Article page generation methods, devices, storage media, and program products

Abstract

Embodiments of the present application provide an article page generation method and device, a storage medium and a program product. In the embodiments of the present application, a user can search for a search result item containing a first target article page published by a target platform based on a search word, and jump to the first target article page based on the search result item pointing to the first target article page. Further, the first target article includes multiple display areas, the first area displays multi-modal article content generated by an artificial intelligence-based AI multi-modal generation model related to the search word, and the entry information of the second area can enter a knowledge square page, and the user can continue to view any article page under any category information through the knowledge square page. The user can view more related article pages through the searched article page, and such a page structure can provide the user with more rich and user-interested related information, meet the user's information needs, and thus attract user traffic.

Description

Technical Field

[0001] This application relates to the field of computer processing technology, and in particular to a method, apparatus, storage medium, and program product for generating article pages. Background Technology

[0002] With the continuous development of mobile internet technology, attracting user traffic by publishing relevant articles on service platforms has become a mainstream trend. Taking a housekeeping platform recruiting users for its partner merchants as an example, the platform can generate promotional articles based on the merchants' personalized information and publish these articles as web pages to attract user traffic.

[0003] Currently, users can only obtain limited information through promotional articles, which fails to meet their information needs. Furthermore, the content of these articles is relatively simple and remains unchanged after publication. If an article has been published for a long time, it may no longer meet users' current information needs, resulting in unsatisfactory user acquisition effects based on promotional articles. Summary of the Invention

[0004] This application provides a method, device, storage medium, and program product for generating article pages, in order to solve the problem of unsatisfactory user traffic generation results based on promotional articles and improve user traffic generation effectiveness.

[0005] This application also provides an article page generation method, including: periodically acquiring popular search terms and popular comments; determining hot topics in the current period based on popular search terms and popular comments, with each hot topic belonging to a category; if the hot topics in the current period have changed relative to the hot topics in the previous period, acquiring historical article content related to the changed target hot topic; inputting the target hot topic, the category information to which the target hot topic belongs, and the historical article content into a multimodal generation model, and generating multimodal content for the target hot topic and related historical article content under the same category information to obtain the article content corresponding to each category information.

[0006] This application also provides an electronic device, including: a memory and a processor; the memory for storing a computer program; and the processor, coupled to the memory, for executing the computer program to implement the steps in the methods described above.

[0007] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps in the methods described above.

[0008] This application also provides a computer program product, which includes a computer program / instructions that, when executed by a processor, enable the processor to implement the steps in the methods described above.

[0009] In this embodiment, based on periodically acquiring popular search terms and comments corresponding to at least one category of information, an AI multimodal generation model is used to generate article content for each category. This content is periodic and represents popular topics within the current period. Furthermore, the generated content is rich in multimodal comparisons, satisfying users' information needs within the current period and attracting user traffic. Additionally, based on the target page structure, target article pages corresponding to each category can be obtained. For example, if the page structure includes a first area and a second area, the article content for each category can be added to the first area of ​​the target page structure, while the second area includes entry information pointing to a knowledge square page. Users can then view more related article pages through a searched article page. This page structure provides users with richer and more interesting information, satisfying their information needs and attracting user traffic. Attached Figure Description

[0010] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0011] Figure 1 A schematic diagram of the structure of an information processing system provided in an exemplary embodiment of this application;

[0012] Figure 2a A schematic diagram of a page provided for an exemplary embodiment of this application;

[0013] Figure 2b A schematic diagram of another page provided for another exemplary embodiment of this application;

[0014] Figure 2c A schematic diagram of yet another page provided for yet another exemplary embodiment of this application;

[0015] Figure 2d A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0016] Figure 2e A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0017] Figure 2f A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0018] Figure 2g A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0019] Figure 2h A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0020] Figure 2i A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0021] Figure 2j A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0022] Figure 2k A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0023] Figure 2l A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0024] Figure 2m A schematic diagram of yet another page provided as yet another exemplary embodiment of this application;

[0025] Figure 3 A flowchart illustrating an exemplary embodiment of this application for generating an article page;

[0026] Figure 4 A schematic diagram of the structure of an electronic device provided for an exemplary embodiment of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0028] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0029] To address the technical problem of unsatisfactory user acquisition effects based on promotional articles, this application embodiment addresses this issue by periodically acquiring popular search terms and comments corresponding to at least one category of information. The embodiment then utilizes an AI-based multimodal generation model to generate article content for each category, ensuring its periodicity and reflecting popular topics within the current period. Furthermore, the generated content is rich in multimodal comparisons, satisfying users' information needs within the current period and thus attracting user traffic. Additionally, based on the target page structure, target article pages corresponding to each category can be obtained. For example, if the page structure includes a first area and a second area, the article content for each category can be added to the first area of ​​the target page structure, while the second area includes entry information pointing to a knowledge plaza page. Users can then view more related article pages through a search result. This page structure provides users with richer and more interesting information, satisfying their information needs and attracting user traffic.

[0030] The following describes in detail a solution provided by an embodiment of this application, with reference to the accompanying drawings.

[0031] Figure 1 This is a schematic diagram of the structure of an information processing system provided for an exemplary embodiment of this application. (See diagram below.) Figure 1 As shown, the system includes: a user terminal 101 and a target platform 102.

[0032] In this embodiment, a user terminal refers to a device used by a user to access internet information resources; that is, a user terminal can serve as an entry point for users to interact with internet content and services. This embodiment does not limit the implementation form of the user terminal. For example, a user terminal can be a smart handheld device, such as a smartphone, tablet, laptop, or desktop computer, etc.; another example is a smart wearable device, such as a smartwatch, smart bracelet, etc.; yet another example is various smart home appliances with display screens, such as smart TVs, smart large screens, or smart robots, etc. Furthermore, various applications can be installed on the user terminal. Applications can be standalone apps or mini-programs that depend on standalone apps; there is no limitation on this. For example, applications can be browser applications (referred to as browsers), smart search applications, service applications (used to provide various information, such as service information), etc. Among these, service applications can be, for example, home service applications, real estate transaction applications, shopping applications, or comprehensive applications that combine the functions of these applications.

[0033] A browser is an application used to retrieve, present, and browse information resources transmitted over the internet, primarily referring to web page information resources. Users access these resources through a browser running on their terminal. The terminal provides the hardware platform, while the browser provides the software functionality; together, they provide a complete internet experience. In practice, users can request web pages from a server by entering the URL or using a search engine, and then receive the returned data. It's important to note that this server can be a browser server or the server of other applications (such as home service applications). Furthermore, the browser can parse this data (such as HTML, CSS, and JavaScript) and render it into an interactive web page interface on the user's terminal. Alternatively, after obtaining the web page's address by entering the URL or using a search engine, the user can be redirected to the corresponding web page.

[0034] A search engine is an online tool or service that allows users to quickly find, retrieve, and access information on the internet. In other words, a browser can act as a host application, providing a platform for users to access and browse internet resources. A search engine can be embedded within the browser as a component or function, allowing users to directly access the search engine page from the browser; alternatively, users can enter the search engine's address into the browser to find the search engine, enabling them to directly search for information based on their search terms through the search engine page displayed in the browser. Regardless of the method used for information retrieval, the search engine relies on the browser to provide the user interface and network connection, allowing users to enter search terms and view search results. The browser, in turn, relies on the search engine to provide search functionality, enhance the user experience, and enable users to quickly find information matching their search terms.

[0035] In practical applications, browsers can typically set a default search engine, and users can also customize and select other search engines built into the browser. Specifically, users can directly enter search terms into the browser's address bar or search bar. The browser will then forward these search terms to the default or user-defined search engine for information retrieval. The search results are then displayed to the user through the browser, which is responsible for rendering the search results page. Furthermore, browsers can be linked to other business platforms (such as target platforms) so that any search engine built into the browser can retrieve information matching the search terms from associated platforms; or, any search engine built into the browser can be linked to other platforms so that when a user uses any search engine within the browser for information retrieval, information matching the search terms can be retrieved from associated platforms.

[0036] In this embodiment, the search engine can also be implemented as a standalone application or a mini-program that depends on an application to run, which can be called an intelligent search application. That is, an intelligent search application provides both a user interface and network connectivity so that users can enter search terms and view search results, and also provides information search functionality. Furthermore, the intelligent search application can be associated with other business platforms (such as target platforms) so that when users conduct information searches, they can find information matching their search terms from associated platforms.

[0037] In addition to browsers and smart search applications with built-in search engines that can provide information search functions, service applications can also provide information search pages and information search functions. Furthermore, service applications can publish various types of service information, and users can search for information based on search terms within service applications to obtain search results.

[0038] In this embodiment, the target platform is a server for an application installed on a user terminal, such as any service application, or it can be a server shared by multiple applications. This application does not limit the type of application; for example, the application can be a service application, such as a housekeeping service application, a real estate transaction application, a shopping application, or a comprehensive application that integrates the functions of these applications. This embodiment does not limit the implementation form of the target platform; for example, the target platform can be a physical server, a cloud server, a server cluster, etc.

[0039] In this embodiment, regardless of which method is used for information search, the user terminal will display a search results page corresponding to the search terms. For ease of description and distinction, this search results page can be referred to as the first page. The first page can be a search results page generated by the user's own search operation, or a search results page generated by others and shared with the user; this embodiment does not limit this. The generation process of the first target article page and other target article pages in the following embodiments can be found in the relevant descriptions of the following embodiments, and will not be repeated here.

