Page generation method and device, electronic equipment, medium and computer program product

Abstract

The embodiment of the invention discloses a page generation method and device, electronic equipment, a medium and a computer program product, and the page generation method comprises the steps: determining the preference degree of each user for a page based on the page interaction data, obtained in real time, of each user; obtaining a user score set corresponding to each user based on the preference degree of each user for the page; the user score set comprises real-time scores of the user on the multiple pages obtained based on the preference degree of the user on the pages; determining similar users of the target user according to the similarity between the user score sets corresponding to the users; the target user is one of the users; obtaining real-time preference characteristics of the target user through the user score set corresponding to the target user and the user score set corresponding to the similar user; and updating the page of the target user in real time based on the real-time preference characteristics.

Description

Technical Field

[0001] The present application belongs to the field of computer technology, and in particular relates to a page generation method, device, electronic device, medium and computer program product. Background Art

[0002] Existing page layout and content arrangement methods often rely on fixed templates and manual adjustments, and are unable to dynamically adjust according to users' personalized preferences and real-time behaviors, resulting in the generated pages being unable to meet user needs and having low intelligence. Summary of the Invention

[0003] Embodiments of the present application provide a page generation method, apparatus, electronic device, medium, and computer program product.

[0004] This embodiment of the present application provides a page generation method, the method comprising:

[0005] Determining each user's preference for a page based on real-time page interaction data of each user; obtaining a user rating set corresponding to each user based on the user's preference for the page; the user rating set including real-time user ratings for multiple pages obtained based on the user's preference for the page;

[0006] Determine similar users of a target user based on similarities between user rating sets corresponding to each user; the target user is one of the users;

[0007] The real-time preference features of the target user are obtained through the user rating set corresponding to the target user and the user rating set corresponding to the similar users; and the page of the target user is updated in real time based on the real-time preference features.

[0008] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method also includes: based on the user rating set corresponding to the target user and the user rating set corresponding to the similar users, real-time training of the preference model is performed, so that the preference model can obtain the user's real-time preference data based on the user rating set; obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users includes: processing the user rating set corresponding to the target user and the user rating set corresponding to the similar users through the preference model to obtain the real-time preference characteristics of the target user.

[0009] As can be seen, real-time training of the preference model using the real-time user rating set helps improve the accuracy of the real-time preference features obtained by the preference model. Real-time updating of user pages using the real-time preference features output by the preference model helps achieve intelligent updating of user pages.

[0010] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method also includes: obtaining a page rating set corresponding to each page based on the preference degree of each user for the page; the page rating set includes real-time ratings of multiple users corresponding to the page obtained based on the user's preference degree for the page; determining the similarity between each page through the similarity between the page rating sets corresponding to each page; and updating the target user's page in real time based on the real-time preference characteristics, including: updating the target user's page in real time based on the real-time preference characteristics and the similarity between the each page.

[0011] As can be seen, the similarity between pages helps to further explore the preferences of target users. Based on similar pages, updating the current page helps to improve the accuracy of user page generation while meeting user preferences.

[0012] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method further includes: determining one or more page sets based on the similarity between each two pages; each page set of the one or more page sets includes two or more pages, and the similarity between each page in the page set is less than a page similarity threshold; obtaining the characteristics of each page set based on each page in the each page set; obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users includes: obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the characteristics of each page set.

[0013] It can be seen that combining the features of the page set with the user's rating is conducive to obtaining more accurate real-time preference features. Based on the real-time preference features of the target user, it is conducive to obtaining updated pages that meet the user's preferences.

[0014] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the characteristics of each page set, the method also includes: based on the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the characteristics of each page set, real-time training of the preference model is performed, so that the preference model can obtain the user's real-time preference data based on the user rating set and the characteristics of the page set; obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the characteristics of each page set includes: processing the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the characteristics of each page set through the preference model to obtain the real-time preference characteristics of the target user.

[0015] In some embodiments, before determining each user's preference for a page based on the page interaction data of each user obtained in real time, the method should also include: obtaining each user's page interaction data in real time; performing feature extraction on the page interaction data of each user to obtain each user's interaction features; constructing a user feature database based on the interaction features of each user; determining each user's preference for a page based on the page interaction data of each user obtained in real time includes: determining each user's preference for a page based on the interaction features updated in real time in the user feature database.

[0016] It can be seen that extracting features from user page interaction data can help improve the accuracy of real-time user preference feature analysis. Establishing a user feature database can further expand the use scenarios of user page interaction data.

[0017] The present application also provides a page generation device, the device comprising:

[0018] a processing module configured to determine each user's preference for a page based on page interaction data obtained in real time for each user; obtain a user rating set corresponding to each user based on the user's preference for the page; the user rating set including real-time ratings of multiple pages obtained by the user based on the user's preference for the page; and determine similar users of a target user based on similarities between the user rating sets corresponding to the respective users; the target user being one of the respective users;

[0019] The updating module is used to obtain the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users; and to update the page of the target user in real time based on the real-time preference characteristics.

[0020] An embodiment of the present application provides an electronic device, comprising a processor and a memory for storing a computer program that can be run on the processor; wherein,

[0021] The processor is configured to run the computer program to execute any one of the above page generation methods.

[0022] An embodiment of the present application provides a computer storage medium having a computer program stored thereon, which implements any of the above-mentioned page generation methods when executed by a processor.

[0023] An embodiment of the present application provides a computer program product, including a computer program, which implements any of the above-mentioned page generation methods when executed by a processor.

