Content display method, apparatus, device, and medium

Abstract

Embodiments of the present application disclose a content display method, device, equipment and medium, which are applied to the technical field of Internet. The method comprises: displaying a collection page; the collection page comprises at least one published content and a recommended conversation content area, the recommended conversation content area comprises at least one recommended conversation content, and the recommended conversation content is determined according to the at least one published content; in response to a selection operation on a target recommended conversation content in the at least one recommended conversation content, the target recommended conversation content is sent to a target intelligent object, a conversation page with the target intelligent object is displayed, and a target reply content for the target recommended conversation content is displayed on the conversation page; the target reply content is determined based on a target published content, and the target published content is a published content related to the target recommended conversation content in the at least one published content. By adopting the embodiments of the present application, the convenience and efficiency of a user in extracting information content in a content collection scenario can be improved.

Description

Technical Field

[0001] This application relates to the field of Internet technology, specifically to the field of data display technology, and in particular to a content display method, apparatus, device, and medium. Background Technology

[0002] Currently, users can bookmark published content for easy access later, viewing it on a bookmark page. However, this bookmarking process typically displays content in chronological order of bookmarking time, such as in reverse chronological order. This means that when a user wants to find specific information (e.g., information related to a particular fitness exercise), they must navigate to each relevant post's details page to view the data and extract useful information (e.g., a tutorial video for a specific fitness exercise). Therefore, in a bookmarked scenario, users often find it difficult to directly access the information they want. Summary of the Invention

[0003] This application provides a content display method, apparatus, device, and medium, which can improve the richness of the display on the collection page in the content collection scenario, and improve the convenience and efficiency for users to view the information they want to know from the published content on the collection page.

[0004] On one hand, embodiments of this application provide a content display method, the method comprising: The Favorites page is displayed; the Favorites page includes at least one published content that has been saved and a recommended conversation content area, the recommended conversation content area includes at least one recommended conversation content, wherein the recommended conversation content is determined based on at least one published content; In response to a selection operation of a target recommended session content in at least one recommended session content, the target recommended session content is sent to the target smart object, and the session page with the target smart object is displayed. The target response content is displayed on the conversation page in relation to the target recommended conversation content; wherein the target response content is determined based on the target published content, which is at least one published content that is related to the target recommended conversation content.

[0005] On one hand, embodiments of this application provide a content display device, the device comprising: The page display module is used to display the collection page; the collection page includes at least one published content that has been collected and a recommended conversation content area, the recommended conversation content area includes at least one recommended conversation content, wherein the recommended conversation content is determined based on at least one published content; The content display module is used to respond to the selection operation of the target recommended session content in at least one recommended session content, send the target recommended session content to the target smart object, and display the session page with the target smart object; The content display module is also used to display target response content for the target recommended session content on the session page; wherein the target response content is determined based on the target published content, and the target published content is at least one published content that is related to the target recommended session content.

[0006] On one hand, embodiments of this application provide an electronic device including a processor and a memory, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute some or all of the steps in the above method.

[0007] On one hand, embodiments of this application provide a computer-readable storage medium storing a computer program, the computer program including program instructions, which, when executed by a processor, are used to perform some or all of the steps in the above-described method.

[0008] Accordingly, according to one aspect of this application, a computer program product or computer program is provided, which includes computer instructions that, when executed by a processor, can implement some or all of the steps in the above-described method.

[0009] In this embodiment, at least one published content that has been collected can be displayed on the collection page, along with at least one recommended conversation content obtained from the at least one published content. The recommended conversation content can be regarded as the user's possible query intent for the at least one published content (such as the query intent for content related to a certain fitness movement). In other words, the query intent that the user may have when collecting these published contents can be predicted and displayed in advance, so that the user can quickly engage in intelligent dialogue, thereby improving the richness of the content displayed on the collection page. In addition, by selecting the target recommended conversation content (i.e. the target query intent), an intelligent dialogue can be initiated with the target intelligent object, and the target reply content determined from the target published content related to the at least one published content can be viewed on the conversation page. That is, the user can quickly view the useful content extracted from the at least one published content through intelligent dialogue, without the user having to search for it himself. This can further improve the convenience and efficiency of the user in viewing the published content on the collection page and finding the information he / she wants to know. Attached Figure Description

[0010] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 A schematic diagram of a network architecture provided for an embodiment of this application; Figure 2 A schematic diagram illustrating a content display scenario provided in an embodiment of this application; Figure 3 A flowchart illustrating a content display method provided in this application embodiment. Figure 1 ; Figures 4a-4f A schematic diagram illustrating a content collection scenario provided in an embodiment of this application; Figure 5 A flowchart illustrating a content display method provided in this application embodiment. Figure 2 ; Figures 6a-6f A schematic diagram illustrating a display scenario of reply content provided in an embodiment of this application; Figure 7 A flowchart illustrating a content display method provided in this application embodiment. Figure 3 ; Figure 8 This is a schematic diagram of the structure of a content display device provided in an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0012] The content display method proposed in this application is implemented in an electronic device, which can be a server or a terminal. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, and big data and artificial intelligence platforms. The terminal can be a smartphone, tablet computer, laptop computer, desktop computer, etc., but is not limited to these.

[0013] In the description of the embodiments of this application, the published content may refer to notes (including text and images), short videos, medium-length videos, etc., posted by users on content platforms. Alternatively, it may also include live streams, instant (a type of short content that is shared and exists in real time), evaluations of products or any type of interest (such as locations, music, etc.) or media data (such as movies, music), comments in notes, bullet comments, topic discussions, points of interest (such as group chats, live streams, etc.), routes (such as cycling routes, travel guides, etc.). The specific content information included in the published content applicable to different scenarios can be adjusted accordingly, and the content information in the published content can be determined by the publisher (such as the target audience). The specific type of published content is not limited here.

[0014] This application allows users to view saved posts while simultaneously displaying recommended conversation content tailored to those posts, reflecting their potential viewing intent (query intent) when saving the content. Viewing intent refers to the latent intent identified for a specific type of saved content. For example, if a user saves fitness-related posts, they might want to learn about other fitness-related topics. If the saved fitness-related posts include content mentioning "shoulder exercises at home," the recommended conversation content for that portion of the posts could be "shoulder exercises at home."

[0015] In other words, the recommended conversation content can be determined based on similar posts in the saved list. For example, at least one post displayed on the saved list page can be for a specific content collection, or at least one post can be for multiple content collections, where each content collection includes at least one post.

[0016] In this context, a content collection corresponds to a content tag (that is, it represents a collection scenario, such as "fitness" or "fashion"). A recommended conversation content is the conversation content related to the corresponding content tag that is determined for the published content in a content collection (for example, if the content tag is "fitness", the recommended conversation content could be "home shoulder workout", and if the content tag is "fashion", the recommended conversation content could be "commuter outfits").

[0017] Optionally, content collections and content tags can be determined intelligently or by the target account; this is not limited here. For example, the target account can create content collections and add collected published content to the corresponding content collections (e.g., published content 1 is added to content collection 1, and published content 2 is added to content collection 2). In this case, the content tags corresponding to the content collection can be specified by the target account (e.g., the collection name, such as "fitness"), or they can be determined intelligently, such as by using an intelligent model to analyze the published content in the content collection to obtain the corresponding content tags.

[0018] For example, content collections can be intelligently obtained by classifying collected published content. For instance, an intelligent model can be used to cluster collected published content to obtain published content under multiple categories. Published content under a category can be a content collection, and content tags can be extracted from the published content under a content collection.

[0019] The target intelligent object refers to a virtual object (such as an intelligent robot) associated with an intelligent model (such as a large language model). This intelligent model can generate corresponding responses by processing the input text (for example, when initiating a conversation, it generates corresponding responses based on the input). Therefore, this application allows users to initiate an intelligent conversation with an intelligent object by selecting target recommended conversation content on the favorites page (such as relevant information about the target recommended conversation content). At this time, the intelligent model corresponding to the target intelligent object generates target response content based on the received information, and then displays the conversation page, where the target response content can be displayed.

[0020] Among them, the Big Prophecy model refers to a pre-trained model with massive parameters trained based on a deep learning architecture. In this application, it is mainly used to understand the deep semantics of the content collected by users to generate recommended conversation content, analyze implicit preferences that represent the user's account characteristics (such as clothing style), and generate corresponding reply content for the recommended conversation content.