[0040] In this embodiment, the first page includes at least one result search item that matches the search term and points to at least one article page. The at least one article page includes an article page published by the target platform, referred to as the first target article page. Furthermore, this embodiment does not limit the implementation form of the at least one search result item; for example, the at least one result search item can be an interactive interface element such as a card, link, title, or video cover.

[0041] Taking a user's information search through a target search engine associated with a target platform as an example, the target search engine can be associated with at least one business platform (such as the target platform). When a user uses the target search engine, it can search for information matching the search terms from multiple associated platforms as search results. Based on this, displaying the first page can be achieved as follows: In response to a user's search operation on the target search engine's search page, based on the user's input search terms, a search is performed on the text pages provided by the business platforms associated with the target search engine to obtain at least one article page matching the search terms; wherein, the business platform includes the target platform; further, based on the address information of at least one article page, at least one search result item pointing to at least one article page is generated, and at least one search result item is displayed on the first page. The address information of the article page can be the Uniform Resource Locator (URL) of that page. The URL can be considered a string identifying the location of the article page; based on this string, the user can access and obtain the content of that page through the target search engine.

[0042] This embodiment does not limit the method of obtaining the address information of at least one article page. For example, before the target search engine performs a search on a text page provided by an associated business platform, the associated business platform may pre-send the address information of at least one article page to the target search engine, so that the target search engine can obtain the corresponding page data based on the address information of at least one article page, or the target search engine may directly jump to the corresponding page based on the address information of at least one article page. As another example, in response to a user's search operation on the target search engine's search page, the target search engine obtains the address information of at least one article page matching the search term from the associated business platform, so that the target search engine can obtain the corresponding page data based on the address information of at least one article page, or the target search engine may directly jump to the corresponding page based on the address information of at least one article page.

[0043] Based on the above-mentioned method of obtaining the address information of at least one article page, the above-mentioned response to a user's search operation initiated on the search page of the target search engine, and the search performed on the text pages provided by the business platform associated with the target search engine according to the search terms entered by the user, in order to obtain at least one article page matching the search terms, includes at least the following methods:

[0044] Method 1: The business platform associated with the target search engine sends the address information of at least one article page to the target search engine in advance; in response to the user's search operation on the search page, the target search engine reads the article content of at least one article page corresponding to the address information of at least one article page from the associated business platform; if the article content matches the search term, it obtains the article page corresponding to the article content to obtain at least one article page that matches the search term.

[0045] Method 2: The business platform associated with the target search engine pre-sends the address information of at least one article page to the target search engine, and the address information of at least one article page carries corresponding category information, which indicates the category to which at least one article page belongs in the business platform associated with the target search engine; In response to the search operation initiated by the user on the search page, the target search engine determines the target category information matching the search term from the category information carried by the address information of at least one article page; Based on the at least one target address information corresponding to the target category information, it retrieves at least one article page corresponding to at least one target address information from the business platform associated with the target search engine to obtain at least one article page matching the search term.

[0046] Optionally, the target search engine may be associated with multiple business platforms, and the address information of at least one article page may also carry the platform identifier of the business platform to which it belongs. In response to the search operation initiated by the user on the search page, the target search engine determines the target category information that matches the search term and its corresponding at least one target address information from the category information carried by the address information of at least one article page; based on the platform identifier carried by the at least one target address information and the at least one target address information, it obtains at least one article page corresponding to the at least one target address information from the business platform corresponding to the corresponding platform identifier, so as to obtain at least one article page that matches the search term.

[0047] Method 3: The target search engine responds to the user's search operation on the search page, and searches the text pages provided by the business platform associated with the target search engine based on the search terms entered by the user, and obtains the article pages whose content matches the search terms, so as to obtain at least one article page that matches the search terms.

[0048] This application does not limit the implementation of generating at least one search result item pointing to at least one article page based on the address information of at least one article page, but depends on the implementation form of the search result item. At least one search result item can be an interactive interface element such as a card, link, title, or video cover. For example, when the search result item is implemented as a card, an interactive card can be generated based on at least one of the following: the address information of each article page and the theme, core content, or image content of the article content on that article page. The cover of the card can display at least one of the theme, core content, or image content of the corresponding article page. As another example, when the search result is implemented as a link, a corresponding interactive link can be generated based on at least one of the following: the address information of each article page and the theme or core content of the article content on that article page. The display format of the link includes, but is not limited to, "theme / core content" or "theme / core content + corresponding article page address information". Alternatively, an interactive link can be generated directly based on the address information of each article page. For example, if the article content on an article page includes videos and the search results are implemented as video covers, an interactive video cover image can be generated based on the video cover in the article content and the address information of the article page. It should be noted that the above is merely illustrative and does not constitute a limitation on this application.

[0049] Figure 2aThis is an exemplary diagram of the first page corresponding to the search term "furniture recycling". The first page includes a search box and multiple search result items. The search box displays the search term "furniture recycling", and the multiple search result items include search result items pointing to the first target article page published on the target platform. Users can view the first target article page through these search result items. It should be noted that the diagrams in the context of this application are merely illustrative and do not constitute a limitation on the technical solution of this application.

[0050] In this embodiment of the application, in response to a user's triggering operation on a search result item pointing to a first target article page, the user can be redirected to the first target article page, thereby allowing the user to view the article content on the first target article page to meet the user's information needs.

[0051] It should be noted that the user interaction operations, access operations, or triggering operations involved in the context of this application include, but are not limited to, various interaction operations such as touch operations, gesture operations, voice operations, head movement operations, and eye movement operations; among them, touch operations include, but are not limited to, click operations, double-click operations, long-press operations, swipe operations, pinch operations, or mouse hover operations. Swipe operations include, but are not limited to, straight-line swipes and curved-line swipes. Furthermore, it should be noted that when the context of this application involves jumping between two pages, the jumping methods involved in the embodiments of this application include, but are not limited to, jumping directly from one page to another page, or jumping from one page to a task interface and completing the corresponding task operation on the task interface before jumping to another page. Therefore, in response to the user's triggering operation on the search result item pointing to the first target article page, when jumping to the first target article page, one can jump directly from the first page to the first target article page, or jump from the first page to a task interface and complete the task operation such as identity verification on the task interface before jumping to the first target article page.

[0052] It should be noted that if a user searches for information using the target search engine on a PC, the first target article page may appear as follows: Figure 2b The webpage shown includes a first target article page and a QR code for the first target article page. Users can scan the QR code using mobile devices such as smartphones and smart wearable devices to directly display the first target article page on their smartphones or smart wearable devices, allowing them to view the article content on their mobile devices.

[0053] In this embodiment, to further meet user information needs and attract user traffic, merchant information can be embedded in the first target article page, allowing users to learn more about the merchant's details and leave their contact information. The merchant information embedded in the first target article page refers to merchant information that can provide services or goods within the category to which the first target article page belongs. In other words, the merchant information embedded in the first target article page is merchant information that matches user needs. Matching user needs can be understood as being able to provide the user with the required services or goods, and the service or goods being provided within the user's desired timeframe. This embodiment does not limit the implementation form of the embedded first target merchant information. For example, the first target merchant information can be interactive interface elements such as cards, links, titles, and video covers. Different implementation forms of first target merchant information may correspond to different generation methods. For details, please refer to the detailed description in the above embodiment of generating at least one search result item pointing to at least one article page based on the address information of at least one article page; it will not be repeated here.

[0054] In this embodiment, to more accurately recommend merchant information that can provide services or goods to users, merchant information that can provide services or goods to the user's current location can be selected from the merchant information that provides services or goods under the category to which the first target article page belongs, referred to as the first target merchant information. This first target merchant information is then embedded into the first target article page. The first target merchant information is thus the merchant information that can provide services or goods under the category to which the first target article page belongs. Based on this, responding to the user's trigger operation on a search result item pointing to the first target article page and redirecting to the first target article page includes at least two of the following implementation methods:

[0055] Method 1: Respond to the user's trigger action on the search results item pointing to the first target article page, obtain the user's current location information; send the current location information to the target platform so that the target platform can embed merchant information adapted to the current location information into the first target article page before returning to the first target article page. This merchant information can be referred to as the first target merchant information; display the first target article page returned by the target platform. The first target article page includes the first target merchant information.

[0056] Method 2: Respond to the user's trigger action on the search results item pointing to the first target article page, obtain the user's current location information, and send the browsing request for the first target article page and the user's current location to the target platform. The target platform then embeds merchant information adapted to the current location information into the first target article page. After the target platform agrees to the browsing request, the target search engine's target search results page redirects to the target platform's first target article page. Figure 1(This is a schematic diagram using this process as an example.)

[0057] This embodiment does not limit the implementation method of sending the current location information to the target platform as described above. Optionally, sending the current location information to the target platform includes at least the following implementation methods:

[0058] Method 1: In response to the user's trigger action on the search result item pointing to the first target article page, send a request to the target platform to obtain the first target article page; in response to the request to obtain the first target page, the target platform sends a request to the target search engine to obtain the user's location information; in response to the location information request, obtain the user's current location information and send the user's current location information to the target platform.

[0059] Method 2: Respond to the user's trigger action on the search results item pointing to the first target article page, obtain the user's current location information, and send a request to the target platform to obtain the first target article page and the user's current location information.