[0024] The embodiments of the present application provide a page generation method, apparatus, electronic device, medium, and computer program product. By acquiring a set of user ratings for a target user and then combining this with a set of user ratings for similar users similar to the target user, the method helps expand the amount of data used for analyzing the target user's preferences and improves the accuracy of analyzing the target user's real-time preference characteristics. Furthermore, using the obtained real-time preference data of the target user, the target user's page is updated in real time, facilitating the generation of a page that meets the user's needs and preferences, thereby improving the intelligence and accuracy of page generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flow chart of a page generation method provided in an embodiment of the present application;

[0026] Figure 2 A preference model training flow chart provided in an embodiment of the present application;

[0027] Figure 3 A schematic diagram of a modeling method provided in an embodiment of the present application;

[0028] Figure 4 A flowchart of user data collection and analysis provided in an embodiment of the present application;

[0029] Figure 5 This is a diagram of the user preference analysis system architecture provided in an embodiment of the present application;

[0030] Figure 6 A flow chart of a page recommendation method provided in an embodiment of the present application;

[0031] Figure 7 A schematic diagram of the structure of a page generation device provided in an embodiment of the present application;

[0032] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0033] Currently, page layout and content arrangement methods often rely on fixed templates and manual adjustments, failing to dynamically adapt to user preferences and real-time behavior. This static approach to page arrangement struggles to meet individual user needs, resulting in a poor user experience and potentially reducing page views and user engagement. Manual adjustments to page arrangement are labor-intensive and time-consuming, and are unable to quickly respond to changes in real-time user behavior.

[0034] In response to the above-mentioned problems, the embodiments of the present application provide a page generation method, device, electronic device, medium and computer program product. By performing real-time analysis on the user's page interaction data, the user's real-time preference characteristics are obtained, and the user's page layout and content arrangement are dynamically adjusted based on the real-time preference characteristics, thereby effectively improving the user experience.

[0035] The following is a further detailed description of the embodiments of the present application in conjunction with the accompanying drawings and examples. It should be understood that the embodiments provided herein are merely for explaining the embodiments of the present application and are not intended to limit the embodiments of the present application. In addition, the embodiments provided below are partial embodiments for implementing the present application, rather than providing all embodiments for implementing the present application. In the absence of conflict, the technical solutions described in the embodiments of the present application can be implemented in any combination.

[0036] It should be noted that, in the embodiments of the present application, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a method or apparatus comprising a series of elements includes not only the elements explicitly stated, but also other elements not explicitly listed, or also includes elements inherent to the implementation of the method or apparatus. In the absence of further restrictions, an element defined by the sentence "include..." does not exclude the presence of other related elements (such as steps in the method or units in the apparatus, for example, a unit in the apparatus may be a portion of a circuit, a portion of a processor, a portion of a program or software, etc.) in the method or apparatus comprising the element.

[0037] The page generation method provided in the embodiment of the present application includes a series of steps, but the page generation method provided in the embodiment of the present application is not limited to the recorded steps. Similarly, the page generation device provided in the embodiment of the present application includes a series of modules, but the device provided in the embodiment of the present application is not limited to including the modules explicitly recorded, and may also include modules required to obtain relevant information or perform processing based on the information.

[0038] The present application embodiment provides a page generation method, such as Figure 1 As shown, Figure 1 The page generation method shown includes:

[0039] Step 101: Based on the page interaction data of each user obtained in real time, determine each user's preference for the page; based on each user's preference for the page, obtain a user rating set corresponding to each user; the user rating set includes the user's real-time ratings for multiple pages obtained based on the user's preference for the page.

[0040] User page interaction data represents the process of data interaction between a user and the system on a web page or application interface. Specifically, page interaction data can include actions performed by users on a web page or application interface, such as clicking, scrolling, sliding, and adjusting the page window. Page interaction data can also include which web pages or application interfaces a user visited, how long the user stayed on the web page or application interface, and the time at which the user visited the web page or application interface.

[0041] In this embodiment, JavaScript scripts or front-end frameworks (such as React, Vue) can be used to add event listeners to various elements (such as buttons, links, input boxes, etc.) in the web page to record the user's page interaction data in the web page, such as browsing pages, clicking links, dwell time, mouse movement trajectory, etc. Alternatively, AJAX (Asynchronous JavaScript and XML) technology or Fetch Application Programming Interface (Fetch API) technology can be used to send asynchronous requests to the server to collect the user's page interaction data and send it to the server without reloading the web page; or by analyzing the logs, the user's access behavior and page interaction situation can be understood. This embodiment does not specifically limit the method for obtaining the user's page interaction data.

[0042] By analyzing each user's real-time page interaction data, such as time spent on a page and scrolling time, we can understand the user's interest and patience with the content on the page. If a user spends a long time in a certain area of ​​the page, it may indicate that the content in that area is attractive to the user; conversely, the content or layout of that area may need to be optimized. Alternatively, based on the user's page interaction data, a page click heat map can be generated to visually display the distribution of user click behavior. Through heat map analysis, it is possible to determine the page areas that users pay the most attention to and those that they ignore, thereby optimizing the page layout and content.

[0043] The degree of preference of each user for a page obtained here includes the degree of preference of each user for multiple pages. By presetting the scoring criteria, based on the degree of preference of the user for each page, the user's score for each page is obtained. For any user, specifically, the user's page interaction data can be analyzed to obtain the user's preference for the current page. For example, the longer the user stays on the current page, the higher the user's preference for the current page is considered; or, the more times the user clicks on the current page, the higher the user's preference for the current page is considered. After obtaining the user's preference for each page, the user's score for each page is obtained based on the user's preference. The user's score for each page is positively correlated with the user's preference for each page, that is, the higher the user's preference for a page, the higher the user's score for the page. Through the user's score for each page, the user's corresponding user score set is obtained, and the user score set corresponding to each user is obtained.

