A method and system for intelligent updating of interactive content on a cloud platform

Abstract

This application relates to the field of interactive content update technology, and discloses a method and system for intelligent updating of interactive content on a cloud platform. The method includes: analyzing and calculating heatmap values ​​of user operation behavior based on pre-acquired user operation data to construct a cloud platform click heatmap; analyzing user operation intentions from the cloud platform click heatmap using a preset intent analysis model to construct intent feature vectors; classifying user intentions according to the intent feature vectors using a preset content optimization model, and performing layered optimization of the cloud platform's interactive content based on the classified user intentions to obtain first interactive content; updating the cloud platform click heatmap within a preset time period, analyzing user interaction with the first interactive content in conjunction with the updated cloud platform click heatmap, and updating the first interactive content accordingly. This application can update and optimize cloud platform interactive content in a timely manner, improving the platform's interactive effect.

Description

Technical Field

[0001] This invention relates to the field of interactive content update technology, and in particular to a method and system for intelligent updating of interactive content on a cloud platform. Background Technology

[0002] Currently, cloud platforms serve as the core carriers of information exchange and service provision, and the quality of interactive content on live streaming platforms directly impacts user experience and platform competitiveness. As users' demands for personalized and intelligent services continue to rise, it is necessary to dynamically update and optimize the interactive content of live streaming platforms based on user behavior and needs. Existing methods for updating platform interactive content rely on manual experience or simple rule engines, failing to utilize user operation data deeply and comprehensively enough, resulting in poor effectiveness in updating interactive content.

[0003] Existing technologies have the following problems: the analysis of user operation data is simple and lacks analysis of user operation behavior based on different areas of the cloud platform operation interface; the analysis of user operation intent is based solely on user operation data, resulting in inaccurate user intent; when optimizing interactive content, a single optimization method is used, which cannot provide differentiated content optimization solutions for different levels of user intent, leading to poor optimization results and difficulty in improving user interaction experience and participation; to solve at least one of the above problems, this invention proposes a method and system for intelligent updating of interactive content on a cloud platform. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the main objective of this invention is to provide a method and system for intelligent updating of interactive content on a cloud platform, effectively solving the problems described in the background section. The specific technical solution of this invention is as follows:

[0005] A method for intelligent updating of interactive content on a cloud platform includes:

[0006] Based on the pre-acquired user operation data, analyze and calculate the heat value of user operation behavior, and construct a click heat map of the cloud platform.

[0007] By using a pre-defined intent analysis model, the user's operational intent is analyzed from the click heatmap of the cloud platform, and an intent feature vector is constructed.

[0008] Based on the intent feature vector, user intent is classified using a preset content optimization model, and the interactive content of the cloud platform is layered and optimized based on the classified user intent to obtain the first interactive content.

[0009] Within a preset time period, the cloud platform click heatmap is updated, and the user's interaction with the first interactive content is analyzed in combination with the updated cloud platform click heatmap. The first interactive content is then updated accordingly, so as to intelligently update the cloud platform interactive content.

[0010] Specifically, the step of analyzing and calculating heatmap values ​​of user actions based on pre-acquired user operation data to construct a cloud platform click heatmap includes:

[0011] Based on the pre-acquired user operation data, analyze the user's click behavior on interactive content on the cloud platform on different devices to obtain click analysis results;

[0012] On the cloud platform operation interface, the heat values ​​of different locations on the interface are calculated based on the click analysis results to construct a cloud platform click heat map.

[0013] Specifically, the step of calculating heat values ​​at different locations on the cloud platform operation interface based on the click analysis results and constructing a cloud platform click heatmap includes:

[0014] The cloud platform operation interface is divided into multiple grid areas according to the preset grid size.

[0015] For each grid area, the click analysis results are analyzed using a preset thermal analysis model to calculate the thermal value;

[0016] Construct a thermal value matrix based on the thermal values ​​of each grid region;

[0017] Based on the aforementioned heat value matrix, a click heatmap for the cloud platform is constructed.

[0018] Specifically, the step of analyzing user operation intent from the click heatmap of the cloud platform using a preset intent analysis model and constructing an intent feature vector includes:

[0019] Based on the cloud platform click heatmap, the cloud platform operation interface is divided into gradients and the corresponding gradient values ​​are calculated. The first gradient region with the gradient value greater than the preset gradient threshold is then selected.

[0020] By using a pre-defined intent analysis model, the user's operational intent within the first gradient region is analyzed, and an intent feature vector is constructed.