[0060] In addition, the user's current location information can also be obtained by the target platform. Specifically, the target platform obtains the user's current location information in response to a request to retrieve or browse the first target article page sent by the target search engine.

[0061] Correspondingly, after obtaining the user's current location information, the target platform can obtain merchant information that can provide services or products under the category information to which the first target article page belongs, based on the category information to which the first target article page belongs; select target merchant information that matches the current location information from the merchant information based on the current location information; embed the target merchant information into the first target article page, and then send the first target article page to the user terminal for display to the user terminal, or for the target search engine's target result page to jump to the target platform's first target article page.

[0062] In this embodiment, the first target article page has a page structure, referred to as the target page structure. The page structure refers to the organization and layout of the page content, defining the position and relationship of each element on the page. The target page structure can be implemented using a target page template. Specifically, the first target article page includes: a first area, where article content related to the search term is displayed, referred to as the first article content. The first article content is the article content corresponding to the category information matched with the search term. The first article content includes, but is not limited to, at least one of: image content, article theme, and text content. The image content can be one or more images. If there are multiple images, each image can be switched between, for example, by swiping left or right within the same area. The article theme is the topic content related to the category information matched with the search term. The text content includes at least one hot topic and topic content under that hot topic, with different hot topics and their topic content located in independent sub-areas within the first area. Taking furniture recycling as an example, the hot topics corresponding to furniture recycling could be: "What can be recycled?", "What is the recycling price?", "Will you accept a piece of furniture?", etc.

[0063] In this embodiment, the first article content can be multimodal content generated by the target platform's AI multimodal generation model. Multimodal content refers to content in multiple forms. For example, the first article content includes, but is not limited to, the article title, image content, and text content for each category. Furthermore, there is certain layout information between content of the same modality and content of different modalities. Layout information may include, for example, the size, color, and position of content of the same modality, as well as the relative position information between content of different modalities. Multimodal content can provide users with richer and more relevant information, meeting their information needs and thus attracting user traffic.

[0064] The AI-based multimodal generation model can be jointly trained using a large language model (LLM) and a text-to-graph model. The large language model possesses deep semantic understanding and generation capabilities across modalities and languages. In generating multimodal article content, at least one method can be employed, such as natural language processing, searching a database, and / or content creation based on semantic understanding. Large language models include, but are not limited to, GLM (Generated Linear Model) series models and Qwen series models. The text-to-graph model is used to output an image that matches the prompt words based on the input text content after a certain number of iterations. This embodiment does not limit the type of text-to-graph model; for example, it can be an SD (Stabular Diffusion) model. Additionally, when selecting an SD model, one can choose a model that has been pre-trained and exhibits generalization ability, is adapted to the user's ethnicity, and generates high-resolution images, such as the Staple Diffusion XL (SDXL) model. The SDXL model is a newer version of the SD model series, with significant improvements and optimizations in image generation quality. The prompts include at least positive and negative prompts. Positive prompts are textual descriptions used to guide the model in generating specific types or styles of high-quality images. Negative prompts are textual descriptions used to guide the model in avoiding the generation of certain features or elements. Negative prompts, in contrast to positive prompts, work together in the model's generation process to achieve more precise control over the elements in the image.

[0065] For details on how to generate the first article content using an AI-based multimodal generation model, please refer to the relevant descriptions in the following embodiments, which will not be repeated here.

[0066] Figure 2cThis is an exemplary diagram of the first area of ​​the first target article page, assuming a user search for articles related to "furniture recycling". The first sub-area of ​​the first area can display at least one image, and the second sub-area can display the article title and text content. The article title could be, for example: "Sisters, Beijing furniture recycling, 30-minute pick-up, even one piece!" The text content could be, for example: "Recently, many sisters have asked us questions about recycling. Today, we've compiled these questions for everyone, selecting a few frequently asked questions for a unified answer. 1. 'What can we recycle?': Currently, we can recycle items including air conditioners, refrigerators, water heaters, televisions, computers, sofas, beds, cabinets, tables and chairs, office furniture, etc. The prerequisite is that these appliances and furniture still have some value; if they have no value at all, they naturally cannot be recycled. 2. 'How much is the recycling price?': Recycling quotes are a relatively complex issue, requiring assessment based on the condition of the furniture and appliances themselves, the purchase price, and market conditions. Therefore, a fixed price cannot be given; a quote can only be given after specific information is obtained. Quotes are usually given online. 3. 'Do you buy one piece of furniture?'" "First, it's important to acknowledge that even a single piece of furniture can be collected. However, if your furniture is of low value and not worth the merchant's trip, you might need to wait until other customers in your neighborhood need their furniture recycled, and then they can handle it together. This is beneficial for both parties!" Additionally, the first area also includes layout information. The first sub-area can be at the top of the first area, and multiple images within it can be switched. The second sub-area is below the first, and the article title within it can be at the top, with the text content below. Different topics within the text content have corresponding sub-areas with different topic content. Furthermore, the first area can also include the author's username and avatar, and this personal information can be displayed anywhere within the first area, such as the top left corner. It should be noted that the layout information for the first area is not limited to this.

[0067] This application does not limit the specific location where the first target merchant information is embedded on the first article page. For example, the first target merchant information can be embedded in a first sub-region or a second sub-region of the first area. Figure 2c As shown, the first target merchant information is also embedded in a third sub-area within the first area. This third sub-area can be located at the bottom of the first area, below the second sub-area. Therefore, after users understand the specific content related to the search term from the first article content, they can immediately view the first target merchant information, allowing them to further understand the merchant's details and leave their contact information. The embedded first target merchant information includes, but is not limited to: the merchant's store name, store rating, and store ranking.

[0068] Furthermore, such as Figure 2c Therefore, the first target merchant information can also set the lead generation method and its corresponding lead generation control, referred to as the first lead generation method and its corresponding first lead generation control, for users to initiate lead generation operations. The first lead generation method includes at least one of online consultation, audio / video consultation, and appointment consultation. Additionally, the first area also includes a page return control, such as... Figure 2c The "<" control in the text can return to the previous page in response to a triggering operation on the control.

[0069] Furthermore, in response to a user's action to leave a contact using any of the first contact leaving controls, the user's contact information can be obtained. Specifically, obtaining the user's contact information in response to a user's action to leave a contact using any of the first contact leaving controls includes the following implementation methods:

[0070] Method 1: For example Figure 2d , Figure 2e , Figure 2f As shown, when the lead generation method is appointment consultation, in response to the user's triggering operation on the first lead generation control corresponding to the appointment consultation lead generation method, a lead generation information collection page is displayed. This page includes category information items and user information items. User information items include, but are not limited to, at least one of city information items and contact information items. In response to the user's triggering operation on the category information items, a category list is displayed, which includes multi-level categories. In response to the user's selection operation on each level of category, the selected level of category is displayed in the category information item. Furthermore, in response to the user's selection / filling operation on the user information items, the selected / filled user information is displayed in the user information item. Additionally, the category information items and user information items have attribute information, including either required or optional fields. Required attributes mean that the information item must be selected / filled, while optional attributes mean that the information item can be selected / filled selectively. If a required information item is not selected / filled, a prompt message will be sent to the user indicating that no information has been selected / filled.

[0071] This embodiment does not limit the implementation form of the lead generation page. For example, the lead generation page can be a separate page; correspondingly, in response to the user's trigger operation on the first lead generation control corresponding to the appointment consultation method, the user is redirected to the lead generation page. As another example, the lead generation page can also be a pop-up page; correspondingly, in response to the user's trigger operation on the first lead generation control corresponding to the appointment consultation method, the pop-up page is displayed on top of the first target article page. Furthermore, the pop-up page has an initial size, which can be smaller than the screen size; furthermore, in response to the user's enlargement operation on the pop-up page, the pop-up page is gradually enlarged to full-screen display.

[0072] Optionally, in response to a user's triggering action on a category information item, a category list can be displayed, including: responding to a user's triggering action on a category information item by redirecting to a category information page, which includes a category information list; or, responding to a user's triggering action on a category information item by displaying a category information list within the lead generation information page or in an associated area of ​​the lead generation information page. It should be noted that category information items may also have a manually entered attribute; in this case, in response to a user's input action on a category information item, the user-entered category information is retrieved and displayed in the category information item. Furthermore, the category information page also includes a return control, used to return to the lead generation information collection page. The return control can be, for example, "<", and in response to a user's triggering action on the return control, the lead generation information collection page is displayed.

[0073] Optionally, in response to a user's selection / filling operation on a user information item, the selected / filled user information is displayed in the user information item. This includes: if the user information item is a city information item, in response to a user's trigger operation on the city information item, redirecting to a city information page, where multiple city information items and their corresponding controls are displayed; or, in response to a user's trigger operation on a city information item, displaying a list of city information items within the data collection information page or in an associated area of ​​the data collection information page, where the city information list includes multiple city information items and their corresponding controls. Further, in response to a selection operation on any city information control, the corresponding city information is retrieved and displayed in the city information item. Further, the city information page also includes a return control, used to return to the data collection information page. The return control can be, for example, "<", in which case, in response to a user's trigger operation on the return control, the user returns to the data collection information page.