[0044] In this embodiment, the user rating set corresponding to each user may also include each user's detailed ratings for each component included in each page across multiple pages. Here, the components included in each page include, but are not limited to, navigation bars, content areas, and ad spaces. In a specific implementation, an interactive rating tool may be embedded in the page, allowing the user to directly click and drag a slider or select a star rating to rate different components on the page.

[0045] Step 102: Determine similar users of the target user based on the similarity between the user rating sets corresponding to each user; the target user is one of the users.

[0046] After obtaining the user rating set corresponding to each user, the user rating set corresponding to each user can be preprocessed. For example, for unrated pages (or page components), the unrated pages can be filled with ratings to obtain a page rating matrix consisting of the ratings of each page. The rows of the page rating matrix can be set to represent users, the columns to represent pages, and each element in the matrix represents the user's rating of the page. By calculating the similarity between every two rows in the page rating matrix, the similarity between every two users is determined, and similar users corresponding to the target user are obtained.

[0047] The constructed page score matrix U can be expressed as follows:

[0048]

[0049] The 0 in the page rating matrix U represents data filling for the unrated page. In the page rating matrix U, the target user can be one of the users 1 to 4.

[0050] Here, the similarity between each two rows in the page rating matrix is ​​calculated. Specifically, this can be achieved by calculating the cosine similarity, Pearson correlation coefficient, Jaccard (Jaccard Similarity Coefficient) similarity coefficient, etc. between each two rows in the page rating matrix. For example, when using the Jaccard similarity coefficient to determine the similarity between user 1 and user 2 in the page rating matrix U, the row corresponding to user 1 in the page rating matrix U can be regarded as set 1, and the row corresponding to user 2 in the page rating matrix U can be regarded as set 2. The intersection of set 1 and set 2, and the union of set 1 and set 2 are calculated, and the similarity between user 1 and user 2 is determined by the quotient of the intersection and the union. In the Jaccard similarity coefficient, the higher the quotient of the intersection and the union, the higher the similarity between the users. By determining whether the similarity is greater than the user similarity threshold, similar users to the target user are determined.

[0051] For example, in the page rating matrix U, the Jaccard similarity coefficient can be used to determine that the similarity between user 1 and user 2 is the highest. When user 1 is the target user, the similar user can be determined to be user 2.

[0052] Step 103: Obtain the target user's real-time preference features through the user rating set corresponding to the target user and the user rating set corresponding to similar users; and update the target user's page in real time based on the real-time preference features.

[0053] After determining the target user and similar users, the target user's real-time preference characteristics can be determined using the user rating sets corresponding to the target user and the user rating sets corresponding to similar users. Taking the page rating matrix U as an example, let's assume that the target user is User 1 and the similar user is User 2. It can be seen that although User 2 is a similar user to User 1, User 2 also includes a rating for Page 2. Therefore, User 2's rating for Page 2 can be used to estimate User 1's rating for Page 2 and determine User 1's preference for Page 2.

[0054] By acquiring real-time page interaction data between User 1 and User 2 and obtaining a real-time user rating set for User 1 and User 2, we can not only determine User 1's preference for the pages they have interacted with, but also predict User 1's preference for Page 2 based on the ratings of similar Users 2 on Page 2 (pages for which User 1 has not generated page interaction data, or pages for which User 1 has not rated), thereby obtaining a more comprehensive real-time preference profile for the target user. In practical applications, the target user's page interaction data, similar users' page interaction data, the target user's user rating set, and similar users' user rating set can be combined to jointly determine the target user's real-time preference profile.

[0055] In practical applications, since the user's real-time rating is determined based on the user's preference level, the target user's real-time preference level can also be re-determined based on the user rating sets corresponding to the target user and similar users. Based on the re-determined target user's real-time preference level, the target user's real-time preference characteristics are obtained, and the target user's page is updated in real time.

[0056] For example, a recommendation algorithm can be used to recommend content that the target user may be interested in, or to adjust the page layout of the target user, based on the target user's real-time preference characteristics and / or the target user's page interaction data. In this embodiment, the target user's page is updated in real time, which can specifically include real-time updates to the target user's page layout and / or content arrangement. In actual applications, after the server obtains the user's page interaction data, the back-end technologies such as Node.js and Python can be used to process and analyze the user behavior data in real time. Based on the analysis results, the updated page content is immediately pushed to the client through WebSocket or other push technologies, and the real-time update strategy is integrated with the content management system to realize automatic update and management of page content. The template and component functions of the CMS are used to quickly build and update the page layout and content.

[0057] After the target user's page is updated in real time, the target user's page interaction data is continuously acquired in real time, and based on the method provided in the embodiment of the present application, the target user's page is continuously updated so that the updated page can meet the needs and preferences of the target user.

[0058] In practical applications, steps 101 to 103 can be implemented based on a processor, and the processor can be at least one of an application-specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, and a microprocessor.

[0059] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the method further includes: based on the user rating set corresponding to the target user and the user rating set corresponding to similar users, performing real-time training on the preference model, so that the preference model can obtain the user's real-time preference data based on the user rating set; obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users includes: processing the user rating set corresponding to the target user and the user rating set corresponding to similar users through the preference model to obtain the real-time preference characteristics of the target user.