[0021] Specifically, the step of analyzing the user's operational intent within the first gradient region using a preset intent analysis model and constructing an intent feature vector includes:

[0022] Within the first gradient region, the user's operational intent is analyzed using a pre-defined intent analysis model to obtain the first intent feature vector;

[0023] According to the preset feature mapping rules, the first intent feature vector is mapped to the cloud platform operation interface to obtain the second intent feature vector;

[0024] Combine the first intent feature vector and the second intent feature vector to construct an intent feature vector.

[0025] Specifically, based on the intent feature vector, user intents are classified using a preset content optimization model, and the interactive content on the cloud platform is then layered and optimized based on the classified user intents to obtain the first interactive content, including:

[0026] Based on the intent feature vector, the intensity of user intent is analyzed through a preset content optimization model, and user intent is classified into levels to obtain multi-level user intent;

[0027] Based on the multi-level user intents, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content, wherein the multi-level user intents include high-level user intents, intermediate-level user intents, and low-level user intents.

[0028] Specifically, based on the multi-level user intent, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content, including:

[0029] For advanced user intent, generate the first optimized instruction by creating a real-time interactive scenario;

[0030] For intermediate user intent, a second optimized instruction is generated by dynamically reorganizing and enhancing the interactive content.

[0031] For basic user intent, a third optimization instruction is generated by adjusting the dynamic effects of the interactive content;

[0032] By combining the first optimization instruction, the second optimization instruction, and the third optimization instruction, the interactive content of the cloud platform is optimized in layers to obtain the first interactive content.

[0033] Specifically, the process of updating the cloud platform click heatmap within a preset time period, analyzing user interaction with the first interactive content based on the updated cloud platform click heatmap, and then updating the first interactive content to intelligently update the cloud platform interactive content includes:

[0034] Within a preset time period, the cloud platform click heatmap is updated based on real-time user operation data to obtain an updated cloud platform click heatmap.

[0035] Based on the updated cloud platform click heatmap, analyze user interaction with the first interactive content and construct an interaction effect vector;

[0036] The first interactive content is updated using the interactive effect vector to intelligently update the interactive content on the cloud platform.

[0037] Specifically, the first interactive content is updated using the aforementioned interactive effect vector to intelligently update the interactive content on the cloud platform, including:

[0038] Based on the interaction effect vector, the first interaction content is decomposed to obtain the first content, the second content, and the third content.

[0039] Using a pre-defined content update model, the corresponding content in the first, second, and third content is updated respectively to construct an updated content sequence;

[0040] Based on the updated content sequence, the corresponding interface elements in the first interactive content are replaced and updated to intelligently update the interactive content on the cloud platform.

[0041] A cloud platform interactive content intelligent update system, used to implement the aforementioned cloud platform interactive content intelligent update method, includes:

[0042] The cloud platform click heatmap construction module analyzes and calculates heat values ​​based on pre-acquired user operation data to construct a cloud platform click heatmap.

[0043] The intent feature vector construction module analyzes the user's operation intent from the click heatmap of the cloud platform using a preset intent analysis model, and constructs an intent feature vector.

[0044] The interactive content optimization module classifies user intents according to the intent feature vectors using a preset content optimization model, and then performs layered optimization of the interactive content on the cloud platform based on the classified user intents to obtain the first interactive content.

[0045] The interactive content update module updates the cloud platform click heatmap within a preset time period, analyzes user interaction with the first interactive content based on the updated cloud platform click heatmap, and updates the first interactive content accordingly, thereby intelligently updating the cloud platform interactive content.

[0046] Compared with the prior art, this application has the following beneficial effects:

[0047] This application constructs a cloud platform click heatmap based on user operation data and the cloud platform's user interface. It analyzes user operation intentions in different areas based on this heatmap and employs different interactive content optimization strategies for different levels of user operation intentions. The heatmap is updated in conjunction with real-time user operation data, and the interactive content is optimized accordingly. By constructing the cloud platform click heatmap, the application can reflect the hotspots and distribution patterns of user operations on the interface, accurately grasp the user's true intentions, and provide a clear direction for content optimization. Layered optimization can meet the personalized needs of different users, improve user interaction experience and engagement, and allow for timely updates and optimization of interactive content based on user feedback, achieving intelligent updates of interactive content and improving the effectiveness of interactive content on the cloud platform. Attached Figure Description

[0048] Figure 1 This is a flowchart illustrating the process of a cloud platform interactive content intelligent update method according to Embodiment 1 of the present invention.

[0049] Figure 2 This is a schematic diagram of the cloud platform click heatmap in Embodiment 1 of the present invention;

[0050] Figure 3 This is a flowchart illustrating the workflow of layered optimization of interactive content on the cloud platform in Embodiment 1 of the present invention.