[0074] Figure 2dThis is an exemplary diagram of a contact information collection page for the recycling category (also known as a contact information pop-up page). In this diagram, the contact information collection page is displayed as a pop-up page above the first article page. This contact information collection page includes recycling category information, city information, and audio / video information, all of which are required fields. Responding to the user's trigger action on the recycling category information, the category information page is displayed. The category information page includes two levels of category information: primary category information and secondary category information, which are displayed in different sub-areas of the category information page. The system is divided into two sub-areas. The first sub-area displays the primary categories and their corresponding controls. Primary category information includes: home appliances, furniture, hotel equipment, kitchen equipment, stationery, beauty and personal care equipment, and industrial metals. Taking "home appliances" as an example, in response to user interaction with the corresponding control, "home appliances" is displayed in the category information item. The second sub-area then displays the secondary category information under the "home appliances" category and its corresponding controls. For example, if the secondary category information is "refrigerator," the secondary category information includes: refrigerator, washing machine, air conditioner, television, other small appliances, and multiple appliances. Further, using "refrigerator" as an example, in response to user interaction with the corresponding control, "refrigerator" is displayed in the category information item. The information displayed in the category information item can be either "home appliances - refrigerator" or simply "refrigerator." Finally, in response to user interaction with the "<" control, the system returns to the lead generation page. The returned lead generation page displays either "home appliances - refrigerator" or simply "refrigerator" in the category information item.

[0075] Method 2: For example Figure 2g As shown, when the lead generation method is online consultation, in response to the user's trigger operation on the first lead generation control corresponding to the online consultation lead generation method, the online consultation page is displayed. The online consultation page includes a content input control. In response to the user's content input operation initiated through the content input control, the user's lead generation information is obtained. Furthermore, after obtaining the user's consultation information, an AI customer service representative can provide a corresponding response based on the user's lead generation information. The response content can be generated in real time based on the lead generation information using the large language model described in the above embodiment; alternatively, after obtaining the user's consultation information, a human customer service representative can provide a corresponding response based on the user's lead generation information. This embodiment does not limit this approach.

[0076] Optionally, in response to a user's triggering action on the first lead-giving control corresponding to the online consultation lead-giving method, the online consultation page can be displayed. This includes: responding to a user's triggering action on the first lead-giving control corresponding to the online consultation lead-giving method by redirecting to the online consultation page; or, responding to a user's triggering action on the first lead-giving control corresponding to the online consultation lead-giving method by displaying the online consultation page (e.g., a pop-up page) in the associated area of ​​the lead-giving information collection page (e.g., a pop-up page). Furthermore, if the online consultation page is a pop-up page, the pop-up page has an initial size, which can be smaller than the screen size; furthermore, in response to a user's gradual enlargement action on the pop-up page, the pop-up page can be gradually enlarged to full screen. Further, the online consultation page also includes a return control, used to return to the lead-giving information collection page. The return control can be, for example, "<", and in response to a user's triggering action on the return control, the user returns to the lead-giving information collection page.

[0077] Method 3: For example Figure 2g As shown, when the lead generation method is audio / video consultation, the system responds to the user's trigger operation on the first lead generation control corresponding to the audio / video consultation lead generation method, and displays the audio / video consultation page (also known as the call page). The audio / video consultation page includes the target merchant's contact information and its corresponding first lead generation control; it also responds to the user's audio / video call operation initiated through the first lead generation control corresponding to the target merchant's contact information, and obtains the user's lead generation information.

[0078] Optionally, in response to a user's triggering action on the first lead-collection control corresponding to the audio / video consultation lead-collection method, the audio / video consultation page can be displayed, including: responding to a user's triggering action on the first lead-collection control corresponding to the audio / video consultation lead-collection method and redirecting to the audio / video consultation page; or, responding to a user's triggering action on the first lead-collection control corresponding to the audio / video consultation lead-collection method and displaying the audio / video consultation page (e.g., a pop-up page) in the associated area of ​​the lead-collection information collection page (e.g., a pop-up page). Furthermore, if the audio / video consultation page is a pop-up page, the pop-up page has an initial size, which can be smaller than the screen size; furthermore, in response to a user's gradual enlargement action on the pop-up page, the pop-up page can be gradually enlarged to full-screen display. Further...

[0079] Optionally, in response to a user's call initiated through the first contact information control corresponding to the target merchant's contact information, the system obtains the user's contact information, including: establishing an audio / video channel between the user and the merchant and displaying an audio / video call page; and transmitting communication information between the user and the merchant through the audio / video channel to obtain the user's contact information. Further, if the audio / video call page includes a call end control, the system responds to the user's triggering of the call end control to end the call and close the audio / video call page. Further, if the audio / video consultation page also includes a cancel control used to close the audio / video consultation page, the system responds to the user's triggering of the cancel control to close the audio / video consultation page.

[0080] Optionally, in response to a user's triggered action regarding the first target merchant's information, the merchant's store details page can be displayed to meet the user's information needs. For example... Figure 2h As shown, the merchant's store details page includes, but is not limited to, at least one of the following: service or product links, image content, business content, merchant lead generation methods and their corresponding lead generation controls. The functions of the merchant lead generation methods and their corresponding lead generation controls can be found in the relevant descriptions of the lead generation controls in the above embodiments, and will not be repeated here.

[0081] Furthermore, to facilitate user contact information collection, a contact information card is provided. Specifically, in response to a swipe operation on the first target article page, a contact information card corresponding to the first target merchant information can be displayed if set swipe conditions are met. This embodiment does not limit the specific content of the set swipe conditions. For example, the set swipe conditions could be that the contact information card corresponding to the first target merchant information is displayed when the number of consecutive swipe operations reaches a set value. Another example is that when a specific area or position on the first target article page is swiped into the display screen, such as when the first target merchant information is swiped into the display screen, the contact information card corresponding to the first target merchant information is displayed. This embodiment also does not limit the display position of the contact information card; for example, the contact information card can be displayed in the upper, middle, or bottom area of ​​the screen. This embodiment also does not limit the display direction and size of the contact information card; for example, the contact information card can be displayed horizontally or vertically on the screen. The final display direction and size of the contact information card are designed for aesthetic appeal and minimal user interference. Figure 2f This is an exemplary schematic diagram of a data entry card.

[0082] Optionally, the contact card includes at least one contact method selected from online consultation, audio / video consultation, and appointment consultation, along with a corresponding second contact control, allowing users to initiate a contact action. Based on this, in response to a user's contact action initiated on the contact card, the user's contact information can be obtained.

[0083] Optionally, in response to a user's contact information creation action on the contact information card, the user's contact information can be obtained in the following ways:

[0084] Method 1: If the lead generation card includes the appointment consultation lead generation method and its corresponding second lead generation control, respond to the user's trigger operation on the second lead generation control corresponding to the appointment consultation lead generation method, display the lead generation information collection page, which includes category information items and user information items; respond to the user's trigger operation on the category information items, display the category list, which includes multi-level categories; respond to the user's selection operation on each level of category, display each selected level of category in the category information item.

[0085] Method 2: If the lead card includes an online consultation lead generation method and its corresponding second lead generation control, respond to the user's trigger operation on the second lead generation control corresponding to the online consultation lead generation method, and display the online consultation page, which includes a content input control; respond to the user's content input operation initiated through the content input control, and obtain the user's lead generation information.

[0086] Method 3: If the lead generation card includes an audio / video consultation lead generation method and its corresponding second lead generation control, respond to the user's trigger operation on the second lead generation control corresponding to the audio / video consultation lead generation method, display the audio / video consultation page, which includes the target merchant's contact information and its corresponding call control; respond to the user's audio / video call operation initiated through the audio / video call control, and obtain the user's lead generation information.

[0087] It should be noted that the operation mode of the second lead retention control corresponding to each lead retention method in this embodiment and the specific implementation mode of each step can be found in the relevant description of the above embodiment of obtaining the user's lead retention information in response to the lead retention operation initiated by the user on any first lead retention control, and will not be repeated here.

[0088] In this embodiment, the first target article page further includes a second area, which includes entry information pointing to the Knowledge Plaza page. Access to the Knowledge Plaza page is possible through this entry information. This embodiment does not limit the implementation form of the entry information pointing to the Knowledge Plaza page. For example, the entry information can be interactive interface elements such as cards, links, titles, or video covers. Different forms of entry information for the Knowledge Plaza page may correspond to different generation methods. For details, please refer to the detailed description in the above embodiment of generating at least one search result item pointing to at least one article page based on the address information of at least one article page; this will not be repeated here. For another example, the entry information pointing to the Knowledge Plaza page can also be a hierarchical link control. This control at least contains a hierarchical link relationship between the first target article page and the Knowledge Plaza page. In response to a trigger operation on the corresponding level in the control, the corresponding page is displayed. For yet another example, the second area can simultaneously include one interactive interface element from cards, links, titles, and video covers, as well as a hierarchical link relationship control, and the two are located in different sub-areas of the second area.

[0089] Figure 2i This is an exemplary diagram of the second area, located below the first area. The first sub-area of ​​the second area includes an interactive interface element in the form of a title, namely the entry information "Related Searches. Knowledge You Didn't Know." In response to a user's triggering action on this entry information, the Knowledge Plaza page is displayed. The second sub-area of ​​the second area includes hierarchical relationship controls, namely, hierarchical linking controls of "Homepage < Article Page < Knowledge Plaza." These hierarchical linking controls include: a Homepage control, an Article Page control, and a Knowledge Plaza page control, used for switching between the corresponding pages. In response to a user's triggering action on any of the sub-controls at any level, the corresponding page is displayed.