[0060] In this embodiment, the preference model can be trained in real time using a set of user ratings corresponding to the target user and a set of user ratings corresponding to similar users. This allows the preference model to process the real-time user rating set to obtain similar users of the target user and the target user's real-time preference profile. The preference model can be located on a backend server and subscribe to user page interaction data, obtaining the user's page interaction data in real time. Based on each user's page interaction data, the similarity between users is determined, similar users to the target user are identified, and the target user's real-time preference profile is obtained using the user rating sets corresponding to the target user and the similar users.

[0061] Figure 2 A preference model training flow chart is shown, such as Figure 2 As shown, the training of the preference model includes:

[0062] Step 201: Preference model subscription data update.

[0063] JavaScript can be used to record user page interaction data and report each user's page interaction data to the server through AJAX or FetchAPI for storage and further processing. The preference model subscribes to the data obtained by the server. When the server obtains the user's page interaction data, the preference model processes the user interaction data to obtain a user rating set for each user, and uses this user rating set for model training.

[0064] Alternatively, the server may process the received user page interaction data to obtain a user rating set for each user, and subscribe to the user rating set for each user obtained by the server through a preference model, so that the preference model is trained based on the user rating set for each user.

[0065] Step 202: Train the preference model.

[0066] After the preference model obtains each user's rating set, it is trained using this set. The preference model can be constructed based on a self-attention network, enabling model training based on the user rating set. During model training, the preference model's self-attention layer captures the dependencies between elements at different positions in the user rating set. This dependency determines the user's preferences for different pages, generating real-time preference features.

[0067] In practical applications, machine learning frameworks such as TensorFlow or PyTorch can be used to obtain user rating sets and integrate them with deep learning algorithms. This approach leverages the strengths of both to address sparsity and insufficient feature representation, improving the accuracy and diversity of training data. Modeling can then be performed using the acquired user page interaction data to calculate user preference features and generate a user preference model.

[0068] It can be seen that by obtaining the user's page interaction data in real time, we can obtain the user's real-time user rating set for each page. Through the user's real-time user rating set, the model parameters of the preference model can be updated in real time, and the model parameters of the preference model can be continuously iterated, so that the preference model can learn the user's page interaction data and obtain the user's real-time preference characteristics.

[0069] During the training of the preference model, the currently acquired user rating set is processed synchronously through the preference model to obtain the real-time preference features of the target user, and the target user's page is updated in real time through the real-time preference features.

[0070] It can be seen that the method provided in this embodiment can realize real-time updating of the model parameters of the preference model, and obtain the user's real-time preference characteristics from the real-time updated preference model, which is conducive to obtaining accurate real-time preference characteristics and obtaining a page that meets the user's preferences and needs.

[0071] In some embodiments, before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the method further includes: obtaining a page rating set corresponding to each page based on each user's preference for the page; the page rating set includes real-time ratings of multiple users corresponding to the page obtained based on the user's preference for the page; determining the similarity between each page through the similarity between the page rating sets corresponding to each page; the real-time update of the target user's page based on the real-time preference features includes: real-time updating of the target user's page based on the real-time preference features and the similarity between each page.

[0072] Based on the method provided in the above embodiment, after obtaining the ratings of different users on different pages, a page rating set for each page can be determined based on the page. Taking the page rating matrix U shown in step 102 as an example, the page rating set for each page corresponds to the columns of the page rating matrix U. For page 1, the page rating set corresponding to page 1 is (3, 5, 0, 0), and the page rating set corresponding to page 2 is (0, 4, 0, 2), where each element in the page rating set represents the rating of the page by different users.

[0073] After obtaining the page score set corresponding to each page, the similarity between each two pages can be obtained by calculating the cosine similarity, Pearson correlation coefficient, Jaccard similarity coefficient, etc. between each two page score sets. Taking the Jaccard similarity coefficient as an example, the Jaccard similarity coefficient shows that the similarity coefficient between page 1 and page 2 is 1 / 3, the similarity coefficient between page 1 and page 3 is 0, the similarity coefficient between page 1 and page 4 is 1, and the similarity coefficient between page 1 and page 5 is 1 / 3. Therefore, page 1 and page 4 are considered to be the most similar, and the similarity between page 1 and page 2 and page 3 is next. It can be considered that page 1 and page 4 are similar.

[0074] After obtaining similar pages of a page through the similarity between pages, the target user's page can be updated in real time based on the page similarity, through pages with similarity greater than the page similarity threshold and the real-time preference characteristics of the target user.

[0075] Specifically, for pages with high user ratings, similar pages can be recommended. Pages with low user ratings can be analyzed, and through the similarity between pages, common issues between low-rated pages, such as duplicate content and poor layout, can be found. These low-rated pages can then be optimized to improve user satisfaction. For example, when User 2 is the target user, it can be seen that User 2 has a high rating for Page 1, meaning User 2 has a higher preference for Page 1. At the same time, based on the method described in the above embodiment, it can be determined that User 2's similar user is User 1, and User 1 also has a relatively high rating for Page 1. In this case, through the similarity between pages, it can be determined that Page 4 is a similar page to Page 1. Page 4 can be recommended to User 2, or, based on the common features of Page 1 and Page 4, User 2's page can be optimized and updated.

[0076] It can be seen that in the method provided in this embodiment, by combining the similarity between pages with user ratings, the target user's page is updated in real time, which can further improve the accuracy of page updates and make the updated page more in line with the preferences and needs of the target user.

[0077] In some embodiments, before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the method further includes: determining one or more page sets based on the similarity between each two pages; each page set of the one or more page sets includes two or more pages, and the similarity between each page in the page set is less than a page similarity threshold; obtaining the features of each page set based on each page in each page set; obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users includes: obtaining the real-time preference features of the target user through the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the features of each page set.

