Text adaptive matching expansion system based on AI

Abstract

The invention discloses an AI-based text adaptive matching expansion system, which belongs to the technical field of text expansion, and specifically comprises the following steps: acquiring a website URL (Uniform Resource Locator) provided by a user, and acquiring operation area information and HTML (Hypertext Markup Language) content of a website according to the website URL; the method comprises the following steps of: analyzing a webpage through a Request library and a BeautifulSoup library, capturing all links in a webpage page, eliminating empty links and external links, and repeatedly capturing the links for filtering; analyzing the page elements, checking whether the contents of the page elements are blank or unreasonable or not, if so, generating corresponding complementation contents according to the context information, and filling or replacing the contents of the original elements through JavaScript; if not, analysis is continued; and regularly monitoring the ranking and exposure of the user website in the search engine, and adjusting the website according to the monitoring result.

Description

Technical Field

[0001] The present invention relates to the technical field of text expansion, and in particular to an AI-based text adaptive matching and expansion system. Background Art

[0002] Artificial intelligence analysis has a broad prospect and current status in SEO. With the continuous development and application of artificial intelligence technology, the role of AI in SEO is becoming more and more important. In recent years, with the emergence of a series of artificial intelligence technologies such as GPT, the impact of AI in SEO will be more significant. GPT is a natural language processing technology based on deep learning that can automatically generate text and simulate human language behavior, which enables AI to better understand users' search queries, provide more relevant search results, and optimize the ranking of websites.

[0003] A well-developed AI SEO tool is critical to the success of your digital content. Once visitors view your website, SEO keeps them on each page longer, retaining their attention as they learn more about your meaningful solutions. Among the many elements of SEO, page elements such as meta titles and descriptions are the most critical factors for achieving high rankings on search engines such as Google. There are many tools on the market that can give you optimization suggestions, but for most website owners, they can only operate the website simply. There are too many professional modification suggestions that they don’t understand and don’t know how to modify, and they also lack the corresponding development capabilities. The existing SEO methods are based on keyword optimization, but this method has limited effect in overseas markets because users in different countries and regions have large differences in search habits and languages. Summary of the invention

[0004] The purpose of the present invention is to provide an AI-based text adaptive matching and expansion system to solve the following technical problems: For most website owners, they can only perform simple operations on the website. They cannot understand too many professional modification suggestions and do not know how to modify them. They also lack the corresponding development capabilities. The existing search engine optimization method is based on keyword optimization, but this method has limited effect in overseas markets because the search habits and languages ​​of users in different countries and regions vary greatly.

[0005] The purpose of the present invention can be achieved through the following technical solutions: An AI-based text adaptive matching and expansion system, comprising: The data collection module is used to obtain the website URL provided by the user, and obtain the website's operating area information and text content based on the website URL; The text filtering module is used to parse web pages through the Requests library and the BeautifulSoup library, crawl all links in the web pages, exclude empty links and external links, and filter duplicate crawled links; The text completion module is used to analyze page elements and check whether the page element content is blank or unreasonable. If so, it generates corresponding completion text content based on context information and fills or replaces the original element content through JavaScript; if not, it continues to analyze; The normal monitoring module is used to regularly monitor the ranking and exposure of user websites in search engines and adjust the websites according to the monitoring results.

[0006] As a further solution of the present invention: the process of analyzing the page elements is: Identify the text elements of the page, including the title, keywords and introduction content, and check whether the title content, keyword content and introduction content are blank or unreasonable; Identify the image tags on the page, check whether the alt attribute of the image exists, check whether the alt content is reasonable, check whether the image opens normally, and check whether the image size meets the requirements; Identify link tags in the page and check the quantity and quality of internal and external links; Identify the h1 tag in the page and check where and how many h1 tags are used in the page.

[0007] As a further solution of the present invention: the process of generating the corresponding complementary content is: Generate corresponding titles, keywords and introductions based on the content of the article on the web page, and update them to the corresponding text elements on the page; Analyze the context of the image itself and the paragraph it is in, generate accurate alt attribute content, and adjust the image size; Adjust the density of internal links, adjust the number and quality of external links, and adjust the location and number of h1 tags.

[0008] As a further solution of the present invention: the process of updating the title and keywords according to the content of the article is: The text with marked keywords is divided into several words, and the words are mapped into vectors and input into the BERT model after position embedding encoding, and the content features in the text are extracted. The BERT model is back-propagated and updated through the marked keywords until convergence. The web page articles to be processed are input into the BERT model after position embedding encoding, and the content features in the web page articles are extracted. The pending keywords are clustered and generated, and the model is updated by back-propagation using the pending keywords. The pending keywords are obtained after repeated updates for several times, marked as preferred keywords, updated to the corresponding positions of the page, and the corresponding page titles are generated and updated based on the preferred keywords in grammatical order.