[0090] Optionally, the first target article page further includes a third area, which includes a comment control and a comment section. Based on this, in response to a user's comment operation initiated through the comment control, the user's input comment content can be obtained and displayed in the comment section. This embodiment does not limit the display order of comment content in the comment section. For example, comment content can be displayed in chronological order; or, comment content can be displayed in reverse chronological order; or, comment content can be displayed randomly, etc. In addition, each comment content is located in a different sub-area, and each sub-area can also display the corresponding user information, comment time, and interactive controls. User information includes, but is not limited to, user avatar and username. Interactive controls include, but are not limited to, reply controls and like controls. The reply control is used to reply to the corresponding comment content, and each user can also delete their own comment or reply. Correspondingly, in response to a user's triggering operation on the reply control, an information input control is displayed; in response to a user's comment content posting operation triggered by the information input control, the comment content is displayed in the comment section. The like control is used to support liking comment content or reply content. Accordingly, in response to the user's first triggering of the like control, the control is activated, and the number of likes is counted and displayed synchronously; in response to subsequent triggering of the like control, the number of likes is accumulated and updated synchronously. Additionally, the same user can also cancel their own like, and in the event of cancellation, the number of likes will be updated again.

[0091] It should be noted that this embodiment does not limit the content of user comments and replies. For example, it can be questions about the article content or target merchant information embedded in the first target article page. In this regard, the target platform can periodically collect comments and replies and make the comments and replies with high frequency the popular comments of the current period, so as to update the page content of each target article page based on the popular comments.

[0092] Figure 2j This is an exemplary diagram of the third area. The comment control is located at the top of the third area, and the comment area is located below the comment control. The comments in the comment area are displayed in their respective sub-areas in reverse chronological order. Each comment also displays the commenter's username, avatar, comment time, and location when the comment was made. The reply to each comment is displayed below the corresponding comment, and the number of likes is displayed in the associated area of ​​the comment or reply.

[0093] Optionally, the first target article page further includes a fourth area, which displays at least one category tag and at least one third target article page associated with the currently selected category tag. Each category tag is associated with at least one third target article page, and the third target article page has the same or similar page structure as the first target article page. The third target article page and the second target article page may be duplicate pages. The at least one category tag is determined based on at least one of the following: search terms, user search behavior information, and search behavior information of the user's associated users. Associated users refer to geographically proximate users who have viewed the same article content, users located in the same business district, or custom-defined associated users. Search behavior information can be search behavior information or browsing behavior information. If the user does not select a tag, a category tag can be selected by default, and at least one third target article page associated with that tag can be displayed. Preferably, the default selected tag can be the category tag to which the first target article page belongs, so that the user can learn more information related to that category. When the user selects a target category tag, in response to the user's selection of the target category tag, at least one third target article page associated with the target category tag is displayed. Furthermore, in response to the user's swiping operation, more third-party article pages related to the target category and target tag can be displayed by flipping through pages. The swiping operation can be up and down or left and right to flip through pages. This embodiment does not limit this.

[0094] It should be noted that at least one third target article page associated with each target category tag may be pre-loaded into the fourth area of ​​the first target article page, or it may be loaded into the fourth area of ​​the first target article page in real time in response to the user's selection operation of the target category tag. This embodiment does not limit this.

[0095] Figure 2k This is an exemplary diagram of the fourth area, using recycling as an example. The fourth area can be considered a recommended reading area, comprising a first sub-area and a second sub-area. The first sub-area displays multiple interactive category information tags, and the second sub-area displays at least one third-target article page associated with the selected category tag. Taking the default category tag "Must-Read for Recycling" as an example, the second sub-area displays at least one third-target article page associated with the "Must-Read for Recycling" tag by default. Other tags in the first sub-area could be "Hot Recycling," "Old Furniture Recycling," "Used Electric Vehicles," etc. The third-target article page can be displayed in the second sub-area as an interactive card pointing to the corresponding third-target article page. The card includes a cover image, topic information, author information, and the number of views for the article.

[0096] Following the first target article page, the second area includes an embodiment with knowledge square entry information. In response to a user's access operation initiated by the entry information, the knowledge square page can be displayed. The knowledge square page includes at least one category information related to the search term and at least one target article page associated with each category information, referred to as a second target article page, allowing the user to continue viewing any second target article page under any category information. By accessing one article page found through a search, a user can view more related article pages. This page structure provides users with richer and more interesting related information, satisfying their information needs and attracting user traffic.

[0097] This application does not limit the implementation form of the knowledge square page. For example, the knowledge square page can be a separate page; correspondingly, in response to the user's access operation initiated by the entry information, the user is redirected to the knowledge square page. As another example, the knowledge square page can also be a pop-up page; correspondingly, in response to the user's access operation initiated by the entry information, the pop-up page is displayed above the first target article page, and the size of the pop-up page is smaller than the size of the first target article page.

[0098] Figure 2l This is an exemplary diagram of a knowledge square page using the recycling category as an example. The knowledge square page includes three sub-areas: the first sub-area is located at the top of the page, and the second and third sub-areas are displayed side by side. The first sub-area displays a background image, which can be generated using an AI multimodal generation model based on artificial intelligence. The specific generation process can be found in the relevant descriptions of the following embodiments, and will not be repeated here. The background image includes the theme information of the current category, a hierarchical link control embedded at a designated location in the sub-area ("Homepage > Article Content > Knowledge Square"), and a return control used to return to the previous page. The second sub-area displays at least one category information and a corresponding control for each category information. The at least one category information includes, but is not limited to: Must-Read for Recycling, Furniture Recycling, Home Appliance Recycling, Mobile Phone Recycling, Metal Scrap Recycling, Machinery and Equipment Recycling, and Digital Product Recycling. Furthermore, in response to a trigger operation on the control corresponding to any category information, multiple second target article pages under that category information are displayed in the third sub-area. Furthermore, in response to a trigger operation on any second target article page, the corresponding second target article page is displayed.

[0099] To further meet user information needs and attract user traffic, merchant information can be added to the second target article page. In response to a user's access to any second target article page under any category on the Knowledge Square page, the system obtains the user's current location information; sends this location information to the target platform, allowing the platform to embed second target merchant information adapted to the current location information into the second target article page before returning to the second target article page; and finally, displays the second target article page returned by the target platform. Other implementation methods for obtaining the user's current location information and displaying the target article page described in the above embodiment are also applicable to this embodiment and will not be repeated here.

[0100] Optionally, the second target article page has the same or similar page structure as the first target article page. Figure 2m This is an exemplary diagram of the first target article page. The first and second areas of the second target article page respectively include the second article content and entry information pointing to the Knowledge Square page. The second target article page also includes second target merchant information, which refers to merchants that can provide services or goods under the category to which the second target article page belongs. For the other areas of the second target article page and their details, please refer to the relevant description of the first target article page, which will not be repeated here.

[0101] Following the above embodiment of generating the first article content using an AI multimodal generation model based on artificial intelligence, the first article content can be generated based on article content related to hot topics. Hot topics may be periodic. For example, in summer, the hot topic may be the recycling of air conditioners or electric fans, while before the Spring Festival, due to the renovation of the home, the corresponding hot topic may be the recycling of furniture. Based on this, in order to improve the richness of the article content and meet users' needs for real-time information, the first article content is generated based on a multimodal generation model. An optional implementation method is as follows: Periodically acquire popular search terms and popular comments corresponding to at least one category. Popular search terms are those whose search frequency meets a set threshold within a period, and popular comments are those whose comment count and / or like count meet a set threshold within a period. Based on popular search terms and popular comments, determine the hot topics within the current period. Each hot topic belongs to one category. A hot topic is a topic whose mention count by users exceeds a preset threshold within a set period. If the hot topics of the current period change compared to the hot topics of the previous period, acquire historical article content related to the changed target hot topic. Input the target hot topic, the category information to which the target hot topic belongs, and the historical article content into an AI multimodal generation model based on artificial intelligence. Generate multimodal content for the target hot topic and related historical article content under the same category information to obtain the article content corresponding to each category. The article content corresponding to each category includes the first article content.

[0102] Optionally, popular search terms and comments can be acquired periodically, including periodically acquiring popular search terms and comments from the target platform and / or other platforms. More specifically, popular search terms and comments can be acquired periodically from the target platform and other platforms; or, popular search terms and comments can be acquired periodically from the target platform; or, popular search terms and comments can be acquired periodically from other platforms. This embodiment does not specifically limit the length of the period, for example, 7 days, 15 days, one month, one quarter, etc. In addition, the length of the period may also vary depending on the method of acquiring popular search terms and comments periodically. For example, the period set when acquiring popular search terms and comments only from the target platform may be longer, and the period set when acquiring popular search terms and comments from other platforms may be shorter, or vice versa. Taking furniture recycling as an example, the hot topics of the current period may be such as "What can I buy?", "How much is the recycling price?", "Will you buy a piece of furniture?", etc.

[0103] Optionally, the article content corresponding to each category information includes the article theme, image content, and text content belonging to that category information. The text content includes at least one hot topic and topic content under each hot topic. Then, multimodal content generation is performed on the target hot topic and related historical article content under the same category information to obtain the article content corresponding to each category information. This includes: for any category information, determining the theme structure based on the attribute information of that category information, and generating the article theme for that category information based on the theme structure; based on the built-in prompts of the multimodal generation model, integrating the historical hot topics and historical topic content from the related historical articles of that category information according to the target hot topic under that category information, to obtain at least one hot topic under that category information and topic content under each hot topic; inputting at least one hot topic and topic content under each hot topic under that category information into the text-to-image model in the multimodal generation model for text-to-image processing to obtain the image content belonging to that category information.