[0078] In combination with the method provided in the above embodiment, this embodiment further provides a method for determining the real-time preference characteristics of the target user.

[0079] Based on the method provided in the above embodiment, the similarity between each two pages can be determined by each user's rating of each page. Alternatively, the similarity between each two pages can be determined by one or more of the similarity of the text in each page, the similarity of the page structure, the similarity of the image content contained in the page, the domain name of the page, and the Uniform Resource Locator (ULR) information. Based on the similarity between each two pages, pages with a similarity greater than a page similarity threshold are clustered into a page set, thereby obtaining one or more page sets.

[0080] In each page set, keywords are extracted based on the text content of each page in each page set, and semantic analysis is performed on the keywords of each page to determine the common features and differences between the keywords. Based on the common features extracted between the keywords, the characteristics of each page set are obtained. Alternatively, external links of each page in each page set can be extracted, and the link relationships between the pages can be analyzed. The characteristics of each page in the page set can be determined based on the link relationships as the characteristics of the page set. Technical means such as text analysis, image processing, and link analysis can be used to extract feature vectors of pages, and clustering algorithms (such as K-means, DBSCAN, etc.) can be used to find common features of similar pages.

[0081] After obtaining the characteristics of each page set, that is, among the multiple pages visited by the user, the type of page visited by the user is obtained. By combining the user rating set corresponding to the target user and the user ratings corresponding to similar users, pages with higher ratings can be obtained. Furthermore, based on the pages with higher ratings of the target user and similar users, a page set containing pages with higher ratings of the target user can be obtained, and the characteristics of the page set can be determined as the real-time preference characteristics of the target user. Alternatively, the weight of each page can be determined by combining the user rating sets corresponding to the target user and similar users, and a weighted sum is performed based on the weight of each page and the characteristics of the page set where each page is located to determine the real-time preference characteristics of the target user. Based on the real-time preference characteristics of the target user, the page is updated in real time.

[0082] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set, the method further includes: based on the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set, real-time training of the preference model is performed, so that the preference model can obtain the user's real-time preference data based on the user rating set and the characteristics of the page set; obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set includes: processing the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set through the preference model to obtain the real-time preference characteristics of the target user.

[0083] Based on the method provided in the above embodiment, after obtaining the user rating set and the features of each page set, a preference model can be jointly established based on the user rating set of the target user, the user rating set of similar users, and the features of each page set, and the preference model can be trained. The rating data in the user rating set and the features of the page set can be combined to construct a page dataset, which can include user identifiers, page identifiers, rating scores, and page feature vectors. The preference model can specifically be a deep learning model constructed based on a self-attention network.

[0084] The preference model is trained using the page dataset, and the performance of the model is optimized by adjusting the model parameters of the preference model. During the training process, the cross-validation method can also be used to further evaluate the generalization ability of the model.

[0085] In the process of acquiring page interaction data, the preference model after preliminary training is deployed in an online environment, and the user's real-time preference characteristics are predicted based on the user's page interaction data through the preference model.

[0086] The user rating set and page rating set corresponding to the user given in the above embodiment can be specifically processed and implemented based on the collaborative filtering layer of the preference model. Based on the method given in the above embodiment, Figure 3 A schematic diagram of a modeling method is shown, including:

[0087] Step 301: Calculate similarity through the collaborative filtering layer.

[0088] In this step, the user's preference for the page can be preliminarily calculated based on the user's page interaction data, and the preference of each user for the page can be obtained.

[0089] First, we need to construct a user-page interaction matrix, using a two-dimensional matrix to represent user ratings of pages. We obtain the page rating matrix U shown in step 102. Here, we define the user set Us = [u1, u2, …, um] and the page set Ui = [i1, i2, …, in].

[0090] Construct a page rating matrix U, where each element R(u, i) in the page rating matrix U represents the rating of user (u) on page (i) (explicit rating), or whether user (u) interacts with page (i) (implicit rating, such as click, browse).

[0091] The page rating matrix U includes the user's ratings R(u, i) for each page, with unrated pages represented by 0. The page rating matrix U can also be broken down into the set N(u) of pages rated by each user and the set N(i) of users for whom each page was rated. The set N(u) of pages rated by each user is the row of the page rating matrix U, and the set N(i) of users for whom each page was rated is the column of the page rating matrix U.

[0092] By calculating the similarity between each set N(u) of user-rated pages, similar users of the target user are obtained. By calculating the similarity between each set N(i) of user-rated pages, the similarity between pages is obtained. Since the Jaccard similarity coefficient is generally used to measure the ratio of the intersection and union of users or items (applicable to implicit rating scenarios), the Jaccard similarity coefficient can be preferred for determining similarity in the page recommendation update scenario involved in this embodiment.

[0093] After obtaining the similarity between users and the similarity between pages, a preliminary recommendation list can be formed based on the similarity between users and the similarity between pages. That is, through the collaborative filtering layer of the preference model, the preliminary real-time preference data corresponding to the target user is obtained.

[0094] Step 302: Perform data processing through the data extension layer.

[0095] After obtaining the user rating set and the page rating set, the similarity between users and the similarity between pages can be obtained. A data expansion layer can be constructed to process the data obtained in step 301, perform feature fusion on the ratings of the target user and similar users, and perform feature fusion on the features between similar pages to generate data in multiple dimensions. The data in multiple dimensions is used as input data for the deep learning layer of the preference model.

[0096] Step 303: Construct a deep learning layer to determine real-time preference features.

[0097] The deep learning layer of the preference model is primarily used to capture complex nonlinear relationships in the data. By learning the complex nonlinear relationships in the input data, the deep learning layer has the ability to extract complex feature combinations and has strong generalization capabilities.