[0009] As a further solution of the present invention: the process of updating the introduction content according to the article content is: The text content is divided into paragraphs and sentences, and interference elements are removed, including advertisements, headers, and footers. The content at the beginning of the text is detected to find out whether there are paragraphs with different styles and formats from the rest of the content. Sentences containing specific words in the paragraphs are extracted, including themes, purposes, methods, results, and arguments. The extracted sentences are combined, and the context information is verified using natural language processing tools to form and update the introduction content.

[0010] As a further solution of the present invention: the process of generating the alt attribute of the image is: Get basic information about the image from its metadata, including the title and description, use image recognition tools to automatically identify the image content, generate preliminary description text, preprocess the identified text to remove unnecessary characters, punctuation, and formatting, get contextual information about the image's location, cluster the contextual information with the text, get keyword information, and generate an alt attribute, insert the generated alt attribute into the alt description of the img tag.

[0011] As a further solution of the present invention: analyze the page elements to check whether the content of the page elements is blank or unreasonable. If so, generate corresponding completion content based on the context information, and fill or replace the original element content through JavaScript; if not, continue to analyze; and regularly monitor the ranking and exposure of the user's website in the search engine, and adjust the website according to the monitoring results.

[0012] Beneficial effects of the present invention: The present invention detects web pages, supplements missing or unreasonable elements on the web pages, generates web content that is easy to retrieve and attracts attention, removes some unreasonable links, and uses AI technology to analyze and predict search behaviors and user preferences in overseas markets, thereby optimizing the ranking and exposure of websites in overseas search engines, thereby improving the visibility and traffic of websites in overseas markets. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention will be further described below in conjunction with the accompanying drawings.

[0014] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0016] See also Figure 1 As shown, the present invention is an AI-based text adaptive matching and expansion system, comprising: The data collection module is used to obtain the website URL provided by the user, and obtain the website's operating area information and text content based on the website URL; The text filtering module is used to parse web pages through the Requests library and the BeautifulSoup library, crawl all links in the web pages, exclude empty links and external links, and filter duplicate crawled links; The text completion module is used to analyze page elements and check whether the page element content is blank or unreasonable. If so, it generates corresponding completion text content based on context information and fills or replaces the original element content through JavaScript; if not, it continues to analyze; The normal monitoring module is used to regularly monitor the ranking and exposure of user websites in search engines and adjust the websites according to the monitoring results.

[0017] A well-developed AI SEO tool is critical to the success of your digital content. Once visitors view your website, SEO keeps them on each page longer, retaining their attention as they learn more about your meaningful solutions. Among the many elements of SEO, page elements such as meta titles and descriptions are the most critical factors for achieving high rankings on search engines such as Google. There are many tools on the market that can give you optimization suggestions, but for most website owners, they can only operate the website simply. There are too many professional modification suggestions that they don’t understand and don’t know how to modify, and they also lack the corresponding development capabilities. The existing SEO methods are based on keyword optimization, but this method has limited effect in overseas markets because users in different countries and regions have large differences in search habits and languages.

[0018] The present invention uses AI technology to analyze and predict search behaviors and user preferences in overseas markets, thereby optimizing the ranking and exposure of websites in overseas search engines, thereby improving the visibility and traffic of websites in overseas markets.

[0019] In a preferred embodiment of the present invention, the process of analyzing the page elements is as follows: Identify the text elements of the page, including the title, keywords and introduction content, and check whether the title content, keyword content and introduction content are blank or unreasonable; Identify the image tags on the page, check whether the alt attribute of the image exists, check whether the alt content is reasonable, check whether the image opens normally, and check whether the image size meets the requirements; Identify link tags in the page and check the quantity and quality of internal and external links; Identify the h1 tag in the page and check where and how many h1 tags are used in the page.

[0020] In another preferred embodiment of the present invention, the process of generating the corresponding completion content is: Generate corresponding titles, keywords and introductions based on the content of the article on the web page, and update them to the corresponding text elements on the page; Analyze the context of the image itself and the paragraph it is in, generate accurate alt attribute content, and adjust the image size; Adjust the density of internal links, adjust the number and quality of external links, and adjust the location and number of h1 tags.

[0021] In another preferred embodiment of the present invention, the process of updating the title and keywords according to the article content is as follows: The text with marked keywords is divided into several words, and the words are mapped into vectors and input into the BERT model after position embedding encoding, and the content features in the text are extracted. The BERT model is back-propagated and updated through the marked keywords until convergence. The web page articles to be processed are input into the BERT model after position embedding encoding, and the content features in the web page articles are extracted. The pending keywords are clustered and generated, and the model is updated by back-propagation using the pending keywords. The pending keywords are obtained after repeated updates for several times, marked as preferred keywords, updated to the corresponding positions of the page, and the corresponding page titles are generated and updated based on the preferred keywords in grammatical order.

