0012] Preferably, the basic profile entry module includes a website basic profile entry window and a website basic profile database, and the information entered through the website basic profile entry window is stored in the website basic profile database.

[0013] Preferably, the generate competitor site module includes an external Agentic RAG and general language model docking port, a competitor site management database, and a competitor site modification and editing window, and the competitor site modification and editing window is used to modify and edit the competitor site.

[0014] Preferably, the generate competitor report module includes an external Agentic RAG and general language model docking port, a competitor report management database, and a competitor site report modification and editing window, and the competitor site report modification and editing window is used to modify and edit the competitor site report.

[0015] Preferably, the generate website architecture module includes an external Agentic RAG and general language model docking port, a website architecture management database, and a website architecture modification and editing window, and the website architecture modification and editing window is used to modify and edit the website architecture.

[0016] Preferably, the website keyword generation module includes an external Agentic RAG and general language model docking port, a website keyword management database, and a website keyword modification and editing window, which is used to modify and edit keywords.

[0017] Preferably, the website topic generation module includes an external Agentic RAG and general language model docking port, a website topic management database, and a website topic modification and editing window, which is used to modify and edit website topics.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: With the independent website building and decoration system and method based on Agentic RAG and general language model adopting the above technical solution, users can input the basic information of website materials, and the system calls Agentic RAG and general language model to automatically mine competitor sites, generate competitor reports of competitor sites, generate the navigation structure of the website, automatically mine keywords for website pages, automatically plan website page topics, automatically plan website page templates, and automatically plan the content of page template components; the entire content generation is automatically generated by Agentic RAG and general language model based on the standard specifications of SEO, and the content can be manually proofread and modified; it can not only quickly generate website material content through Agentic RAG and general language model, but also realize manual modification and adjustment of the generated content materials; improve efficiency and ensure the SEO effect of the independent website at the same time. Brief Description of the Drawings

[0019] Figure 1 It is a flow chart of the present invention; Figure 2 It is a flow chart of the basic information input module in the present invention; Figure 3 It is a flow chart of the competitor site generation module in the present invention; Figure 4 It is a flow chart of the competitor report generation module in the present invention; Figure 5 It is a flow chart of the website architecture generation module in the present invention; Figure 6 It is a flow chart of the website keyword generation module in the present invention; Figure 7 It is a flow chart of the website topic generation module in the present invention. Detailed Embodiment

[0020] Next, in combination with the accompanying drawings and specific embodiments, the present application will be further described. It should be noted that, on the premise of non-conflict, any combination of the following-described steps or technical features can form a new step.

[0021] Embodiment 1: As Figure 1 shown, an independent website building and decoration system based on Agentic RAG and a general language model includes a basic information input module, a competing product site generation module, a competing product report generation module, a website architecture generation module, a website keyword generation module, a website topic generation module, a page template planning module, and a website data generation module connected to Agentic RAG and the general language model; the basic information input module is used for users to fill in the basic information of the website; the competing product site generation module is used to mine and generate competing product sites; the competing product report generation module is used to analyze the competing product sites and obtain a competing product report on the advantages and disadvantages of the competing product websites; the website architecture generation module is used to generate the architecture of the website; the website keyword generation module is used to plan and expand keywords for the pages within the website; the website topic generation module is used to plan relevant topics for the pages within the website; the page template planning module is used to plan templates for each page type within the website; the website data generation module is used to generate the component content of each page, and finally generate the page data of the independent website, and present the effect of independent website building and decoration on the page.