[0021] For intelligent models, generating target response content first requires identifying relevant published content, specifically target published content related to the target recommended session content. When constructing the target prompt text from the target recommended session content, the intelligent model can be instructed to perform content analysis on the target published content based on the prompt text to obtain target response content tailored to the target recommended session content. For example, if the target recommended session content is the query intent "home shoulder exercises" determined from published content within the "fitness" content collection, then related published content can be identified from the "fitness" content collection, and content analysis can be performed on this related content to obtain target response content (such as text tutorials or corresponding video clips for "home shoulder exercises").

[0022] The target response content may include: text content compiled in response to the target posting, multimedia content (such as related videos, images, and audio), and component content (such as product components, location components for relevant online POI locations, route components derived from multiple locations, and map components that display map information). No specific restrictions are placed on the specific content types within the target response content.

[0023] For example, if the target recommendation conversation asks a question about fitness tutorials, the target response could include text tutorials, related videos, etc. Similarly, if the target recommendation conversation asks a question about item recommendations, the target response could include text descriptions of the recommended items, related images, related products, etc. And if the target recommendation conversation asks a question about travel guides, the target response could include text guides, specific routes, etc. It can be understood that this is determined by the intelligent model based on the specific type of the target recommendation conversation content.

[0024] The term "in response to" indicates a state where a corresponding event occurs or a condition is met. The timing of subsequent actions performed in response to this event or condition is not necessarily strongly correlated with the time when the event occurs or the condition is met. For example, "in response to XX operation, execute the target step" means that the target step is triggered upon detection of "XX operation".

[0025] For example, in some cases, subsequent actions can be performed immediately when the event occurs or the condition is met; while in other cases, subsequent actions can be performed some time after the event occurs or the condition is met, meaning there can be an intermediate process between responding to the corresponding event or condition and the subsequent actions performed.

[0026] The term "trigger operation" refers to an action performed on certain information (such as a control). This operation sends a corresponding instruction to the electronic device, triggering it to execute the next task. The next task triggered by different trigger operations on different information can all be pre-set. This trigger operation can be a contact-based operation such as clicking, double-clicking, long-pressing, or swiping, or other preset event operations, or it can be a non-contact operation (such as voice instructions). In some cases, the trigger operation can also be executed by the electronic device based on a pre-defined program. No further limitations are specified here.

[0027] One of the network architecture diagrams proposed in the content display method is as follows: Figure 1 As shown, the network architecture may include server 100 (the number of servers is not limited) and a cluster of terminal devices (the number of terminal devices is not limited, such as terminal device 200a, terminal device 200b, ..., terminal device 200n), wherein communication connections may exist between servers. Simultaneously, a server may have a communication connection with any terminal device, so that the server can interact with the terminal device through this communication connection. The aforementioned communication connection is not limited in method; it can be a direct or indirect connection via wired communication, a direct or indirect connection via wireless communication, or other methods, which are not limited herein. Furthermore, it is understood that the electronic devices involved in the embodiments of this application may be... Figure 1 The terminal device shown can also be Figure 1 The server shown.

[0028] It should be understood that, such as Figure 1 Each terminal device in the terminal device cluster shown can have an application client installed for content display. This application client can be of any type, such as a social networking client, instant messaging client (e.g., a conferencing client), entertainment client (e.g., a live streaming client), multimedia client (e.g., a video client), information client (e.g., a news client), shopping client, or any other client capable of displaying text, images, audio, and video data. No specific type of application client is limited here.

[0029] For example, an application client refers to a client that can send and receive internet messages instantly and has information functions. Specifically, terminal devices (such as…) Figure 1 The application client on the terminal device 200a) shown displays a map page and selects at least two routes on the map page to instruct the server to generate a target route based on the at least two routes, which is then displayed on the map page by the application client.

[0030] Optionally, the aforementioned terminal devices and servers can be logically separated. Therefore, when referring to terminal devices and servers below, they may be physically the same device or different devices.

[0031] For further information, please refer to [link / reference]. Figure 2 , Figure 2 This is a schematic diagram illustrating a content display scenario provided in an embodiment of this application. A favorites page 20a can be displayed, which includes at least one saved published content. For example, this favorites page is used to view published content under a specific content collection (such as "outfits"). At this time, a recommended conversation content area 20b can be displayed on the favorites page. This recommended conversation content area displays at least one recommended conversation content derived from the published content in the content collection. This recommended conversation content represents the user's query intent regarding the published content in the favorites content collection.

[0032] At this point, the target recommended session content 20c can be selected to send the target recommended session content to the target smart object, thereby initiating a smart session for the target smart object.

[0033] This includes a conversation page 21a that can be displayed with the target smart object, and the target response content that can be displayed on the conversation content recommended to the target.

[0034] Optionally, the conversation page can also display a conversation entry 21b to continue the conversation. For example, a question can be asked, and the target smart object can generate a corresponding reply based on the question, the replies already displayed on the conversation page, and the posts on the favorites page.

[0035] Therefore, the collection page not only allows users to view collected posts, but also recommended conversation content, enhancing the richness of the display and stimulating users to engage in intelligent conversations about collected posts. This not only improves interactivity but also allows users to view useful information more quickly and conveniently.

[0036] In specific embodiments of this application, scenarios involving the acquisition of user information and related data, such as the acquisition of user information (e.g., published content), require user permission or consent. That is, when these embodiments are applied to specific products or technologies, the collection, use, and processing of relevant user data comply with the relevant laws, regulations, and standards of the relevant regions. For example, interactive pages can be used to provide prompts indicating which data will be collected or acquired. Specifically, lists or other methods can be used to present the types and content of this data to the user. Data collection and processing will only proceed after a confirmation or instruction to allow data collection is received on the interactive page.

[0037] The scenarios described above are merely examples and do not constitute a limitation on the application scenarios of the technical solutions provided in this application. The technical solutions of this application can also be applied to other scenarios. For example, as those skilled in the art will know, with the evolution of system architecture and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0038] Based on the above description, this application proposes a content display method, which can be executed by the aforementioned electronic device, specifically... Figure 1 The terminal device shown. Please refer to [link / reference]. Figure 3 , Figure 3 A flowchart illustrating a content display method provided in this application embodiment. Figure 1 .like Figure 3 As shown, the flow of the content display method in this application embodiment may include the following: S101, Display the favorites page.

[0039] The collection page can be used to display at least one collected published content. For example, the collection page may include at least one of the following: a first collection page for displaying at least one published content in the order of collection; a first collection page for displaying at least one collection of content derived from at least one published content; and a third collection page for displaying at least one published content included in a collection of content. No further limitations are imposed here.

[0040] For example, a collection page can display all collected posts in chronological order, or it can display one or more collections of content derived from collected posts. A content collection must include at least one post. This collection can be used to represent a type of content, such as fitness or fashion. Optionally, at least one content collection can be a collection obtained by categorizing at least one post, or a collection created by the target account.

[0041] For example, it could be multiple content collections generated by intelligently categorizing collected published content, such as fitness-related collections, food-related collections, etc. When collecting published content later, users can specify the content collection containing that content, or it can be intelligently added through content matching. For instance, when collecting published content A, the relevance between published content A and existing content collections can be determined (such as relevance to published content within the collection, or relevance to the content tags corresponding to the collection), and published content A can be added to the corresponding content collection based on similarity. Alternatively, if it is determined that published content A is irrelevant to any existing content collections, a new content collection can be created to house published content A.

[0042] Optionally, the content tags for a content collection can represent the collection scenario of the content collection. For example, they can be intelligently determined based on the published content in the content collection, such as representing "fitness" or "food".

[0043] For example, content collections can be created by the target account. The target account can create a content collection when saving and publishing content, specifying the collection to which the published content belongs. In this case, the content tags for the content collection can be the collection name configured by the target account, or they can be intelligently determined based on the published content within the collection. Alternatively, after the target account creates a content collection, the saved published content can be automatically added to the corresponding content collection, such as based on the relevance between the content tags and the content collection. In this case, the content tags for the content collection created by the target account can be configured by the target account.