[0104] Optionally, the article topics corresponding to different category information may have different topic structures, depending on the attribute information of the category information. Attribute information includes, but is not limited to, the specific target audience related to the category information and the service information that merchants under the category information can provide. For example, for some category information related to a specific target audience, the corresponding article topic may include the user profile of that specific target audience. For example, for the lipstick recycling category, the specific target audience involved is women, that is, the corresponding user profile can be "sisters", so the article topic for the lipstick recycling category can be "sisters, lipstick recycling...". In addition, some merchants may also provide some specific services, such as lipstick recycling that can be picked up within 30 minutes or even one piece. In order to highlight the specific service information and attract user traffic, the corresponding article topic can also include the specific service information, so the article topic for the lipstick recycling category can be "sisters, lipstick recycling, pick-up within 30 minutes, even one piece...". In addition, the article topic may also include other information, which is not limited in this embodiment.

[0105] Optionally, the AI-based multimodal generation model incorporates built-in prompt words to enable it to generate article content. These prompt words are textual content used to guide the AI-based multimodal generation model in generating article content with a specific style, theme, or structure. The prompt words include both positive and negative prompt words. For generating article themes, the corresponding prompt words can guide the model on how to select the theme structure, word count, and font style, etc., to guide the AI-based multimodal generation model in generating article themes with the theme structure, word count, and font style that match the search terms. For generating text content, the corresponding prompt words can guide the model on how to select and determine the number of hot topics included in the text content, the specific hot topics, the specifications of the topic content under each hot topic, and the word count of the text content, etc., to guide the AI-based multimodal generation model in generating text content with the number of hot topics, the specific hot topics, the specifications of the topic content under each hot topic, and the word count of the text content that match the search terms. For the generated image content, the corresponding prompts can be textual information guiding the model on how to determine the number of images, the subject content of each image, and the size of the images, so as to guide the AI ​​multimodal generation model based on artificial intelligence to generate image content that matches the category information of the search terms, including the number of images, the subject content of each image, and the size of the images.

[0106] Based on this, optionally, based on the built-in prompt words of the multimodal generation model, the historical hot topics and historical topic content in the historical articles related to the category information are integrated according to the target hot topics under the category information to obtain at least one hot topic under the category information and topic content under each hot topic. This includes: determining the number of hot topics contained in the article content to be generated and the corresponding target hot topics based on the prompt words, the category information corresponding to the article content to be generated, and the number of hot topics corresponding to the category information. For example, the number of target hot topics contained in the article content corresponding to different category information is different according to the prompt words, and / or the number of target hot topics contained in each article content. Select from the current category's trending topics according to a certain proportion; if each target characteristic topic is included in the historical characteristic topics related to the category, extract the corresponding content from the historical topic content corresponding to the relevant historical trending topics according to the requirements of the prompt words for the content structure of each topic, and fill the corresponding content into the content structure of each topic to obtain at least one trending topic under the category and the topic content under each trending topic; if at least one target characteristic topic is missing in the historical characteristic topics related to the category, generate the corresponding topic content using an AI multimodal generation model based on artificial intelligence, according to the requirements of the prompt words for the content structure of each topic and the relevant network knowledge of the topic.

[0107] It should be noted that for a certain category of information, if the trending topics in the current period have not changed compared to the previous period, and the corresponding historical topic content does not need to be updated, there is no need to regenerate the article content related to that category of information. Alternatively, for a certain category of information, if its related historical article content contains the target trending topics and target feature topics to be generated in the current period, and the topic content corresponding to these target features does not need to be updated, but the related historical article content is missing one or more target trending topics, then it is only necessary to regenerate the topic content corresponding to the missing features, and then add the missing target trending topics and their regenerated content to the related historical article content (i.e., update based on the original related historical article content).

[0108] Optionally, for any category of information, a topic structure is determined based on the attribute information of that category. An article topic for that category is then generated based on this topic structure. This includes determining the topic structure for any category based on the prompts corresponding to the article topic and the attribute information of the category. The topic structure includes at least one structural piece of information. At least one structural piece of information is then extracted from the topic structure and added to the corresponding position in the topic structure to obtain the article topic for that category. For example, if the article topic "Sisters, lipstick recycling, door-to-door pickup within 30 minutes, even one piece..." includes the following four items: "Sisters," "lipstick recycling," "door-to-door pickup within 30 minutes," and "even one piece," then before generating the article topic, these four structural pieces of information need to be extracted from the article content based on the topic structure, and then added to the corresponding position in the topic structure to obtain the article topic.

[0109] Based on this, optionally, at least one hot topic under the category information and the topic content under each hot topic are input into the text-to-image model in the multimodal generation model for text-to-image processing to obtain image content belonging to the category information. This includes: inputting at least one hot topic under the category information and the topic content under each hot topic into an AI multimodal generation model based on artificial intelligence; determining the number of images, the element features contained in each image, and their relative position information based on at least one hot topic, the topic content under each hot topic, and prompt words; determining the resolution of the images and creating a latent space for the images based on the resolution, with the vector dimension of the latent space matching the feature space dimension of the images; using an encoder to map the element features and relative position information in each image to the latent space to obtain the latent space features of the element features in the latent space; using a sampler to collect features that match at least one prompt word from the latent space features of the element features; and using a decoder to decode the element feature vectors to obtain each image.

[0110] In some embodiments, to enhance the functionality and flexibility of the text-based image model, plug-in models can be integrated into it. This embodiment does not specifically limit the type of plug-in model; for example, the plug-in model could be an InstantID model. The InstantID model includes an element feature constraint model and an element feature extraction model. The element feature constraint model could be a ControlNet model, and the element feature extraction model could be an Ipadapter model. The coexistence of the element feature constraint model and the element feature extraction model, i.e., by decoupling the cross-attention mechanism to perform image processing separately, can improve the accuracy of the image.

[0111] Furthermore, the target page also has a target page structure, and target article pages are generated based on this structure. After obtaining the article content corresponding to each category, the target page structure can be obtained. This structure includes a first area and a second area. The second area includes entry information pointing to the Knowledge Square page. The article content corresponding to each category can be added to the first area of ​​the target page structure to obtain the target article page for each category. The target article pages include a first target article page, a second target article page, and a third target article page.

[0112] Furthermore, if the target page structure includes a third area, a comment control can be added to the third area. The comment control is used by users to make comments; it receives the user's comment content sent from the user's terminal and adds the comment content to a specified position in the comment area; wherein, the comment content is entered by the user through the comment control.

[0113] Furthermore, the target page structure includes a third area, which can determine at least one category target tag and at least one target article page associated with each category target tag based on at least one of the following: search terms, user search behavior information, and search behavior information of the user's related users. The at least one category target tag and at least one target article page associated with the default category target tag are embedded in the fourth area of ​​the target article page corresponding to the default category to obtain the target article page, namely, the first target article page, the second target article page, and the third target article page.

[0114] Furthermore, the address information of the target article page corresponding to each category information can be published on the target search engine, so that users can perform the process described in the above embodiment of searching for the target article page corresponding to any category information through the target search engine and accessing it if the target page corresponding to any category information is found.

[0115] The following example illustrates the cue words for close-up / close-up shots. Cue words for close-up / close-up shots can include the following ("##" is used to represent each cue word):

[0116] "##Character Name: Expert in Creating Illustrated Prompts for Popular Science Articles, {{input}} is a fragment of a popular science article entered by the user."

[0117] ##Objective: Analyze {{input}} and rewrite it into a prompt from a third-party perspective. This prompt should then be used to create an image using the Static Image drawing tool. This image will be used as an illustration for a science article and needs to accurately convey the meaning, scenario, and even feelings that this article segment is meant to express.

[0118] ##Focus Area: Please focus on the task of rewriting excerpts from popular science articles into vivid descriptions, providing high-quality output. Please describe rich details, with highly detailed environmental and scene descriptions. Output prompts in both Chinese and English, with and without characters.

[0119] ## Output language: Chinese and English.

[0120] ## Description: An expert in creating visual aids for popular science articles. They can accurately understand article segments and rewrite the original content into a descriptive text of a scene. The rewritten text must be able to be used by the Staled IFFUTION drawing tool to create an image that accurately conveys the article segment.

[0121] ###Skill: Sometimes, text passages may lack visual descriptions, requiring you to use your imagination to fill in the missing elements. When filling in the blanks, you need a complete description of the visual elements, a thorough understanding of the original text, and the identification of key keywords to compose the elements. Consider the context and the text, make reasonable inferences, optimize or adjust inappropriate elements, or appropriately supplement missing elements.

[0122] ##Constraints: Please rewrite the input text according to the context. Do not change the meaning, omit anything, or fabricate information; the text must be logical. Follow the original text's logic and avoid introducing ambiguity. The revised text must be at least 100 characters long. Do not output text that cannot form a visual image, such as "selected." Do not output scenes unrelated to the central theme of the text. Use third-person pronouns for character names, such as: a man, a woman, a boy, a girl, a little boy, a little girl, a middle-aged woman, a middle-aged man, an elderly woman, an elderly man, two girls, three boys, etc. If the text contains occupational names, such as cleaner, nanny, repairman, etc., use "a female cleaner," "a female nanny," "a male repairman," etc. to represent the character names. Do not use names like "you," "I," or "he / she." Do not use broad terms like "everyone" or "people."