[0098] Specifically, the input layer of the deep learning layer receives input data and embeds it into a low-level dense vector representation. The self-attention layer of the deep learning layer captures dependencies within the input data. By optimizing the parameters of the deep learning layer, the loss function is reduced to a preset value.

[0099] As can be seen, by incorporating the concept of collaborative filtering, similar users and pages to the target user are obtained, thus expanding the input data for the target user. Processing this expanded data through the preference model helps improve the accuracy of the target user's real-time preference characteristics obtained through the preference model.

[0100] In some embodiments, before determining each user's preference for a page based on the page interaction data of each user obtained in real time, the method should also include: obtaining each user's page interaction data in real time; performing feature extraction on each user's page interaction data to obtain each user's interaction features; and constructing a user feature database based on each user's interaction features. The method of determining each user's preference for a page based on the page interaction data of each user obtained in real time includes: determining each user's preference for a page based on the interaction features updated in real time in the user feature database.

[0101] Based on the method given in the above embodiment, JavaScript scripts or front-end frameworks (such as React, Vue) can be used to record the user's behavior data on the web page to obtain real-time page interaction data for each user. User behavior log data can be generated through real-time page interaction data, and the user behavior log data can be reported to the log database. By processing the page interaction data in the log database, the features of the page interaction data are extracted to obtain the interaction features of each user. Based on the interaction features of each user, a user feature database is obtained. The user feature database includes the features of the page interaction data of each user on each page.

[0102] In practical applications, after acquiring user page interaction data, intelligent recommendation algorithms such as collaborative filtering, content filtering, or hybrid recommendation systems can be used to clean and extract features from the data. Finally, the data is vectorized and stored in a user feature database for fast retrieval and real-time updates. The user feature database can be a NoSQL (Not Only Structured Query Language) database, such as MongoDB.

[0103] The preference model can be used to subscribe to the update message of the user feature database. After receiving the update message of the user feature database, the preference model is based on the interactive features updated in real time in the user feature database. Figure 3 The method shown determines the real-time preference characteristics of each user for pages.

[0104] Based on the data processing method given in this embodiment, Figure 4 A user data collection and analysis flow chart is shown, including:

[0105] Step 401: Obtain user page interaction data.

[0106] Get each user's page interaction data in real time.

[0107] Step 402: Establish a log database based on user page interaction data.

[0108] Based on each user's page interaction data, user behavior log data corresponding to each user is obtained; based on the user behavior log data corresponding to each user, a log database is established. The subsequent user's page interaction data is stored in the log database in the form of a log. Alternatively, the page log containing the user's page interaction data can be directly read and stored in the log database.

[0109] Step 401: Extract features from page interaction data and build a user feature database.

[0110] The feature extraction is performed on the page interaction data of each user in the log database. The user feature database is constructed by extracting the interaction features of each user. The user feature database includes the interaction features corresponding to the page interaction data of each user on each page.

[0111] It can be seen that the user page interaction data obtained in real time is uniformly stored in the log database and the user feature database, so that the log database and the user feature database can be used for subsequent user behavior analysis or expansion of other businesses.

[0112] In some embodiments, a K-means clustering algorithm can also be used to cluster the page interaction data of each user. First, K cluster centers are initialized, and the page interaction data of each user is assigned to the cluster center with the closest distance (Euclidean distance); the cluster center is updated to the mean of the page interaction data in each cluster, and the above steps are repeated until the cluster center no longer changes or changes very little.

[0113] In the specific implementation process, it is assumed that the user's page interaction data can be represented as a vector X, X = [x1, x2, x3, ..., x n], where x i Represents a type of user page interaction data, for example: x1 represents the average dwell time, x2 represents the page click rate, x3 represents the proportion of visited content types, and x4 represents the interaction frequency.

[0114] If a user's page interaction data consists of the above four dimensions, the corresponding page interaction data can be expressed as: X = [x1, x2, x3, x4]. After K-means clustering, the output user interaction features are converted into high-dimensional data vectors, and the cluster center and corresponding user label are obtained to assist in subsequent model training and personalized recommendations.

[0115] During the K-means clustering process, you can also evaluate the clustering effect based on the silhouette coefficient to determine whether the K value needs to be adjusted. If an evaluation metric (such as the silhouette coefficient) does not meet the preset conditions, increase the K value, re-cluster, and evaluate the clustering effect until the optimal K value is found.

[0116] The embodiment of the present application provides a page generation method. Through the page generation method provided in the embodiment of the present application, the user's web page layout and content arrangement can be dynamically adjusted to provide a personalized user experience; by dynamically adjusting the user's page, it is beneficial to increase page visits and user stickiness, and enhance user satisfaction; by dynamically adjusting the user's page, the workload of manual page adjustment is effectively reduced, and page operation efficiency is improved; through the page generation method provided in the embodiment of the present application, the update effect of the user's page is effectively improved, so that the updated page can be more in line with the user's preferences, meet the user's current usage needs, respond to changes in user behavior and preferences in real time, and quickly adapt to market demand.

[0117] Based on the page generation method given in the above embodiment, Figure 5 A user preference analysis system architecture diagram is shown in FIG. Figure 5 As shown, the user preference analysis system includes a data collection module 501, a data processing module 502, a preference model module 503, an application service module 504, and a database 505. Among them, the data collection module 501 is responsible for collecting the page interaction data of each user delivered by the JavaScript script in the controller web page. The data collection module 501 includes a data collection service and a user log service. The data collection service is used to collect and count the page interaction data of each user delivered by the JavaScript script. The user log service is used to generate user behavior log data based on the page interaction data of each user delivered by the JavaScript script. The user behavior log data is stored in the log database in the database 505.