[0044] The collection page can also display content published in only one collection, such as displaying the published content in a collection in the order of collection.

[0045] It is understandable that the favorites page can display a recommended conversation content area, which includes at least one recommended conversation content. The recommended conversation content is determined based on at least one published content displayed on the favorites page.

[0046] For example, at least one recommended conversation content can be determined by one or more content collections on the favorites page. A content collection corresponds to a content tag, a recommended conversation content is associated with a content collection, and a recommended conversation content is conversation content related to the corresponding content tag determined based on at least one published piece of content from the corresponding content collection.

[0047] For example, the favorites page is the first favorites page, and at least one published content on this first favorites page involves published content belonging to content collection 1 and published content belonging to content collection 2. Therefore, the recommended conversation content area can display recommended conversation content determined by the published content of content collection 1, and / or, recommended conversation content determined by the published content of content collection 2. That is, users can view recommended conversation content related to content tags of multiple content collections. When displaying the favorites page, it is necessary to request display data for at least one published content, and at this time, recommended conversation content can be requested simultaneously.

[0048] Optionally, the second collection page can display multiple content collections. For example, viewing a collection on the first collection page can lead to the corresponding second collection page, thus switching at least one published content on the first collection page to display multiple content collections. When displaying the collection page, it is necessary to request display data for each content collection and at least one published content within each content collection. At this time, the recommended conversation content for each content collection can be requested simultaneously.

[0049] At this point, the second collection page displays at least one recommended session content derived from each content collection (e.g., one recommended session content area corresponds to one content collection). For example, it can display the corresponding recommended session content for each content collection separately, or only display the recommended session content for a specific content collection (e.g., displaying only one recommended session content area; for example, in response to a selection operation for a specific content collection, the recommended session content area switches to display the recommended session content for that specific content collection; the selection operation could be a long press, double-click, etc.). Alternatively, when a specific content collection is selected, the recommended session content for that specific content collection can be requested from the server.

[0050] Optionally, the third collection page is a collection page for a specific content set, that is, a page that displays the published content in a content set. In this case, the third collection page displays recommended conversation content determined based on the published content of that content set.

[0051] For example, in response to a view operation on a favorites page (such as a third favorites page), a page view request is generated; the page view request is sent to the server so that when the server obtains at least one published content to be displayed on the favorites page (or, for the second favorites page, the collection of displayed content), it extracts the multimodal features of at least one published content, and generates at least one recommended session content for the target account for at least one published content based on the multimodal features (i.e., extracts multimodal features for at least one published content in a content collection to generate at least one recommended session content for that content collection).

[0052] Therefore, when viewing the favorites page, the system can dynamically determine the user's current potential viewing intent (such as "today's training" or "commuting outfit") in real time. Optionally, when predicting viewing intent, in addition to combining at least one published content, information such as the current target account's search history and interaction data can be further obtained to comprehensively determine the intent. For example, if the search term in the target account's search history is "Kuaishou weight loss meal", when predicting the viewing intent for the "fitness" scenario, it can be determined that the user is currently looking for weight loss meal recipes related to "fitness", so the generated recommended conversation content could be "fitness weight loss meal recipes", etc. Clicking on any recommended conversation content (usually a capsule-shaped shortcut button) can trigger the generation of a dynamically aggregated page (i.e., the target reply content in the conversation page).

[0053] Dynamic page generation refers to a technology where the server does not pre-store a fixed page layout, but instead generates and assembles relevant response content (such as video players, images, and text paragraphs) and renders a personalized page in real time based on the user's real-time query intent and account characteristics.

[0054] When the terminal device receives at least one published content and at least one recommended session content from the server, it can display the corresponding favorites page, and display a content card (or content collection list) of at least one published content and a recommended session content area containing at least one recommended session content on the favorites page.

[0055] Among them, recommended conversation content refers to the viewing intent of the published content in a certain content collection recommended to the target account. For example, it is based on the high-frequency keywords that appear in the published content, the objects (such as people, actions, locations, etc.) that appear in the videos included in the published content, and common search terms related to the content scene (i.e. content tags) in which the published content is located.

[0056] For example, an intelligent model can generate at least one recommended conversation content based on published content in a content collection or the content tags of a content collection.

[0057] Specifically, multimodal features can be extracted from published content within a content collection using intelligent models. This involves determining multimodal features based on multimodal data (such as text, video, images, and audio) within the published content. Multimodal analysis, which refers to the technology of jointly processing and understanding two or more data formats such as text, images, video, and audio within the published content, is another example. In this solution, it specifically refers to the comprehensive analysis of visual features (such as clothing styles and fitness movements) and textual features (such as titles and body text) within the published content to obtain multimodal features.

[0058] For example, it can be divided into visual analysis and text analysis. Visual analysis can identify objects (such as dumbbells, wide-leg pants), human characteristics (such as gender, body type), and style atmosphere (such as retro, minimalist) in images / videos. Text analysis can extract high-frequency words in titles and body text, such as "slimming", "muscle gain", "dormitory". In other words, it can extract multimodal features for the content published in a content collection, thereby sorting out high-frequency keywords (such as the word "home" appearing multiple times) or frequently occurring objects (such as fitness movements that appear multiple times, such as "shoulder exercises").

[0059] Alternatively, common search terms can be generated based on the content scenario. For example, search records related to the content scenario can be obtained (e.g., if the content scenario is fitness, the search records could contain the term "fitness"). In this case, common search terms can be determined based on the search terms in the search records. For example, for the fitness scenario, common search terms could be beginner tutorials, exercise for weight loss, etc.

[0060] Therefore, based on the published content in a content collection, we can determine the user's viewing intent when viewing the published content of that content collection or collecting published content related to the content tags of that content collection. This viewing intent can then be used as recommended conversation content to attract users to initiate intelligent dialogue by selecting recommended conversation content. Furthermore, users can quickly view the key content extracted from the published content without having to view it themselves, thus improving the efficiency and accuracy of information extraction in content collection scenarios.

[0061] S102, in response to the selection operation of the target recommended session content in at least one recommended session content, send the target recommended session content to the target smart object and display the session page with the target smart object.

[0062] Specifically, a target recommended conversation content can be selected from at least one recommended conversation content. For example, triggering the target recommended conversation content will initiate a smart dialogue with the target smart object, thereby displaying the conversation page. Optionally, the target recommended conversation content is associated with content tags, and the content analysis for the target recommended conversation content is performed on at least one published piece of content from the content collection corresponding to the associated content tags.

[0063] S103. Display the target response content for the recommended conversation content on the conversation page.

[0064] The target response content is determined based on the target published content, that is, it is obtained by content analysis of the target published content. The target published content is at least one published content that is related to the target recommended session content, that is, at least one published content related to the target recommended session content in the content set corresponding to the associated content tags.

[0065] For example, in response to a selection operation targeting recommended session content, a content query request for that recommended session content can be generated. This content query request is then sent to the server, which generates target prompt text based on the recommended session content. This target prompt text instructs the target intelligent object to obtain the target response content based on the recommended session content through content analysis.

[0066] At this point, target published content related to the target recommended session content can be obtained from at least one published content. The target prompt text and target published content are then input into the target smart object. The target smart object then queries the target published content based on the target prompt text to find the target reply content related to the target recommended session content (i.e., it calls the smart model associated with the target smart object to perform content analysis on the target published content based on the target prompt text).

[0067] When the terminal device receives the target response content returned by the server, it displays a session page and shows the target response content on the session page.

[0068] Optionally, the conversation page can also display a conversation entry to continue asking questions. For example, in response to a question asked on the conversation page regarding the target response, the question can be sent to the target smart object to display the response to the question on the conversation page.

[0069] Optionally, the response to a question can be determined based on at least one published content (i.e., all published content on the collection page; or, published content in the collection of content corresponding to the associated content tags; or it can be only the target published content) and the target response content already displayed on the conversation page.

[0070] Optionally, if the collection page displays at least one published content, the published content can be filtered on the collection page, such as by time or by keywords.

[0071] For example, if a target account only wants to engage in smart conversations with posts saved within the last 3 days, it can select the corresponding time filter on the save page, which will then display posts that meet the time filter criteria. Similarly, if a target account only wants to engage in smart conversations with posts containing a specific keyword, it can enter that keyword on the save page (e.g., a specific series of fitness videos by a particular blogger), which will then display posts containing that keyword.