[0022] An independent website building and decoration method based on Agentic RAG and a general language model uses an independent website building and decoration system based on Agentic RAG and a general language model, including a basic information input module, a competing product site generation module, a competing product report generation module, a website architecture generation module, a website keyword generation module, a website topic generation module, a page template planning module, and a website data generation module connected to Agentic RAG and the general language model; it includes the following steps: Step 1: The user inputs the basic information of the website through the basic information input module; after the user inputs the basic information of the website in the basic information input module, the system will splice these basic information (website type, site industry, default language, product information, etc.) into relevant content and generate instructions to provide to Agentic RAG and the general language model; Step 2: Generate a competitive product site module Utilize Agentic RAG and a general language model to mine competitive product sites based on the basic information of the website entered in Step 1; Agentic RAG and the general language model, according to the system's instructions (specifically required to filter out certain irrelevant site types such as Amazon, Wikipedia, etc.), mine competitive product sites that match the basic information; Agentic RAG and the general language model return the URLs of the competitive product sites, and the generated competitive product site module parses the content returned by Agentic RAG and the general language model and stores it in the generated competitive product site module; it can automatically mine multiple relevant competitive product sites, and there is no need for manual participation throughout the process. According to the site type of the website construction (B2B or DTC), it automatically mines relevant sites and blocks irrelevant sites, ensuring the accuracy of the competitive product sites; significantly shortens the time required to find competitive product sites, reduces labor costs, and at the same time ensures the accuracy and referenceability of the competitive product sites; Step 3: Generate a competitive product report module Utilize Agentic RAG and a general language model to analyze the competitive product sites based on the SEO optimization standards and rules, and generate a competitive product report for the competitive product sites; based on the URLs of the competitive product sites generated in Step 2; the generated competitive product report module provides the relevant SEO standards to Agentic RAG and the general language model, and Agentic RAG and the general language model conduct competitive product analysis on each competitive product site generated in Step 2, including analyzing the page type (page title, page type, page content), and analyzing whether it conforms to SEO specifications, and then generate a competitive product report; Agentic RAG and the general language model return the competitive product report according to the relevant requirements; the generated competitive product report module receives and parses the data returned by Agentic RAG and the general language model, and finally stores it in the database of the generated competitive product report module; the generated competitive product report module, according to the preset instructions, requires Agentic RAG and the general language model to generate a website construction strategy; Agentic RAG and the general language model generate and return a website construction strategy according to the relevant instructions and prompts; the generated competitive product report module parses the website construction strategy returned by Agentic RAG and the general language model, and finally stores it in the database of the generated competitive product report module; through Agentic RAG and the general language model, it can analyze the competitive product sites according to the specified SEO dimensions, quickly, efficiently, accurately, and can significantly save labor; and ensure the effect and quality of the competitive product analysis; at the same time, it can summarize a website construction strategy based on the competitive product report, providing guidance and assistance for subsequent website construction, whether it is manual or automatic by a large model; Step 4: Generate Website Architecture Module Utilize Agentic RAG and a general language model to plan the page architecture of the website, including the page titles and page content of the header and footer, based on the competitor report in Step 3 and the basic information entered by the user in Step 1. First, concatenate and combine the competitor report, website construction strategy, and basic website construction information generated in Step 3 to generate a prompt. Then, call Agentic RAG and the general language model and input the prompt. Agentic RAG and the general language model generate the website's page architecture information (header, footer, page title, main page content) according to relevant rules and return this page architecture information to the system. The system accepts the returned page architecture information and stores this page architecture information in the page architecture generation module. By using Agentic RAG and the general language model to generate the website's page architecture (header, footer, navigation bar information), it is possible to quickly and conveniently plan the architecture, page types, and main content of the web page, and the planned page content is relatively comprehensive and reasonable. Step 5: Generate Website Keyword Module Utilize Agentic RAG and a general language model to mine keywords for each page based on the page titles and page content generated in Step 4, and compare the relevance of the keywords to the corresponding pages. Finally, retain the keywords relevant to the page content. First, the system concatenates the preset prompt template with the pages generated in Step 4 (page title, page type, page content) to generate a prompt. The system calls Agentic RAG and the general language model, provides the concatenated prompt (for keyword planning or mining), and after receiving the prompt, Agentic RAG and the general language model perform keyword mining and generation according to the requirements and rules and return the keywords to the system. The system receives the content returned by Agentic RAG and the general language model, parses it, and stores it. The system calls Google Keyword Planner to query the search volume of the keywords generated by Agentic RAG and the general language model, filters out the keywords with a search volume of 0, and retains the keywords with a search volume greater than 0, storing them in the website keyword generation module. By using Agentic RAG and the general language model, it is possible to generate and plan keywords, quickly conduct keyword planning and mining, and the system can automatically query the search volume of keywords, significantly reducing the time for keyword planning while ensuring the quality of keyword planning and the relevance of the pages. Step 6: Generate Website Topic Module Utilize Agentic RAG and general language model to plan page topics and content for each page in Step 4 based on the keywords generated for each page in Step 5. First, the system concatenates the preset prompt template with the page generated in Step 4 (page title, page type, page content) and the keywords generated in Step 5 to generate a prompt. Then, the system calls Agentic RAG and the general language model to provide the concatenated prompt (for page topic title and content planning). Next, after receiving the prompt, Agentic RAG and the general language model plan the page, page topic title, and content according to the requirements and rules and return the generated content to the system. Finally, the system receives the content returned by Agentic RAG and the general language model, parses it, and stores it in the generated website topic module. By using Agentic RAG and the general language model for page topic and content planning, it is possible to analyze the current popular topics and content in the relevant field based on big data, thereby increasing the probability that customers are interested in the website topics, ensuring the quality of website topic planning, enhancing the weight of the website and user traffic. At the same time, it can significantly reduce the time required for topic planning and improve the efficiency of building an independent website; Step 7: Page Template Planning Module Utilize Agentic RAG and general language model to plan the templates for each page type of the website based on the competitive product report generated in Step 3 and generate page templates. First, it is necessary to organize and classify the existing components of the page. The system concatenates the preset prompt with the page generated in Step 4 (page title, page type, page content), the keywords generated in Step 5, the page topics generated in Step 6, and the available components of the system to generate a prompt for each page type. Again, the system submits the concatenated prompt to Agentic RAG and the general language model to generate page templates. Then, Agentic RAG and the general language model plan the page type templates based on the content and information in the prompt and return the planned pages to the system. Finally, the system receives the page template data returned by Agentic RAG and the general language model, parses these template data, and stores them in the page template planning module. By using Agentic RAG and the general language model for page template planning, it is possible to greatly improve the efficiency and speed of page template planning; moreover, the planned page templates are planned after analyzing data such as competitive product websites, ensuring the rationality of the planned page templates; Step 8: The website data generation module uses Agentic RAG and a general language model to plan the component content of each template based on the title, content, and topic information of each page, and finally generates an independent website, presenting the effect of independent website construction and decoration on the page. First, the system concatenates the preset prompt with the page generated in Step 4 (page title, page type, page content), the keywords generated in Step 5, the page topics generated in Step 6, and the template information in Step 7 to generate a prompt for each component content. Next, the system submits the concatenated prompt to Agentic RAG and the general language model to generate the page component content. Then, Agentic RAG and the general language model generate the page component content (text, image materials) according to the specified prompt and return this page component content to the system. Finally, the system parses the content returned by Agentic RAG and the general language model, stores the data in the website data generation module, loads the corresponding content, and presents the effect of independent website construction and decoration on the page. By using Agentic RAG and the general language model to generate page content materials, relevant website content can be generated quickly, with high speed and efficiency, meeting the SEO optimization standards and having good effects.