[0118] Data processing module 502 is responsible for processing the collected page interaction data and converting it into a vector representation. Data processing module 502 includes a data processing service, a feature extraction service, and a vectorization module. The data processing service is responsible for preprocessing the original collected user page interaction data, for example, removing duplicate data and stop words. The feature extraction service and vectorization module are responsible for extracting features from the preprocessed page interaction data to obtain interaction features corresponding to the page interaction data. The interaction features processed by data processing module 502 are stored in the user feature database in database 505.

[0119] The preference model module 503 is used to generate a preference model using the machine learning framework TensorFlow based on the vector data stored in the user feature database, and to update and maintain the preference model parameters. The application service module 504 is used to obtain the content and components of the page from the content database in database 505 based on the target user's real-time preference features generated by the preference model, and to provide personalized recommendation, search, and other services to the target user. The API gateway in the user preference analysis system provides an interface for external systems and services to access the preference model.

[0120] based on Figure 5 The user preference analysis system shown, and the page generation method given in the above embodiment, Figure 6 A flow chart of a page recommendation method is shown, including:

[0121] Step 601: The controller obtains a recommendation page request.

[0122] Controller web pages are requested through scripts or front-end frameworks (such as React, Vue.js) Figure 5 The user preference analysis system shown obtains page layout recommendations.

[0123] Step 602: The API gateway forwards the request.

[0124] The API gateway (such as APISIX, Kong, etc.) of the user preference analysis system forwards the received controller web page request to the backend service in the user preference analysis system according to the routing rules, such as the content recommendation service in the application service module 504.

[0125] Step 603: The content recommendation service accesses the preference model and the content database.

[0126] The content recommendation service accesses the preference model through the model access service, obtains the real-time preference characteristics of the target user, and obtains page recommendation data from the content database in the database 505.

[0127] Step 604: The controller updates the page.

[0128] Leverage real-time preference features recommended by user preference models to dynamically adjust webpage content and layout. Page elements can be displayed and hidden using Cascading Style Sheets (CSS) and JavaScript to optimize the user experience. Use template engines (such as Handlebars.js and Jinja2) to generate personalized webpage content.

[0129] Step 605: Continuous monitoring and updating.

[0130] Monitor the user's latest page interaction data in real time (reported by the controller) and continuously update the user's preference model to ensure that the user's preference model and page update plan are constantly improved.

[0131] The present application provides a method for arranging page preferences based on an intelligent recommendation algorithm. By collecting user page interaction data in real time and utilizing a fusion recommendation method of collaborative filtering and deep learning, the performance of the page recommendation system is greatly improved by combining traditional sparse matrix modeling capabilities with the nonlinear modeling capabilities of deep learning. The method dynamically adjusts the layout and content arrangement of web pages based on the preference model to provide a personalized user experience. This method has the advantages of dynamically adjusting arrangement, improving user stickiness, reducing the workload of manual adjustments, and responding to user needs in real time.

[0132] Those skilled in the art will understand that, in the above-mentioned specific implementation method, the writing order of each step does not mean a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of each step should be determined by its function and possible internal logic.

[0133] Based on the page generation method proposed in the above embodiment, the present embodiment further proposes a page generation device. Figure 7 As shown, Figure 7 A schematic diagram of the structure of a page generation device is shown, the device comprising:

[0134] Processing module 701 is used to determine each user's preference for a page based on the page interaction data of each user obtained in real time; obtain a user rating set corresponding to each user based on each user's preference for the page; the user rating set includes the user's real-time ratings for multiple pages obtained based on the user's preference for the page; determine similar users of the target user through the similarity between the user rating sets corresponding to each user; the target user is one of the users.

[0135] The updating module 702 is configured to obtain the target user's real-time preference characteristics through the user rating set corresponding to the target user and the user rating sets corresponding to similar users; and update the target user's page in real time based on the real-time preference characteristics.

[0136] In practical applications, the processing module 701 and the updating module 702 can be implemented based on a processor and a communication device.

[0137] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the processing module 701 is also used to perform real-time training on the preference model based on the user rating set corresponding to the target user and the user rating set corresponding to similar users, so that the preference model can obtain the user's real-time preference data based on the user rating set; the updating module 702 is specifically used to process the user rating set corresponding to the target user and the user rating set corresponding to similar users through the preference model to obtain the real-time preference characteristics of the target user.

[0138] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the processing module 701 is also used to obtain the page rating set corresponding to each page based on the preference degree of each user for the page; the page rating set includes the real-time ratings of multiple users corresponding to the page obtained based on the user's preference degree for the page; the similarity between each page is determined by the similarity between the page rating sets corresponding to each page; the update module 702 is specifically used to update the target user's page in real time based on the real-time preference characteristics and the similarity between each page.

[0139] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to similar users, the processing module 701 is also used to determine one or more page sets based on the similarity between each two pages; each page set of the one or more page sets includes two or more pages, and the similarity between each page in the page set is less than the page similarity threshold; based on each page in each page set, the characteristics of each page set are obtained; the update module 702 is specifically used to obtain the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set.

[0140] In some embodiments, before obtaining the real-time preference characteristics of the target user through the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set, the processing module 701 is also used to perform real-time training on the preference model based on the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set, so that the preference model can obtain the user's real-time preference data based on the user rating set and the characteristics of the page set; the update module 702 is specifically used to process the user rating set corresponding to the target user, the user rating set corresponding to similar users, and the characteristics of each page set through the preference model to obtain the real-time preference characteristics of the target user.