[0072] Optionally, when at least one reference post is displayed on the Favorites page, the recommended conversation content in the recommended conversation content area of ​​the Favorites page can also be dynamically updated accordingly. That is, for the currently displayed at least one reference post, the corresponding reference recommended conversation content is determined and displayed (the method of determination is the same as the method of determining the recommended conversation content described above).

[0073] Alternatively, intelligent dialogue can also be conducted by selecting target reference recommended conversation content. In this case, the corresponding reply content can be obtained based on the published content related to the target reference recommended conversation content from the reference published content filtered on the collection page, and displayed on the conversation page.

[0074] Therefore, compared to content collected in reverse chronological order, users often "save but don't view" or find it "difficult to utilize." This application aims to activate users' "dormant collection data" by intelligently analyzing collected content, automatically predicting viewing intentions, and initiating intelligent dialogues based on those intentions regarding the collected content.

[0075] For example, such as Figures 4a-4f As shown, Figures 4a-4f This application provides a schematic diagram of a content collection scenario; wherein, in Figure 4a In the example of a target account creating a content collection and specifying the addition of collected published content to the corresponding content collection, the target account can collect the content by triggering the collection control 40b when viewing the content details page 40a of the published content. At this time, the content addition prompt message 40c can be displayed to indicate whether to add the collected published content to the corresponding content collection.

[0076] For example, if the content collection page 40d can be displayed, showing a list of content collections, including those already created, the target account can choose to create a new content collection or add published content to an existing one (e.g., to the "Outfit" content collection). After successful addition of published content, a content collection viewing entry 40e can be displayed on the content details page. Triggering this entry will take the user to a third-party collection page for viewing the published content within that collection (the display format can be the same). Figure 2 The example is the collection page 20a).

[0077] Among them, Figure 4bTaking the first collection page 41a as an example, if the first collection page can be the account page of the target account, the collected published content can be viewed through the account page. In this case, while the first account page displays all published content in chronological order of collection time, it can also display a recommended conversation content area 41b. This recommended conversation content area can display at least one recommended conversation content determined by the content collection to which the collected published content belongs. This at least one recommended conversation content can come from multiple content collections. For example, the recommended conversation content determined by the "fitness" content collection is "daily training," and the recommended conversation content determined by the "outfit" content collection is "commuting outfits."

[0078] Among them, Figure 4c In the example of the second collection page 42a, the collection page can be obtained by performing a content collection viewing operation on the first collection page 41a. For example, if the collection viewing entry 42b in the first collection page is triggered, various content collections can be switched and displayed on the first collection page, thus obtaining the second collection page.

[0079] Optionally, on the second collection page, each content collection is associated with a recommended conversation content area. This area displays recommended conversation content specific to the corresponding content collection. For example, the recommended conversation content area 42c associated with the "Fitness" content collection displays recommended conversation content for the "Fitness" content collection (such as "Daily Fitness"). Similarly, the recommended conversation content area 42d associated with the "Fashion" content collection displays recommended conversation content for the "Fashion" content collection (such as "Commuting Outfit"). When a recommended conversation content is triggered, the target reply content can be determined for at least one published content within the corresponding content collection.

[0080] Among them, Figure 4d In the second collection page 42a, it can also display only a recommended session content area 42e. The recommended session content can be triggered by selecting a content collection on the second collection page. For example, long-pressing the "Outfit" content collection requests the server to generate corresponding recommended session content based on the "Outfit" content collection, and displays the recommended session content corresponding to the "Outfit" content collection (such as "Commuting Outfit") in the recommended session content area.

[0081] Among them, Figure 4eIn the example of the third collection page 43a, the content of a specific collection 43b can be viewed on the second collection page 42a. For example, clicking on the "Outfit" content collection will take you to the third collection page 43a of that content collection. The third collection page can display the published content in the "Outfit" content collection, as well as a recommended conversation content area 43c. The recommended conversation content in this recommended conversation content area is obtained from the published content in the specific content collection displayed on the third collection page.

[0082] Among them, Figure 4f Taking the third collection page (43a) as an example, content can be filtered on the third collection page. For example, you can select published content collected within a specified time period, such as selecting to view published content collected only within the last 7 days. After the reference published content is displayed in the filter on the third collection page, the reference recommended session content determined by the reference published content can be dynamically updated in the recommended session content area (43c). At this time, any reference recommended session content can be triggered to enter the session page for smart object processing.

[0083] Optionally, in addition to filtering published content by time, the third collection page can also display a keyword filtering entry 43d, which can be triggered to filter content by entering keywords.

[0084] In this embodiment, at least one published content that has been collected can be displayed on the collection page, along with at least one recommended conversation content obtained from the at least one published content. The recommended conversation content can be regarded as the user's possible query intent for the at least one published content (such as the query intent for content related to a certain fitness movement). In other words, the query intent that the user may have when collecting these published contents can be predicted and displayed in advance, so that the user can quickly engage in intelligent dialogue, thereby improving the richness of the content displayed on the collection page. In addition, by selecting the target recommended conversation content (i.e. the target query intent), an intelligent dialogue can be initiated with the target intelligent object, and the target reply content determined from the target published content related to the at least one published content can be viewed on the conversation page. That is, the user can quickly view the useful content extracted from the at least one published content through intelligent dialogue, without the user having to search for it himself. This can further improve the convenience and efficiency of the user in viewing the published content on the collection page and finding the information he / she wants to know.

[0085] Based on the above description, this application proposes a content display method, which can be executed by the aforementioned electronic device, specifically... Figure 1 The terminal device shown. Please refer to [link / reference]. Figure 5 , Figure 5A flowchart illustrating a content display method provided in this application embodiment. Figure 2 .like Figure 5 As shown, the flow of the content display method in this application embodiment may include the following: S201. Display the Favorites page, showing at least one saved post, a recommended conversation section, and a summary section.

[0086] The recommended session content area includes at least one recommended session content, which is determined based on at least one published piece of content. For details on how the recommended session content is determined, please refer to the relevant description in the above embodiments.

[0087] The summary section displays summary information about the content identified from at least one published content within at least one content collection. Therefore, in addition to displaying recommended conversation content identified from at least one published content, the favorites page can also display summary information about the content identified from at least one published content. For example, it could be based on content collections, determining summary information for each content collection.

[0088] The summary content area can be displayed in conjunction with the recommended conversation content area (e.g., Figure 2 The example collection page displays not only a recommended conversation content area but also a summary content area. The recommended conversation content area displays at least one recommended conversation, while the summary content area displays summary information such as "AI Summary," "Style Positioning: Your aesthetic keyword is Japanese retro," and "Color Aesthetics: Your signature color scheme is earth tones..." This summary information can characterize the account profile of the target account within the content collection. In other words, by analyzing the published content within the collection, a collection profile can be obtained, essentially building a preference profile based on the user's published content within a specific collection. Unlike user profiles determined by user information on the platform, this preference profile focuses more on the content collection characteristics of a vertical field (e.g., in a fashion content collection, the account profile includes "Gender: Male," "Style: Japanese," and "Fit: Loose," etc.).

[0089] Therefore, this summary content area can be displayed in conjunction with the recommended session content area. This summary content area displays summary information about the collection of content corresponding to the recommended session content in the associated recommended session content area.

[0090] When the collection page displays the published content of multiple content collections (such as the first collection page mentioned in the above embodiment), the collection page can display content summary information for each content collection, or display target content summary information obtained by merging the content summary information of multiple content collections.

[0091] Alternatively, when the collection page displays multiple recommended conversation content areas (one content collection corresponds to one recommended conversation content area), each recommended conversation content area can be associated with a summary content area, which displays the content summary information of the content collection corresponding to the associated recommended conversation content area.

[0092] The content summary information can be used to represent a summary of at least one published content in a content collection, so that users can quickly understand at least one published content in the content collection. For example, for a content collection under the fitness tag, the summary content summary could be: "XX published content mentions shoulder training, mainly in the home setting; XX published content mentions exercise and fat-loss recipes, mainly for office workers," and so on.