[0123] ##Analysis process: The analysis can be carried out according to the following workflow: "I. Analyze {{input}}, and divide it into two cases according to whether there are characters in the text: 1. The case where there are characters in {{input}}, which is further divided into three cases: (1) The text describes a person doing something in a certain place. This case is inherently more vivid, so we only need to analyze what the content of the scene that does not appear in the text is, for example: [The nanny patiently communicates with the little girl in the living room] Analysis: This text has the characters [nanny], expressions [patiently], actions [communicating with the little girl], and scene [living room], but it lacks ambient lighting and camera angle. We need to think about how to add these two elements (ambient lighting) (and camera angle). Considering that the scene is the living room, it can be reasonably inferred that the ambient lighting is [daytime]. Since the character's expression and actions are [patiently communicating with the little girl], in order to see the character's expression more clearly, it can be reasonably inferred that the camera angle is [frontal view, half-body shot]. The scene can be appropriately supplemented to enrich the picture. For example, based on [nanny], [little girl], and [living room], it can be inferred that there are objects [toys], [carpets], and [books] in the scene. (2) When a profession appears in the text, the possible character actions in the scene can be inferred from the context. For example: [What kind of person is reliable for house cleaning? House cleaning is an important issue that every family needs to face, and choosing a reliable cleaner is even more crucial.] Important. Analysis: The character is a cleaning staff member. The common clothing of cleaning staff is an apron. The common action is mopping the floor. The location is the living room. The expression is serious. The ambient lighting is daytime. The camera angle is full-body. You can add common clothing and tools for the profession to enrich the picture according to the scene. For example, the common clothing of cleaning staff is an apron and gloves. The common tools are cleaning tools such as rags and mops. Modify according to the different nature of different professions. (3) Extract the central event of the text and determine the staff members who may be involved in the event according to the context. For example: Moving is a tedious thing. Choosing a reliable moving service company can greatly reduce your burden. But how Find a more reliable moving service? Analysis: The central event is moving. In this kind of scenario, the common characters are [moving staff], the action is [moving boxes], the scene is [the neighborhood], the expression is [serious], the ambient lighting is [daytime], and the camera angle is [full-body]. 2. When there are no characters in the text, we need to analyze the scene, ambient lighting and camera angle described in the text according to the context. (1) When there is a location, for example: [the lawyer's office is empty] Analysis: The scene is [the lawyer's office]. According to the context, the ambient lighting is [daytime] and the camera angle is [panoramic]. According to the nature of the scene, there may be objects in the scene such as [contract], [computer], [table], [chair], etc.(2) When specific occupational terms appear in the text, focus on the tools commonly used in that occupation. For example, in the question "What kind of people are reliable for home cleaning?", the commonly used tools for home cleaning are "cleaning tools," the frequently occurring environment is "living room," the ambient lighting is "daytime," and the camera angle is "close-up shot." Note: This step does not require you to output any content; it only represents your workflow. II. Divide the {{input}} into paragraphs. Each paragraph will be separated by punctuation marks

。

!

?

[0124] The technical solutions provided in the above embodiments of this application allow users to search for search results containing links to a first target article page published by a target platform based on search terms, and then jump to the first target article page based on these search results. Furthermore, the first target article includes multiple display areas. The first area displays multimodal article content generated by an AI multimodal generation model related to the search terms. Users can access a knowledge square page through the entry information in the second area, where they can continue to view any article page under any category. By accessing one article page through a search, users can view more related article pages. This page structure provides users with richer and more interesting information, meeting their information needs and attracting user traffic.

[0125] Figure 3 This is a flowchart illustrating an exemplary embodiment of the method for generating an article page provided in this application. Figure 3 As shown, the method includes:

[0126] 301. Periodically retrieve popular search terms and popular comments corresponding to at least one category;

[0127] 302. Based on popular search terms and popular comments, use an AI multimodal generation model to generate multimodal content for each category of information to obtain the article content corresponding to each category of information;

[0128] 303. Obtain the target page structure, which includes a first area and a second area. The second area includes entry information pointing to the Knowledge Square page.

[0129] 304. Add the article content corresponding to each category information to the first area of ​​the target page structure to obtain the target article page corresponding to each category information.

[0130] Optionally, based on popular search terms and popular comments, an AI-based multimodal generation model is used to generate multimodal content for each category of information to obtain the article content corresponding to each category. This includes: determining the hot topics in the current period based on popular search terms and popular comments, with each hot topic belonging to a category; if the hot topics in the current period have changed compared to the hot topics in the previous period, obtaining historical article content related to the changed target hot topic; and inputting the target hot topic, the category information to which the target hot topic belongs, and the historical article content into the AI-based multimodal generation model to generate multimodal content for the target hot topic and related historical article content under the same category, to obtain the article content corresponding to each category.

[0131] Optionally, the article content corresponding to each category information includes the article theme, image content, and text content belonging to that category information. The text content includes at least one hot topic and topic content under each hot topic. The target hot topic, the category information to which the target hot topic belongs, and historical article content are input into an AI multimodal generation model based on artificial intelligence. Multimodal content is generated for the target hot topic and related historical article content under the same category information to obtain the article content corresponding to each category information. This includes: for any category information, determining the theme structure based on the attribute information of the category information, and generating the article theme of the category information based on the theme structure; based on the built-in prompt words of the multimodal generation model, integrating the historical hot topics and historical topic content in the related historical article content of the category information according to the target hot topic under the category information to obtain at least one hot topic under the category information and topic content under each hot topic; inputting at least one hot topic and topic content under each hot topic under the category information into the text-to-image model in the multimodal generation model for text-to-image processing to obtain the image content belonging to the category information.

[0132] Optionally, the address information of the target article page corresponding to each category information can be published on the target search engine, so that users can search for the target article page corresponding to any category information through the target search engine and access it if the target page corresponding to any category information is found.

[0133] Optionally, the system obtains the current location information sent by the user terminal. This current location information is the location information when the user initiates a trigger operation on a search result item pointing to a target article page corresponding to the target category information. The search result item is found based on search terms, and the target category information can be any category. Based on the target category information, the system obtains merchant information that can provide services or goods under the target category information. Based on the current location information, the system selects target merchant information that matches the current location information from the merchant information. After embedding the target merchant information into the target article page corresponding to the target category information, the system sends the target article page corresponding to the target category information to the user terminal for display to the user.

[0134] Optionally, the target page structure may further include a third area, which includes a message area. In this case, a message control is added to the third area. The message control is used by users to make comments. The system receives the user's comment content sent from the user's terminal and adds the comment content to a specified position in the message area. The comment content is entered by the user through the message control.

[0135] Optionally, the target page structure further includes a fourth area. Before sending the target article page corresponding to the target category information to the user terminal, at least one category target tag and at least one target article page associated with each category target tag are determined based on at least one of the following: search terms, user search behavior information, and search behavior information of the user's associated users. The at least one category target tag and at least one target article page associated with the default category target tag are embedded in the fourth area of ​​the target article page corresponding to the default category.

[0136] Optionally, after sending the target article page corresponding to the target category information to the user terminal, a request is received from the user terminal. The request is used to request at least one target article page associated with the target category tag selected by the user. According to the request, at least one target article page associated with the target category tag is sent to the user terminal so that the user terminal can display at least one target article page associated with the target category tag in the fourth area.

[0137] In this embodiment, based on periodically acquiring popular search terms and comments corresponding to at least one category of information, an AI multimodal generation model is used to generate article content for each category. This content is periodic and represents popular topics within the current period. Furthermore, the generated content is rich in multimodal comparisons, satisfying users' information needs within the current period and attracting user traffic. Additionally, based on the target page structure, target article pages corresponding to each category can be obtained. For example, if the page structure includes a first area and a second area, the article content for each category can be added to the first area of ​​the target page structure, while the second area includes entry information pointing to a knowledge square page. Users can then view more related article pages through a searched article page. This page structure provides users with richer and more interesting information, satisfying their information needs and attracting user traffic.

[0138] The detailed implementation methods and beneficial effects of each step provided in the embodiments of this application have been described in detail in the foregoing embodiments, and will not be elaborated here.

[0139] Figure 4 A schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application. For example... Figure 4 As shown, it includes: a memory 40a and a processor 40b; the memory 40a is used to store computer programs; the processor 40b, coupled to the memory 40a, is used to execute the computer programs to perform the following steps:

[0140] Periodically acquire popular search terms and popular comments corresponding to at least one category of information; based on the popular search terms and popular comments, use an AI multimodal generation model to generate multimodal content for each category of information to obtain the article content corresponding to each category of information; acquire the target page structure, which includes a first area and a second area, the second area including entry information pointing to the knowledge square page; add the article content corresponding to each category of information to the first area of ​​the target page structure to obtain the target article page corresponding to each category of information.

[0141] Optionally, when the processor generates multimodal content for each category of information based on popular search terms and popular comments using an AI-based multimodal generation model, to obtain the article content corresponding to each category of information, the specific steps are as follows: Based on popular search terms and popular comments, determine the hot topics within the current period, with each hot topic belonging to a category of information; if the hot topics of the current period have changed compared to the hot topics of the previous period, obtain historical article content related to the changed target hot topic; input the target hot topic, the category information to which the target hot topic belongs, and the historical article content into the AI-based multimodal generation model to generate multimodal content for the target hot topic and related historical article content under the same category of information, to obtain the article content corresponding to each category of information.