[0023] Embodiment 2: As Figure 2 shown, based on Embodiment 1, the basic information input module includes a website basic information input window and a website basic information database, and the information input through the website basic information input window is stored in the website basic information database.

[0024] Embodiment 3: As Figure 3 shown, based on any one of Embodiments 1 to 2, the competing product site generation module includes an external Agentic RAG and general language model docking port, a competing product site management database, and a competing product site modification and editing window, and the competing product site modification and editing window is used to modify and edit the competing product site to realize manual modification and adjustment of the generated content materials.

[0025] Embodiment 4: As Figure 4 shown, based on any one of Embodiments 1 to 3, the competing product report generation module includes an external Agentic RAG and general language model docking port, a competing product report management database, and a competing product site report modification and editing window, and the competing product site report modification and editing window is used to modify and edit the competing product site report to realize manual modification and adjustment of the generated competing product site report.

[0026] Embodiment 5: AsFigure 5 As shown, based on any one of Embodiments 1 to 4, the website architecture generation module includes an external Agentic RAG and a general language model docking port, a website architecture management database, and a website architecture modification and editing window. The website architecture modification and editing window is used to modify and edit the website architecture, enabling manual modification and adjustment of the generated website architecture.

[0027] Embodiment 6: As Figure 6 shown, based on any one of Embodiments 1 to 5, the website keyword generation module includes an external Agentic RAG and a general language model docking port, a website keyword management database, and a website keyword modification and editing window. The website keyword modification and editing window is used to modify and edit the keywords, enabling manual modification and adjustment of the generated website keywords.

[0028] Embodiment 7: As Figure 7 shown, based on any one of Embodiments 1 to 6, the website topic generation module includes an external Agentic RAG and a general language model docking port, a website topic management database, and a website topic modification and editing window. The website topic modification and editing window is used to modify and edit the website topics, enabling manual modification and adjustment of the generated website topics.

[0029] The entire content generation is automatically generated by Agentic RAG and the general language model based on the SEO standard specifications, and the content can be manually proofread and modified; it can not only quickly generate the material content of the website through Agentic RAG and the general language model, but also enable manual modification and adjustment of the generated content materials; improving efficiency while ensuring the SEO effect of the independent website.

[0030] The above embodiments are only the preferred embodiments of the present application and cannot be used to limit the scope of protection of the present application. Any non-substantive changes and substitutions made by those skilled in the art based on the present application fall within the scope of protection required by the present application.

Claims

1. An independent website building and decoration system based on Agentic RAG and universal language model, characterized by: It includes a basic data entry module connected with Agentic RAG and a universal language model, a module for generating competitive product sites, a module for generating competitive product reports, a module for generating website architecture, a module for generating website keywords, a module for generating website topics, a page template planning module, and a module for generating website data; the basic data entry module is used for users to fill in basic website information; the module for generating competitive product sites is used to mine and generate competitive product sites; the module for generating competitive product reports is used to analyze competitive product sites and obtain competitive product reports on the advantages and disadvantages of the website; the module for generating website architecture is used to generate the architecture of the website; the module for generating website keywords is used to plan and expand keywords for pages within the website; the module for generating website topics is used to plan related topics for pages within the website; the page template planning module is used to plan templates for each page type within the website; The website data generation module is used to generate the component content of each page, and finally generate the page data of the independent website, and present the effect of independent website construction and decoration on the page.