[0093] Optionally, the content summary information can also represent a type of account characteristic of the target account within a content collection. In other words, the account profile (such as gender, clothing preference, favorite clothing colors, etc.) of the target account within a content collection can be obtained through the content published within that content collection. That is, the content summary information is used to indicate at least one type of account characteristic of the target account in at least one content collection, with one content collection corresponding to one type of account characteristic.

[0094] For example, account characteristics of a target account can be generated from the multimodal features of published content within a content collection, and content summary information can be generated based on these account characteristics. Alternatively, an intelligent model can determine the multimodal features of a content collection based on its published content; then, recommended conversation content and content summary information for the content collection can be generated using these multimodal features.

[0095] S202: In response to the selection operation of target recommended session content from at least one recommended session content, send the target recommended session content and content summary information to the target smart object to display the session page with the target smart object. For a detailed implementation of step S202, please refer to the relevant description in the above embodiments.

[0096] S203. Display on the conversation page the target reply content that is related to the target recommended conversation content and conforms to the account characteristics of the target account indicated by the content summary information.

[0097] The target response content is determined based on the target published content, such as response content that matches the characteristics of the target account and is determined based on the target published content.

[0098] Among them, the target account features are the account features corresponding to the target content collection associated with the target recommended session content, and the target published content is the published content related to the target recommended session content from at least one published content included in the target content collection.

[0099] In other words, you can combine the target recommended conversation content and content summary information (or the target account characteristics indicated in the content summary information, such as "style positioning: Japanese retro," "color aesthetics: earth tones," etc.). Based on this, you can construct target prompt text. This prompt text instructs the target intelligent object to align its generated response content with the target account characteristics indicated in the content summary information. For example, when generating responses related to "commuting outfits," the primary focus would be on generating outfit recommendations that fit the "Japanese retro" style. This makes the target response content more accurate and more likely to reflect what the user wants to see.

[0100] Specifically, the target response content is determined through content analysis of at least one published content. For example, if the target recommended conversation content is "shoulder exercises at home," then the target response content could be text tutorials and videos related to "shoulder exercises at home." Optionally, the target response content can also be tailored to account characteristics. For instance, if the account characteristics indicate that the target account is female, then the determined text tutorials could be more suitable for women, and the selected videos could be videos by female bloggers.

[0101] For example, if the target recommended conversation content is "commuting outfits," then the target response content can be text descriptions and images or products related to "commuting outfits." Optionally, the target response content can also be tailored to account characteristics. For instance, if the account characteristics indicate that the target account's style is Japanese retro, then the determined text description can be outfit recommendations related to Japanese retro, and the selected images or products can be related to Japanese retro.

[0102] Therefore, the target response content varies depending on the target recommended session content and account characteristics. Furthermore, the content type within the target response content can differ based on the type of query intent represented by the target recommended session content. For example, for a tutorial-related query intent, a text tutorial and corresponding video can be displayed. Alternatively, for a recommendation-related query intent, a text description (or highlighting keywords related to the product and linking them to search components) and corresponding images (or product components, etc.) can be displayed. There are no specific limitations; the intelligent model generates the response content based on the actual query scenario. Compared to generating response content according to a specified template, this approach is more flexible and adaptable to different query scenarios, improving the user experience.

[0103] For example, such as Figures 6a-6f As shown, Figures 6a-6f This is a schematic diagram illustrating a display scenario of reply content provided in an embodiment of this application; wherein, in Figure 6aIn this context, the target response content may include a target image identified from at least one published content that is relevant to the target recommended session content.

[0104] For example, if the target recommended conversation content is "commuting outfits," the target response content could be some recommended commuting outfits and matching images. For instance, the intelligent model could output target text based on at least one published content, and subsequently, an image matching model could be used to select images from at least one published content that match the target text (e.g., the target text includes a description of trench coat recommendations, and the image could be a matched image that matches the account characteristics of the target account).

[0105] Optionally, in response to a viewing operation of the target image 60b on the session page 60a, the content source of the target image and / or the product entry associated with the target image may be displayed.

[0106] Optionally, triggering the target image allows you to view the content source of the target image to determine the target published content in which the target image is located. At this time, you can enter the content details page of the target published content through the content source.

[0107] Optionally, if the target content of the target image also includes a target product (i.e., the target product is the product attached to the target content of the target image), the product entry can also be used to view the target product. For example, the product link (i.e., product component) of the target product can be displayed on the conversation page in association with the target image, triggering entry into the product details page, so that the target reply content can be associated with related products, and users can more conveniently view and purchase products, thereby increasing the product transaction rate.

[0108] Optional, in Figure 6b In the process, the product entry point is also used to trigger a product search. For example, a search control 61b can be displayed for the target image 61a. By triggering the search control, the user can enter the search results page 61c for the target image. The search results page is used to display at least one similar product found for the target image, that is, to search for related products for the target image.

[0109] Optionally, a search operation for the target image can be triggered directly on the session page 60a. For example, by long-pressing the target image, a search results page for the target image can be displayed.

[0110] Optionally, the product entry associated with the target image can also be displayed directly on the session page, in which case the product entry can be directly triggered to search for products or jump to product links.

[0111] Among them, Figure 6cIn this context, the target response content includes target text determined from at least one published content that is relevant to the target recommended session content. Displaying the target response content for the target recommended session content on the session page 60a can be achieved by highlighting the product keywords 62a within the target text when displaying the target text on the session page (e.g., by displaying them as triggerable elements, such as underlining them, or by using a color distinct from other text).

[0112] For example, in a relevant recommendation scenario, the target response content can be used to recommend products. In this case, the target text can include product keywords. For instance, in a relevant product recommendation description, keywords related to the item will be included, and these item keywords can be considered as product keywords. Examples include "short trench coat" and "workwear jacket."

[0113] At this point, product keywords can be highlighted and linked to the search component. Therefore, by triggering a product keyword, users can access the search results page 62b for that keyword. This search results page displays at least one product found using the product keyword. Users can then view the relevant products found, improving the convenience of product searching and viewing without having to exit the session page to access the product search interface or be redirected to another e-commerce platform to search and purchase.

[0114] Therefore, it is possible to display physical product mapping, which refers to the process of associating unstructured content features (such as products in an image or keywords in text) with specific products in the product library, thus enabling "search for similar items" or "product link redirection".

[0115] Among them, Figure 6d In this context, the target response content includes target video clips identified from at least one published content that are relevant to the target recommended session content. For example, for tutorial-type recommended session content, after generating relevant tutorial text, the intelligent model can extract video clips matching the tutorial text from at least one published content. For instance, a tutorial text for "Shoulder Exercises at Home" might include: Step 1: Warm-up; Step 2: Shoulder Exercises; Step 3: Stretching. In this case, corresponding video clips can be extracted from at least one published content for each step and displayed as the target response content. Compared to plain text tutorial content, using adapted video clips allows for a more intuitive and effective way to view the desired content.

[0116] In addition, the target video clip is a video segment in the published content that matches the corresponding tutorial text (such as a set of shoulder exercise videos). This eliminates the need for users to manually open the video and drag the progress bar to find the corresponding video clip. Instead, when the target reply content is generated, the appropriate clip is automatically selected for playback, improving the convenience and efficiency for users to view relevant content.

[0117] For example, video segment matching and extraction can be performed using video temporal localization technology (also known as "video slicing technology"). Specifically, it refers to the technology of automatically locating the start and end times of an event in an unedited long video stream based on a specific semantic query (such as "lateral raise action"), thereby obtaining the corresponding target video segment.

[0118] Optionally, displaying the target response content corresponding to the target query intent block on the session page can be, when displaying the target video clip 63b on the session page 63a, displaying the source of the target video clip and the progress point of the target video clip.

[0119] The segment source is used to represent the target published content where the target video segment is located. Therefore, the segment source of the target video segment can be triggered to enter the content details page 63c of the target published content where the target video segment is located. At this time, the target video segment can be played at the segment progress point on the content details page, that is, jump to the content details page to watch the target video segment in the complete video.