[0142] Optionally, the article content corresponding to each category information includes the article theme, image content, and text content belonging to that category information. The text content includes at least one hot topic and topic content under each hot topic. When the processor inputs the target hot topic, the category information to which the target hot topic belongs, and historical article content into the AI ​​multimodal generation model based on artificial intelligence, and generates multimodal content for the target hot topic and related historical article content under the same category information to obtain the article content corresponding to each category information, it specifically performs the following: For any category information, it determines the theme structure based on the attribute information of the category information, and generates the article theme of the category information based on the theme structure; Based on the built-in prompt words of the multimodal generation model, it integrates the historical hot topics and historical topic content in the related historical article content of the category information according to the target hot topic under the category information to obtain at least one hot topic under the category information and topic content under each hot topic; It inputs at least one hot topic and topic content under each hot topic under the category information into the text-to-image model in the multimodal generation model for text-to-image processing to obtain the image content belonging to the category information.

[0143] Optionally, the processor is also configured to publish the address information of the target article page corresponding to each category information to the target search engine, so that users can search for the target article page corresponding to any category information through the target search engine and access it if the target page corresponding to any category information is found.

[0144] Optionally, the processor is further configured to obtain current location information sent by the user terminal. The current location information is the location information when the user initiates a trigger operation on a search result item pointing to a target article page corresponding to the target category information. The search result item is found based on search terms, and the target category information can be any category. Based on the target category information, the processor obtains merchant information that can provide services or goods under the target category information. Based on the current location information, the processor selects target merchant information that matches the current location information from the merchant information. After embedding the target merchant information into the target article page corresponding to the target category information, the processor sends the target article page corresponding to the target category information to the user terminal for display to the user.

[0145] Optionally, if the target page structure further includes a third area, which includes a message area, then the processor is also used to add a message control to the third area, the message control being used by the user to make a comment; receive the user's comment content sent by the user terminal, and add the comment content to a specified position in the message area; wherein the comment content is entered by the user through the message control.

[0146] Optionally, the target page structure further includes a fourth region, and the processor is further configured to, before sending the target article page corresponding to the target category information to the user terminal, determine at least one category target tag and at least one target article page associated with each category target tag based on at least one of the following information: search terms, user search behavior information, and search behavior information of the user's associated users; and embed at least one target article page associated with at least one category target tag and a default category target tag into the fourth region of the target article page corresponding to the default category.

[0147] Optionally, the processor is further configured to, after sending the target article page corresponding to the target category information to the user terminal, receive an acquisition request sent by the user terminal, the acquisition request being used to request at least one target article page associated with the target category tag selected by the user; and according to the acquisition request, send at least one target article page associated with the target category tag to the user terminal so that the user terminal can display at least one target article page associated with the target category tag in the fourth area.

[0148] Furthermore, such as Figure 4 As shown, the server also includes other components such as a communication component 40c, a display 40d, a power supply component 40e, and an audio component 40f. Figure 4 The diagram only shows some components and does not mean that the electronic device includes only these components. Figure 4 The components shown.

[0149] The detailed implementation methods and beneficial effects of the electronic devices provided in this application have been described in detail in the foregoing embodiments, and will not be elaborated further here.

[0150] Exemplary embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps in the above-described method embodiments.

[0151] An exemplary embodiment of this application also provides a computer program product comprising a computer program / instructions that, when executed by a processor, enable the processor to perform the steps described in the above method embodiments.

[0152] The aforementioned memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0153] The aforementioned communication components are configured to facilitate wired or wireless communication between the device containing the communication components and other devices. The device containing the communication components can access wireless networks based on communication standards, such as WiFi, 2G, 3G, 4G / LTE, 5G, or combinations thereof. In one exemplary embodiment, the communication components receive broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication components also include a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID), Infrared Data Association (IRDA), Ultra Wide Band (UWB), Bluetooth (BT), and other technologies.

[0154] The aforementioned display includes a screen, which may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touchscreen to receive input signals from the user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensors can sense not only the boundaries of touch or swipe actions but also the duration and pressure associated with the touch or swipe operation.

[0155] The aforementioned power supply components provide power to various components within the device in which they reside. These power supply components may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power to the device in which they reside.

[0156] The aforementioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC) configured to receive external audio signals when the device containing the audio component is in an operating mode, such as call mode, recording mode, or voice recognition mode. The received audio signals can be further stored in memory or transmitted via a communication component. In some embodiments, the audio component also includes a speaker for outputting audio signals.

[0157] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-readable storage media (including, but not limited to, disk storage, compact disk read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.

[0158] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1A device that provides the functions specified in one or more boxes.

[0159] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0160] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0161] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0162] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash memory. Memory is an example of computer-readable media.

[0163] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change random access memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, Digital Video Disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media do not include transient computer-readable media, such as modulated data signals and carrier waves.

[0164] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0165] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. An article page generation method characterized by comprising: The method comprises the following steps: Periodically acquiring hot search words and hot comments corresponding to at least one category information; According to the hot search words and the hot comments, generating multi-modal content for each category information by using an AI multi-modal generation model based on artificial intelligence, to obtain article content corresponding to each category information; Acquiring a target page structure, the target page structure comprising a first area and a second area, and the second area comprising entry information pointing to a knowledge square page; wherein the article content comprises an article theme, picture content and text content belonging to the category information, and the text content comprises at least one hot topic and topic content under each hot topic; Acquiring a target page structure, the target page structure comprising a first area and a second area, and the second area comprising entry information pointing to a knowledge square page; Adding the article content corresponding to each category information into the first area of the target page structure respectively, to obtain a target article page corresponding to each category information; wherein the knowledge square page is a page that aggregates article pages corresponding to different category information.

2. The method of claim 1, wherein, According to the hot search words and the hot comments, generating multi-modal content for each category information by using an AI multi-modal generation model based on artificial intelligence, to obtain article content corresponding to each category information, comprising: According to the hot search words and the hot comments, determining hot topics in the current period, each hot topic belonging to a category information; If the hot topics in the current period change relative to the hot topics in the last period, acquiring historical article content related to the changed target hot topic; Inputting the target hot topic, the category information to which the target hot topic belongs and the historical article content into the AI multi-modal generation model based on artificial intelligence, to generate multi-modal content for the target hot topic and the related historical article content under the same category information, to obtain article content corresponding to each category information.

3. The method of claim 2, wherein, Inputting the target hot topic, the category information to which the target hot topic belongs and the historical article content into the AI multi-modal generation model based on artificial intelligence, to generate multi-modal content for the target hot topic and the related historical article content under the same category information, to obtain article content corresponding to each category information, comprising: For any category information, determining a theme structure according to attribute information of the category information, generating an article theme of the category information based on the theme structure; Based on the built-in prompt words of the multi-modal generation model, integrating historical hot topics and historical topic content in the historical article content related to the category information according to the target hot topic under the category information, to obtain at least one hot topic under the category information and topic content under each hot topic; Inputting the at least one hot topic under the category information and the topic content under each hot topic into a text-to-image model in the multi-modal generation model for text-to-image processing, to obtain picture content belonging to the category information. Further comprising: ​ ​ 4. The method of claim 1, wherein, ​ The address information of the target article page corresponding to each category information is published on a target search engine, so that a user searches for the target article page corresponding to any category information through the target search engine and accesses the target article page when the target article page is searched.

5. The method of claim 4, wherein, Further comprising: obtaining current position information sent by a user terminal, the current position information being position information when the user initiates a trigger operation on a search result item pointing to a target article page corresponding to target category information, the search result item being searched based on a search word, and the target category information being any category information; obtaining merchant information capable of providing services or goods under the target category information according to the target category information; selecting target merchant information adapted to the current position information from the merchant information according to the current position information; embedding the target merchant information into the target article page corresponding to the target category information, and sending the target article page corresponding to the target category information to the user terminal to display the target article page to the user.

6. The method of claim 5, wherein, The target page structure further comprises a third area, and the third area comprises a message area; the method further comprises: adding a message control to the third area, the message control being used for a comment operation of a user; receiving comment content of the user sent by the user terminal, and adding the comment content to a specified position of the message area; wherein the comment content is input by the user through the message control.

7. The method according to claim 5 or 6, characterized in that, The target page structure further comprises a fourth area; determining at least one category tag and at least one target article page associated with each category tag based on at least one of the search word, search behavior information of the user, and search behavior information of an associated user of the user; embedding the at least one target article page associated with the at least one category tag and a default category tag in the fourth area of the target article page corresponding to the default category.

8. The method of claim 7, wherein, Further comprising: receiving an acquisition request sent by the user terminal, the acquisition request being used for requesting at least one target article page associated with a target category tag selected by the user; sending the at least one target article page associated with the target category tag to the user terminal according to the acquisition request, so that the user terminal displays the at least one target article page associated with the target category tag in the fourth area.

9. An electronic device, comprising: comprising: a memory and a processor; the memory is used for storing a computer program; the processor is coupled with the memory and is used for executing the computer program to realize the steps in the method of any one of claims 1-8.

10. A computer readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the processor is caused to realize the steps in the method of any one of claims 1-8.

11. A computer program product, characterised in that, The computer program product comprises a computer program, when the computer program is executed by the processor, the processor is caused to realize the steps in the method of any one of claims 1-8.

Images