[0141] In some embodiments, before determining each user's preference for a page based on the page interaction data of each user obtained in real time, the processing module 701 is also used to obtain each user's page interaction data in real time; perform feature extraction on each user's page interaction data to obtain each user's interaction features; and construct a user feature database based on each user's interaction features. The processing module 701 is specifically used to determine each user's preference for a page based on the interaction features updated in real time in the user feature database.

[0142] It should be noted that the description of the above device embodiment is similar to the description of the above method embodiment, and has similar beneficial effects as the same method embodiment. For technical details not disclosed in the device embodiment of this application, please refer to the description of the method embodiment of this application for understanding.

[0143] It should be noted that, in the embodiment of the present application, if the above method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a terminal, server, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a magnetic disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific combination of hardware and software.

[0144] An embodiment of the present application also provides an electronic device. Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown in FIG. Figure 8 As shown, the electronic device 80 may include:

[0145] The memory 801 is used to store executable instructions.

[0146] The processor 802 is configured to implement any of the above page generation methods when executing the executable instructions stored in the memory 801.

[0147] The processor 802 may be at least one of an ASIC, a DSP, a DSPD, a PLD, an FPGA, a CPU, a controller, a microcontroller, and a microprocessor.

[0148] In actual application, the computer-readable storage medium or memory 801 can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); or various terminals including one or any combination of the above memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc. Accordingly, an embodiment of the present application further provides a computer storage medium, on which computer executable instructions are stored, and the computer executable instructions are used to implement any of the page generation methods provided in the above embodiments.

[0149] Correspondingly, an embodiment of the present application further provides a computer program product, which includes computer-executable instructions, and the computer-executable instructions are used to implement any one of the page generation methods provided in the above embodiments.

[0150] In some embodiments, the functions or modules included in the device provided in the embodiments of the present application can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.

[0151] The above description of the various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced with each other and will not be repeated herein for the sake of brevity.

[0152] The methods disclosed in the various method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0153] The features disclosed in the various product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0154] The features disclosed in the various method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0155] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0156] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are protected by this application.

Claims

1. A page generation method, characterized in that: The method comprises: Determining each user's preference for a page based on real-time page interaction data of each user; obtaining a user rating set corresponding to each user based on the user's preference for the page; the user rating set including real-time user ratings for multiple pages obtained based on the user's preference for the page; Determine similar users of a target user based on similarities between user rating sets corresponding to each user; the target user is one of the users; The real-time preference features of the target user are obtained through the user rating set corresponding to the target user and the user rating set corresponding to the similar users; and the page of the target user is updated in real time based on the real-time preference features.

2. The method according to claim 1, characterized in that Before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method further includes: Based on the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the preference model is trained in real time, so that the preference model can obtain the user's real-time preference data based on the user rating set; The step of obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users includes: The user rating set corresponding to the target user and the user rating set corresponding to the similar users are processed by the preference model to obtain the real-time preference features of the target user.

3. The method according to claim 1, characterized in that Before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method further includes: Based on the degree of preference of each user for the page, a page score set corresponding to each page is obtained; the page score set includes real-time scores of multiple users corresponding to the page obtained based on the degree of preference of the user for the page; Determine the similarity between each page by comparing the similarity between the page score sets corresponding to each page; The updating of the target user's page in real time based on the real-time preference feature includes: Based on the real-time preference features and the similarities between the various pages, the target user's page is updated in real time.

4. The method according to claim 1, wherein Before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users, the method further includes: Determining one or more page sets based on the similarity between each two pages; each of the one or more page sets includes two or more pages, and the similarity between each page in the page set is less than a page similarity threshold; Based on each page in each page set, obtaining a feature of each page set; The step of obtaining the real-time preference features of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users includes: The real-time preference features of the target user are obtained through the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the features of each page set.

5. The method according to claim 4, characterized in that Before obtaining the real-time preference features of the target user through the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the features of each page set, the method further includes: Based on the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the features of each page set, the preference model is trained in real time, so that the preference model can obtain the user's real-time preference data based on the user rating set and the features of the page set; The step of obtaining the real-time preference features of the target user by using the user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the features of each page set includes: The user rating set corresponding to the target user, the user rating set corresponding to the similar users, and the features of each page set are processed by the preference model to obtain the real-time preference features of the target user.

6. The method according to claim 1, characterized in that Before determining each user's preference for the page based on the page interaction data of each user obtained in real time, the method may further include: Get each user's page interaction data in real time; Extracting features from the page interaction data of each user to obtain interaction features of each user; Building a user feature database based on the interaction features of each user; Determining each user's preference for a page based on the page interaction data of each user obtained in real time includes: Based on the interactive features updated in real time in the user feature database, each user's preference for the page is determined.

7. A page generating device, characterized in that: The device comprises: a processing module configured to determine each user's preference for a page based on page interaction data obtained in real time for each user; obtain a user rating set corresponding to each user based on the user's preference for the page; the user rating set including real-time ratings of multiple pages obtained by the user based on the user's preference for the page; and determine similar users of a target user based on similarities between the user rating sets corresponding to the respective users; the target user being one of the respective users; The updating module is used to obtain the real-time preference characteristics of the target user through the user rating set corresponding to the target user and the user rating set corresponding to the similar users; and to update the page of the target user in real time based on the real-time preference characteristics.

8. An electronic device, characterized in that: The electronic device comprises a processor and a memory for storing a computer program that can be run on the processor; wherein, The processor is configured to run the computer program to perform the method according to any one of claims 1 to 6.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images