[0120] For example, taking a fitness scenario, the conversation page can display saved posts tagged "fitness," a recommended conversation content area (63d), and a summary content area (63e). The recommended conversation content area displays at least one recommended conversation (such as "home shoulder workout"), while the summary content area displays summary information for the posts in the content collection (which can be used to indicate account characteristics, such as what the target account focuses on in a fitness scenario, and what fitness scenarios the target account typically engages in). At this point, triggering any of the recommended conversation posts will take you to the corresponding conversation page.

[0121] It is understandable that a segment progress point is used to indicate the playback progress point of a target video segment within the video included in the target published content, such as 03:30-04:00.

[0122] Optionally, for a target video segment, the matching published content (such as the high-quality published content among the matched published content) can be determined first based on the determined target text (such as the step text in the tutorial text), and the matching target video segment can be determined from the video of that published content.

[0123] Alternatively, one could first identify matching video clips from multiple published content based on the target text, and then select the video clips with better content quality as the target video clips.

[0124] Alternatively, one could first determine a certain number of matching published content items based on the identified target text, then identify matching video clips from each of these published content items, and finally select the video clips with the best content quality from the selected matching video clips as the target video clips. The method for determining the target video clips is not limited here.

[0125] Optionally, since a target video clip comes from a published content, it is also possible to view video clips similar to the target video clip in at least one published content on the favorites page, that is, to quickly find the video clips you want to learn about and view in other favorite published content.

[0126] For example, in response to a similar segment viewing operation for a target video segment, similar target video segments can be displayed; the similar target video segments are determined from target published content other than the target published content where the target video segment is located.

[0127] For example, in Figure 6e In the context of the target video segment, a similar video viewing entry 64a is associated with it. Triggering this entry leads to a similar video viewing page 64b, which displays similar target video segments identified from other published content. These other published content refers to target published content related to the recommended session content, excluding the target published content containing the target video segment.

[0128] Optionally, when displaying similar target video clips, the clip source and clip progress point can be displayed together, so as to display the content details page of other published content through the clip source and locate and view similar target video clips.

[0129] For example, if the target published content includes published content 1-10, and the target video clip is located in published content 1, then similar target video clips can be identified from published content 2-10. Alternatively, one could directly determine whether similar target video clips exist from published content 2-10 based on the target video clip.

[0130] Alternatively, it could be based on the target text corresponding to the target video segment to determine whether similar target video segments exist in the published content 2-10.

[0131] Optionally, users can not only view similar target video clips, but also edit the conversation page based on similar target video clips. For example, by performing a replacement operation on a similar target video clip (triggering the replacement control), a specified similar target video clip 64c can be used as the new target video clip and updated in the conversation page. That is, the original target video clip 64d is replaced with the specified similar target video clip on the conversation page. This allows the target account to edit the target reply content, thereby creating a more satisfactory reply (for example, when viewing replies related to "home shoulder exercises," if a target video clip is not satisfactory, it can be replaced with a more preferred video clip, and then viewed and followed according to the current reply content).

[0132] Optional, in Figure 6f Furthermore, the currently displayed target reply content can be saved on the conversation page 65a (e.g., by triggering the save control 65b) for viewing as independent media content. For example, if a user wants to view the "home shoulder workout" related reply content on the conversation page again, it can be saved. This could be done by saving the conversation page as a separate media content details page, or by sending the page data of the conversation page to the server, which then generates the media content for the target account.

[0133] Optionally, the media content can be stored on the target account's account page 65c and viewed there. Alternatively, the media content can be triggered on the account page 65d to access the corresponding content details page (displayed in the same format as the conversation page 65a) to view more information (i.e., the content displayed on the conversation page). When the media content is presented as independent content, the displayed content (i.e., the conversation history on the conversation page, including the target's reply) can be edited and modified on the content details page. Alternatively, further questions can be asked on the content details page, and replies can be based on the content displayed on the content details page. Optionally, replies can also be obtained by searching for relevant content on the current content platform or across the entire network (such as related published content, related articles, etc.).

[0134] It is understandable that the target account can edit and modify the content displayed on the conversation page, such as replacing the video in the reply, modifying the text, modifying the location in the route, etc. The target account can also save the content on the conversation page to obtain the associated media content (intelligent dialogue content). Later, the target account can access the content details page of the media content through the account page to view it (i.e., a new conversation page, where questions can be asked, and the questions are still retained in the content details page of the media content).

[0135] Therefore, this application can be specifically defined as follows: First, data analysis and profile building. When a user views the collection page, the published content in the relevant content collection can be analyzed as a whole through an intelligent model. Specifically, multimodal feature extraction can be performed to build account features based on multimodal features.

[0136] For example, in a clothing scenario, the user might be male and prefer a "Japanese retro" style. In a fitness scenario, the user might be interested in "shoulder training," and the training is often conducted in a "home / equipment-free" setting (e.g., recognizing a water bottle or a dormitory background).

[0137] The second step involves intelligent summarization and intent capsule recommendation. This includes generating corresponding summary text based on account characteristics of content collections, as well as generating query intents—that is, what the system predicts users might currently want to know. For example, an intelligent model can perform content analysis on at least one published piece of content based on the constructed prompt text to obtain summary text and query intents.

[0138] For example, in a fitness setting, it could be "Today's workout" or "Shoulder workout at home." In a fashion setting, it could be "Commuting outfit" or "Date outfit."

[0139] The third step, the interaction process, involves responding to commands and generating a dynamic details page (i.e., a conversation page). As the user clicks on different query intents, the system dynamically generates a completely new page by combining the "account profile indicated by the summary text" and the "current click intent." This improves the account profile for content collections, enabling a deeper understanding of the account profile. For example, by analyzing the content posted in a user's favorites, the system can extract the user's deep preferences (style, gender, taste, etc.). Thus, when a user searches for "commuting" related content, the system can provide suggestions that match their personal style.

[0140] Furthermore, dynamic scenario reconfiguration is possible, such as combining a user's "long-term preferences" with their "current intentions" (e.g., wanting to commute, wanting to work out shoulders) to generate customized pages, improving the accuracy and adaptability of content analysis. Specifically, multiple posts in a user's favorites are aggregated and analyzed to extract common preference features (e.g., gender, specific style, specific taste); when a user's specific scenario-based instruction (e.g., commuting, weekend, etc.) is received, the scenario requirements are combined with common preferences to generate a solution that matches the user's unique taste, solving the problem of "one-size-fits-all" in traditional search / recommendation. For example, when clicking on "commuting," a girl who likes Korean style and a boy who likes Japanese workwear will see completely different recommendation results, greatly improving content matching and user purchase intention.

[0141] Specifically, for published content containing videos, multimodal fine-grained processing can be achieved, automatically extracting video segments and improving the utilization rate of long video content. For example, in fitness or tutorial scenarios, users may only be interested in a specific movement in the video (such as the "lateral raise" in the 3rd minute of a 10-minute video). This application eliminates the need for users to manually open the video in the published content and drag the progress bar to find it, improving the user experience. This enables fine-grained slicing and logical reorganization of video content, that is, it can identify specific semantic segments in long videos (such as specific fitness movements or specific cooking steps) and obtain their timeline information; then, multiple video segments from different video sources are displayed sequentially according to the step order indicated by the target text (such as warm-up, training, stretching). This can transform unstructured long videos into structured "atomic knowledge," significantly improving the efficiency of users in follow-up and learning scenarios and solving the pain point of "difficult progress bar dragging."

[0142] Furthermore, for recommendation-based scenarios, products can be automatically associated, such as displaying product links or triggering product search functions, to directly establish a connection with the product library through the session page, enabling product search or purchase. This achieves cross-modal entity mapping and a closed-loop business model, extracting matching product images from user-collected published content and mapping them to search criteria in the e-commerce system, realizing automatic redirection from "product image" to "product search" or "product purchase," thus bridging the final step from community content to e-commerce transactions.