[0022] In another preferred embodiment of the present invention, the process of updating the introduction content according to the article content is as follows: The text content is divided into paragraphs and sentences, and interference elements are removed, including advertisements, headers, and footers. The content at the beginning of the text is detected to find out whether there are paragraphs with different styles and formats from the rest of the content. Sentences containing specific words in the paragraphs are extracted, including themes, purposes, methods, results, and arguments. The extracted sentences are combined, and the context information is verified using natural language processing tools to form and update the introduction content.

[0023] In another preferred embodiment of the present invention, the process of generating the alt attribute of an image is as follows: Get basic information about the image from its metadata, including the title and description, use image recognition tools to automatically identify the image content, generate preliminary description text, preprocess the identified text to remove unnecessary characters, punctuation, and formatting, get contextual information about the image's location, cluster the contextual information with the text, get keyword information, and generate an alt attribute, insert the generated alt attribute into the alt description of the img tag.

[0024] Analyze page elements to check whether the content of page elements is blank or unreasonable. If so, generate corresponding completion content based on context information and fill or replace the original element content through JavaScript; if not, continue to analyze; and regularly monitor the ranking and exposure of user websites in search engines, and adjust the website according to the monitoring results.

[0025] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A text adaptive matching and expansion system based on AI, characterized in that: include: The data collection module is used to obtain the website URL provided by the user, and obtain the website's operating area information and text content based on the website URL; The text filtering module is used to parse web pages through the Requests library and the BeautifulSoup library, crawl all links in the web pages, exclude empty links and external links, and filter duplicate crawled links; The text completion module is used to analyze page elements and check whether the page element content is blank or unreasonable. If so, it generates corresponding completion text content based on context information and fills or replaces the original element content through JavaScript; if not, it continues to analyze; The normal monitoring module is used to regularly monitor the ranking and exposure of user websites in search engines and adjust the websites according to the monitoring results.

2. According to claim 1, the text adaptive matching and expansion system based on AI is characterized in that: The process of analyzing page elements is: Identify the text elements of the page, including the title, keywords and introduction content, and check whether the title content, keyword content and introduction content are blank or unreasonable; Identify the image tags on the page, check whether the alt attribute of the image exists, check whether the alt content is reasonable, check whether the image opens normally, and check whether the image size meets the requirements; Identify link tags in the page and check the quantity and quality of internal and external links; Identify the h1 tag in the page and check where and how many h1 tags are used in the page.

3. According to the AI-based text adaptive matching and expansion system of claim 1, it is characterized in that: The process of generating the corresponding completion content is as follows: Generate corresponding titles, keywords and introductions based on the content of the article on the web page, and update them to the corresponding text elements on the page; Analyze the context of the image itself and the paragraph it is in, generate accurate alt attribute content, and adjust the image size; Adjust the density of internal links, adjust the number and quality of external links, and adjust the location and number of h1 tags.

4. The text adaptive matching and expansion system based on AI according to claim 3 is characterized in that: The process of updating the title and keywords according to the content of the article is: The text with marked keywords is divided into several words, and the words are mapped into vectors and input into the BERT model after position embedding encoding, and the content features in the text are extracted. The BERT model is back-propagated and updated through the marked keywords until convergence. The web page articles to be processed are input into the BERT model after position embedding encoding, and the content features in the web page articles are extracted. The pending keywords are clustered and generated, and the model is updated by back-propagation using the pending keywords. The pending keywords are obtained after repeated updates for several times, marked as preferred keywords, updated to the corresponding positions of the page, and the corresponding page titles are generated and updated based on the preferred keywords in grammatical order.

5. The text adaptive matching and expansion system based on AI according to claim 3 is characterized in that: The process of updating the introduction content according to the article content is: The text content is divided into paragraphs and sentences, and interference elements are removed, including advertisements, headers, and footers. The content at the beginning of the text is detected to find out whether there are paragraphs with different styles and formats from the rest of the content. Sentences containing specific words in the paragraphs are extracted, including themes, purposes, methods, results, and arguments. The extracted sentences are combined, and the context information is verified using natural language processing tools to form and update the introduction content.

6. The text adaptive matching and expansion system based on AI according to claim 3, characterized in that: The process of generating an image alt attribute is: Obtain basic information about the image from the image metadata, including the title and description, use image recognition tools to automatically identify the image content, generate preliminary description text, preprocess the identified description text to remove unnecessary characters, punctuation marks and formats, and obtain contextual information about the location of the image, cluster the contextual information with the description text, obtain the core key information after clustering, generate an alt attribute based on the key information, and insert the generated alt attribute into the alt description of the img tag.

7. The text adaptive matching and expansion system based on AI according to claim 3, characterized in that: include: Analyze page elements to check whether the content of page elements is blank or unreasonable. If so, generate corresponding completion content based on context information and fill or replace the original element content through JavaScript; if not, continue to analyze; and regularly monitor the ranking and exposure of user websites in search engines, and adjust the website according to the monitoring results.

Images