2. The independent website building and decoration system based on Agentic RAG and universal language model according to claim 1 is characterized by: The basic information entry module includes a website basic information entry window and a website basic information database.

3. The independent website building and decoration system based on Agentic RAG and universal language model according to claim 1 is characterized by: The module for generating competitive product sites includes an external Agentic RAG and universal language model docking port, a competitive product site management database, and a competitive product site modification editing window; the module for generating competitive product reports includes an external Agentic RAG and universal language model docking port, a competitive product report management database, and a competitive product site report modification editing window.

4. The independent website building and decoration system based on Agentic RAG and universal language model according to claim 1 is characterized by: The website architecture generation module includes an external Agentic RAG and a universal language model docking port, a website architecture management database, and a website architecture modification editing window.

5. The independent website building and decoration system based on Agentic RAG and universal language model according to claim 1 is characterized by: The module for generating website keywords includes an external AgenticRAG and a universal language model docking port, a website keyword management database, and a website keyword modification editing window; the module for generating website topics includes an external AgenticRAG and a universal language model docking port, a website topic management database, and a website topic modification editing window.

6. An independent website building and decoration method based on Agentic RAG and a universal language model, using an independent website building and decoration system based on Agentic RAG and a universal language model, including a basic data entry module connected with Agentic RAG and a universal language model, a module for generating competitive product sites, a module for generating competitive product reports, a module for generating website architecture, a module for generating website keywords, a module for generating website topics, a page template planning module, and a module for generating website data; characterized in that: The following steps are involved: Step 1: The user enters the basic information of the website through the basic information entry module; Step 2: Generate competitor sites The module uses Agentic RAG and the general language model to mine competitor sites based on the basic information of the website entered in step 1; Step 3: Generate competitive product report module uses Agentic RAG and universal language model to analyze competitive product sites based on SEO optimization standards and rules, and generate competitive product reports for competitive sites; Step 4: Generate website architecture module. Using Agentic RAG and the universal language model, plan the page architecture of the website, including the page title and page content of the header and footer, based on the competitive product report of the competitive product site in step 3 and the basic information entered by the user in step 1; Step 5: Generate website keywords using Agentic RAG and the universal language model, based on the page title and page content generated in step 4, to mine keywords for each page, and compare the relevance of the keywords with the corresponding pages, and finally retain the keywords related to the page content; Step 6: Generate website topic module uses Agentic RAG and universal language model to plan page topic and content for each page in step 4 based on the keywords generated for each page in step 5; Step 7: The page template planning module uses Agentic RAG and the general language model to plan the templates of each page type of the website based on the competitive product report generated in step 3 and generate page templates; Step 8: Generate website data The module uses Agentic RAG and the universal language model to plan the component content of each template based on the title, content, and topic information of each page, and finally generates the page data of the independent station website, and presents the effect of independent station construction and decoration on the page.

7. The independent website building and decoration method based on Agentic RAG and universal language model according to claim 6 is characterized by: The basic information entry module includes a website basic information entry window and a website basic information database, and the information entered in the website basic information entry window is stored in the website basic information database.

8. The independent website building and decoration method based on Agentic RAG and universal language model according to claim 6 is characterized by: The module for generating a competitive site includes an external Agentic RAG and a universal language model docking port, a competitive site management database, and a competitive site modification and editing window, and the competitive site modification and editing window is used to modify and edit the competitive site; the module for generating a competitive report includes an external Agentic RAG and a universal language model docking port, a competitive report management database, and a competitive site report modification and editing window, and the competitive site report modification and editing window is used to modify and edit the competitive site report.

9. The independent website building and decoration method based on Agentic RAG and universal language model according to claim 6 is characterized by: The website architecture generation module includes an external Agentic RAG and a universal language model docking port, a website architecture management database, and a website architecture modification editing window. The website architecture modification editing window is used to modify and edit the website architecture.

10. The independent website building and decoration method based on Agentic RAG and universal language model according to claim 6 is characterized by: The module for generating website keywords includes an external Agentic RAG and universal language model docking port, a website keyword management database, and a website keyword modification and editing window, and the website keyword modification and editing window is used to modify and edit the keywords; the module for generating website topics includes an external Agentic RAG and universal language model docking port, a website topic management database, and a website topic modification and editing window, and the website topic modification and editing window is used to modify and edit the website topics.

Images