[0143] In this embodiment, at least one set of published content and at least one recommended conversation content obtained from the at least one set of published content can be displayed on the collection page. The recommended conversation content can be regarded as the user's possible query intent for the at least one set of published content (such as the query intent for content related to a certain fitness movement). In other words, the query intent that the user may have when collecting these published contents can be predicted and displayed in advance, so that the user can quickly engage in intelligent dialogue, thereby improving the richness of the content displayed on the collection page. In addition, content summary information can also be displayed. The content summary information can characterize the account characteristics of the target account in a content collection. Thus, by selecting the target recommended conversation content (i.e. the target query intent) to initiate an intelligent dialogue with the target intelligent object, the user can view the target reply content that matches the target account characteristics determined from the relevant target published content of the at least one set of published content on the conversation page. That is, the user can quickly view the useful content extracted from the at least one set of published content through intelligent dialogue, and the extracted content matches the account profile obtained through the content collection. This makes the information viewed by the user more accurate and more in line with the user's characteristics, improving the user experience. In other words, it can further improve the convenience and efficiency of the user in viewing the published content on the collection page.

[0144] Based on the above description, this application proposes a content display method, which can be executed by the aforementioned electronic device, specifically... Figure 1 The terminal devices and servers shown are illustrated. Please refer to [link / reference]. Figure 7 , Figure 7 A flowchart illustrating a content display method provided in this application embodiment. Figure 3 .like Figure 7 As shown, the flow of the content display method in this application embodiment may include the following: S301. In response to the operation of viewing the favorites page, the terminal device generates a page viewing request and sends the page viewing request to the server.

[0145] S302. When the server obtains at least one published content, it extracts the multimodal features of at least one published content, generates at least one recommended session content for the target account in relation to at least one published content based on the multimodal features, and generates content summary information to indicate the account features of the target account.

[0146] S303. The server returns at least one published content, at least one recommended session content, and content summary information to the terminal device.

[0147] S304. The terminal device displays a favorites page, and displays at least one content card of published content, a recommended conversation content area containing at least one recommended conversation content, and a summary content area containing content summary information on the favorites page.

[0148] S305. In response to the selection operation for the target recommended session content in at least one recommended session content, the terminal device generates a content query request for the target recommended session content and sends the content query request to the server.

[0149] S306. The server generates target prompt text based on the account characteristics indicated by the target recommended session content and the content summary information, obtains target published content related to the target recommended session content from at least one published content, and inputs the target prompt text and target published content into the target smart object. The target smart object then queries the target published content based on the target prompt text to find target reply content that is related to the target recommended session content and matches the account characteristics.

[0150] S307. The server returns the target response content to the terminal device.

[0151] S308. The terminal device displays a conversation page, showing the target reply content for the recommended conversation content. The specific implementation methods for steps S301-S308 can be found in the relevant descriptions of the above embodiments, and will not be repeated here.

[0152] In this embodiment, the collection page can display at least one collected published content, as well as at least one recommended conversation content and content summary information obtained from the at least one published content. The recommended conversation content can be regarded as the user's possible query intent for the at least one published content, and the content summary information can characterize the account characteristics of the target account. That is, it can display the user's possible query intent when collecting these published contents and the obtained summary content, so that the user can quickly engage in intelligent dialogue, thereby improving the richness of the content displayed on the collection page. Thus, by selecting the target recommended conversation content to initiate an intelligent dialogue with the target intelligent object, the user can view the target reply content that is related to the target recommended conversation content and matches the target account characteristics on the conversation page. In other words, the user can quickly view the useful content extracted from at least one published content through intelligent dialogue, and the extracted content matches the account profile obtained through the content collection. This makes the information viewed by the user more accurate and tailored to the user's characteristics, thereby improving the user experience.

[0153] Please see Figure 8 , Figure 8 This is a schematic diagram of a content display device provided in an embodiment of this application. It should be noted that... Figure 8 The content display device shown is used to execute this application. Figure 3 , Figure 5 , Figure 7 The methods in the illustrated embodiments are shown only in the parts relevant to the embodiments of this application for ease of explanation; specific technical details are not disclosed. Reference to this application is required. Figure 3 , Figure 5 , Figure 7 The example shown. Figure 8 The example content display device 1000 may include: a page display module 1001 and a content display module 1002.

[0154] in: The page display module 1001 is used to display the collection page; the collection page includes at least one collection of published content and a recommended conversation content area, the recommended conversation content area includes at least one recommended conversation content, wherein the recommended conversation content is determined based on at least one collection of published content; Content display module 1002 is used to send the target recommended session content to the target smart object and display the session page with the target smart object in response to a selection operation of the target recommended session content in at least one recommended session content; The content display module 1002 is also used to display target reply content for the target recommended session content on the session page; wherein the target reply content is determined based on the target published content, and the target published content is published content related to the target recommended session content among at least one published content.

[0155] The collection page includes at least one of the following: a first collection page for displaying at least one published content in the order of collection; a first collection page for displaying at least one collection of content obtained from at least one published content; and a third collection page for displaying at least one published content included in a collection of content. Among them, at least one content collection is a collection obtained by categorizing at least one published content, or a collection created by the target account.

[0156] In this context, a content collection corresponds to a content tag, a recommended session content is associated with a content collection, and a recommended session content is a session content related to the corresponding content tag determined based on at least one published content in the corresponding content collection.

[0157] The collection page also includes a summary content area, which is used to display summary information of content determined by at least one published content in at least one content collection. The summary information is used to indicate at least one type of account characteristic of the target account in at least one content collection, with one content collection corresponding to one type of account characteristic. Among them, the target reply content is the reply content that conforms to the target account characteristics based on the target published content. The target account characteristics are the account characteristics corresponding to the target content collection associated with the target recommended session content. The target published content is the published content related to the target recommended session content from at least one of the published content included in the target content collection.

[0158] The page display module 1001 is also used for: In response to a content filtering operation on the Favorites page, the Favorites page displays reference published content filtered from at least one published content, and the recommended conversation content area on the Favorites page is updated to display at least one reference recommended conversation content obtained from the reference published content.

[0159] The content display module 1002 is also used for: In response to a question asked on the conversation page regarding the target reply content, a question is sent to the target smart object to display a reply to the question on the conversation page; the reply to the question is determined based on at least one published content and the target reply content.

[0160] The target response content includes a target image identified from at least one published content that is relevant to the target recommended session content; The content display module 1002 is also used for: In response to a viewing action on the target image on the session page, display the content source of the target image and / or the product entry associated with the target image; The content source is used to trigger access to the content details page of the target published content where the target image is located. The product entry is used to trigger access to the search results page for the target image or the product details page for the target product. The search results page is used to display at least one similar product found for the target image. The target product is the product attached to the target published content where the target image is located.

[0161] The target response content includes target text determined from at least one published content that is relevant to the target recommended conversation content; When displaying the target reply content for the target recommended conversation content on the conversation page, the content display module 1002 is specifically used for: When displaying target text on the conversation page, highlight the product keywords within the target text; Among them, product keywords are used to trigger access to the search results page for product keywords, and the search results page is used to display at least one product found by searching for product keywords.

[0162] The target response content includes target video clips identified from at least one published content that are relevant to the target recommended session content; The content display module 1002 is also used to display the target response content corresponding to the target query intent block on the session page, specifically for: When displaying the target video clip on the session page, show the source of the target video clip and the progress point of the target video clip; The segment source is used to locate the segment progress point on the content details page when the target video segment is triggered and enters the content details page of the target published content. The segment progress point is used to indicate the playback progress point of the target video segment in the videos included in the target published content.

[0163] The content display module 1002 is also used for: In response to a similar segment viewing operation for a target video segment, display similar target video segments; similar target video segments are determined from target published content other than the target published content where the target video segment is located.

[0164] Specifically, when displaying the favorites page, the page display module 1001 is used for: In response to a view request for a favorites page, generate a page view request; A page view request is sent to the server so that when the server obtains at least one published content, it can extract the multimodal features of at least one published content and generate at least one recommended session content for the target account based on the multimodal features. Upon receiving at least one published content and at least one recommended session content from the server, a favorites page is displayed, showing a content card for at least one published content and a recommended session content area containing at least one recommended session content.

[0165] Specifically, when the content display module 1002 sends the target recommended session content to the target smart object and displays the session page with the target smart object in response to a selection operation of the target recommended session content in at least one recommended session content, it is used for: In response to a selection operation on the target recommended session content, a content query request for the target recommended session content is generated; A content query request is sent to the server, which generates target prompt text based on the target recommended session content and retrieves target published content related to the target recommended session content from at least one published content. The target prompt text and target published content are input into the target smart object, which then retrieves the target reply content related to the target recommended session content from the target published content based on the target prompt text. The specific implementation methods of the page display module and the content display module can be found in the description of the above embodiments, and will not be repeated here. It should be understood that the beneficial effects obtained by using the same method will also not be repeated here.

[0166] Please see Figure 9 , Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 9As shown, the electronic device 2600 includes at least one processor 2601 and a memory 2602. Optionally, the electronic device may also include a network interface. The processor 2601, memory 2602, and network interface can exchange data. The network interface, controlled by the processor 2601, is used to send and receive messages. The memory 2602 stores computer programs, including program instructions. The processor 2601 executes the program instructions stored in the memory 2602. The processor 2601 is configured to invoke the program instructions to execute the aforementioned method. The memory 2602 may include volatile memory, such as random-access memory (RAM); the memory 2602 may also include non-volatile memory, such as flash memory, solid-state drive (SSD), etc.; the memory 2602 may also include combinations of the above types of memory.

[0167] Processor 2601 may be a central processing unit (CPU). In one embodiment, processor 2601 may also be a graphics processing unit (GPU). Processor 2601 may also be a combination of a CPU and a GPU. Processor 2601 may be used to invoke device control applications stored in memory 2602 to perform the above-described tasks. Figure 3 , Figure 5 , Figure 7 The description of the content display method in the corresponding embodiment can also be executed as described above. Figure 8 The description of the content display device in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated.

[0168] In specific implementations, the devices, processors, memory, etc., described in the embodiments of this application can execute the implementation methods described in the above method embodiments, or they can execute the implementation methods described in the embodiments of this application, which will not be repeated here.

[0169] This application also provides a computer-readable storage medium storing a computer program. The computer program includes program instructions, which, when executed by a processor, enable the processor to perform some or all of the steps described in the above method embodiments. Optionally, the computer storage medium can be volatile or non-volatile. The computer-readable storage medium may primarily include a program storage area and a data storage area. The program storage area may store an operating system, at least one application program required for a given function, etc.; the data storage area may store data created based on the use of blockchain nodes, etc.

[0170] This application provides a computer program product, which may include a computer program. When the computer program is executed by a processor, it can implement some or all of the steps in the above method, which will not be elaborated here.

[0171] In this article, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0172] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a computer storage medium, which can be a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the methods described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0173] The above-disclosed embodiments are merely some of the embodiments of this application, and should not be construed as limiting the scope of this application. Those skilled in the art can understand that all or part of the processes for implementing the above embodiments, and equivalent changes made in accordance with the claims of this application, still fall within the scope of this application.

Claims

1. A content display method, characterized in that, The method includes: The collection page is displayed; the collection page includes at least one collection of published content and a recommended conversation content area, the recommended conversation content area includes at least one recommended conversation content, wherein the recommended conversation content is determined based on the at least one collection of published content; In response to a selection operation of a target recommended session content from the at least one recommended session content, the target recommended session content is sent to the target smart object, and a session page with the target smart object is displayed. The target response content is displayed on the conversation page in response to the target recommended conversation content; wherein the target response content is determined based on the target published content, and the target published content is the published content related to the target recommended conversation content among the at least one published content.

2. The method according to claim 1, characterized in that, The collection page includes at least one of the following: a first collection page for displaying the at least one published content in the order of collection; a first collection page for displaying at least one content collection obtained from the at least one published content; and a third collection page for displaying at least one published content included in a content collection. The at least one content collection is a collection obtained by categorizing the at least one published content, or a collection created by the target account.

3. The method according to claim 2, characterized in that, A content collection corresponds to a content tag, a recommended session content is associated with a content collection, and a recommended session content is a session content related to the corresponding content tag determined based on at least one published content in the corresponding content collection.

4. The method according to claim 3, characterized in that, The collection page also includes a summary content area, which is used to display content summary information determined by at least one published content in the at least one content collection. The content summary information is used to indicate at least one type of account characteristic of the target account in the at least one content collection, and one content collection corresponds to one type of account characteristic. The target reply content is a reply content that conforms to the target account characteristics determined based on the target published content. The target account characteristics are the account characteristics corresponding to the target content collection associated with the target recommended session content. The target published content is the published content related to the target recommended session content from at least one published content included in the target content collection.

5. The method according to claim 1, characterized in that, The method further includes: In response to the content filtering operation on the favorites page, reference published content filtered from the at least one published content is displayed on the favorites page, and at least one reference recommended conversation content obtained from the reference published content is updated and displayed in the recommended conversation content area of ​​the favorites page.

6. The method according to claim 1, characterized in that, The method further includes: In response to a content-asking operation on the conversation page regarding the target reply content, a question is sent to the target smart object to display a reply to the question on the conversation page; the reply to the question is determined based on the at least one published content and the target reply content.

7. The method according to claim 1, characterized in that, The target response content includes a target image determined from the at least one published content that is related to the target recommended session content; The method further includes: In response to a viewing operation of the target image on the session page, the content source of the target image and / or the product entry associated with the target image are displayed; The content source is used to trigger access to the content details page of the target published content where the target image is located. The product entry is used to trigger access to the search results page for the target image or the product details page of the target product. The search results page is used to display at least one similar product found in the search for the target image. The target product is the product attached to the target published content where the target image is located.

8. The method according to claim 1, characterized in that, The target response content includes target text determined from the at least one published content that is related to the target recommended session content; The step of displaying the target reply content for the recommended conversation content on the conversation page includes: When the target text is displayed on the session page, the product keywords in the target text are highlighted; The product keyword is used to trigger access to the search results page for the product keyword, and the search results page is used to display at least one product obtained by searching for the product keyword.

9. The method according to claim 1, characterized in that, The target response content includes target video segments determined from the at least one published content that are related to the target recommended session content; The step of displaying the target response content corresponding to the target query intent block on the session page includes: When the target video clip is displayed on the session page, the source of the target video clip and the progress point of the target video clip are displayed. The segment source is used to locate the segment progress point on the content details page when the target video segment is triggered and enters the content details page of the target published content. The segment progress point is used to indicate the playback progress point of the target video segment in the video included in the target published content.

10. The method according to claim 9, characterized in that, The method further includes: In response to a similar segment viewing operation for the target video segment, similar target video segments are displayed; the similar target video segments are determined from target published content other than the target published content where the target video segment is located.

11. The method according to claim 1, characterized in that, The display of the favorites page includes: In response to a view operation on the bookmarked page, a page view request is generated; The page viewing request is sent to the server so that when the server obtains at least one published content, it extracts the multimodal features of the at least one published content and generates at least one recommended session content for the target account based on the multimodal features; Upon receiving at least one published content and at least one recommended session content from the server, the collection page is displayed, and the collection page displays the content card of the at least one published content and the recommended session content area containing the at least one recommended session content.

12. The method according to claim 1, characterized in that, The step of responding to a selection operation of a target recommended session content from the at least one recommended session content, sending the target recommended session content to the target smart object, and displaying a session page with the target smart object includes: In response to the selection operation for the target recommended session content, a content query request for the target recommended session content is generated; The content query request is sent to the server, so that the server generates target prompt text based on the target recommended session content and obtains target published content related to the target recommended session content from the at least one published content. The target prompt text and the target published content are input into the target smart object, and the target smart object queries the target reply content related to the target recommended session content from the target published content based on the target prompt text. Upon receiving the target response content from the server, the session page is displayed.

13. A content display device, characterized in that, The device includes: A page display module is used to display the collection page; the collection page includes at least one collection of published content and a recommended conversation content area, the recommended conversation content area includes at least one recommended conversation content, wherein the recommended conversation content is determined based on the at least one collection of published content; The content display module is configured to, in response to a selection operation of a target recommended session content among the at least one recommended session content, send the target recommended session content to the target smart object and display the session page with the target smart object; The content display module is further configured to display target reply content for the target recommended session content on the session page; wherein the target reply content is determined based on the target published content, and the target published content is the published content related to the target recommended session content among the at least one published content.

14. An electronic device, characterized in that, The method includes a processor and a memory, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to perform the method as described in any one of claims 1-12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, the computer program including program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1-12.

16. A computer program product, characterized in that, The method includes a computer program comprising program instructions that, when executed by a processor, implement the method according to any one of claims 1-12.

Images