[0051] Figure 4 This is a schematic diagram of the structure of a cloud platform interactive content intelligent update system according to Embodiment 2 of the present invention. Detailed Implementation

[0052] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0053] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0054] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0055] Example 1:

[0056] This embodiment provides a method for intelligent updating of interactive content on a cloud platform, such as... Figure 1 As shown, a method for intelligently updating interactive content on a cloud platform includes:

[0057] S101. Based on the pre-acquired user operation data, analyze and calculate the heat value of user operation behavior, and construct a click heat map of the cloud platform.

[0058] S102. Using a preset intent analysis model, analyze the user's operational intent from the click heatmap on the cloud platform and construct an intent feature vector;

[0059] S103. Based on the intent feature vector, the user intent is classified according to the preset content optimization model, and the interactive content of the cloud platform is layered and optimized according to the classified user intent to obtain the first interactive content.

[0060] S104. Within a preset time period, update the cloud platform click heatmap, analyze user interaction with the first interactive content based on the updated cloud platform click heatmap, and update the first interactive content accordingly, so as to intelligently update the cloud platform interactive content.

[0061] This embodiment analyzes user operation data to construct a cloud platform click heatmap. Based on the cloud platform click heatmap, it analyzes user operation intentions, classifies user operation intentions, and optimizes the interactive content of the cloud platform in layers. Finally, it updates the interactive content in real time based on user feedback. By analyzing user operation intentions based on user operation data and optimizing interactive content accordingly, it can meet the needs of different users, increase the frequency and depth of user interaction with the platform, dynamically update the cloud platform interactive content, adapt to changes in user needs, and enhance the flexibility and adaptability of the cloud platform interactive content update process.

[0062] In this embodiment, based on the collected user operation data, various user behaviors on the cloud platform are analyzed, specifically including clicks, pauses, and swipes. Different behaviors reflect the degree of user attention to different content areas. By analyzing user operation data, heat values ​​of different areas of the cloud platform interface are calculated. These heat values ​​reflect user clicks on different content areas, thus reflecting the hot spots of user attention on the cloud platform. By analyzing user operation data and constructing heat maps, the hot spots of user behavior on the cloud platform can be displayed intuitively, thereby quickly filtering out interactive content areas that users are interested in and updating the content accordingly for different areas.

[0063] Specifically, based on the constructed cloud platform click heatmap, user operational intent is analyzed, including intents such as browsing information, searching for specific content, and making purchases. A pre-defined intent analysis model is used to analyze the cloud platform click heatmap, extracting features that reflect user intent and constructing intent feature vectors. By constructing these intent feature vectors, the true intent of user operations can be analyzed, transforming abstract user needs into concrete feature vectors. This provides a clear direction for optimizing interactive content, making content optimization more targeted.

[0064] Preferably, the intensity of user intent is analyzed based on intent feature vectors. Different users have different levels of operational intent and different needs for content. User intent is classified into different levels, such as high intent, medium intent, and low intent, through a preset content optimization model. The interactive content of the cloud platform is then optimized in layers according to different levels of intent, and different optimization methods are used for different levels of intent. Personalized interactive content can be provided according to the operational needs of different users, so that the content can accurately meet the operational needs of different users, improve user satisfaction and participation in interactive content, and thus improve the platform's conversion rate and user retention rate.

[0065] Meanwhile, within a preset time period, the cloud platform's click heatmap is updated based on real-time user operation data. The updated heatmap is then used to analyze user interaction with the primary interactive content. By analyzing user feedback on the interactive content, the cloud platform's interactive content is updated in real time. This approach takes into account the dynamic nature of user needs and behaviors, allowing for timely identification and adjustment of issues with the interactive content based on user feedback. This ensures that the cloud platform's interactive content remains attractive and practical, thereby improving user experience and the platform's competitiveness.

[0066] This application constructs a cloud platform click heatmap based on user operation data and the cloud platform's user interface. It analyzes user operation intentions in different areas based on this heatmap and employs different interactive content optimization strategies for different levels of user operation intentions. The heatmap is updated in conjunction with real-time user operation data, and the interactive content is optimized accordingly. By constructing the cloud platform click heatmap, the application can reflect the hotspots and distribution patterns of user operations on the interface, accurately grasp the user's true intentions, and provide a clear direction for content optimization. Layered optimization can meet the personalized needs of different users, improve user interaction experience and engagement, and allow for timely updates and optimization of interactive content based on user feedback, achieving intelligent updates of interactive content and improving the effectiveness of interactive content on the cloud platform.

[0067] Furthermore, based on the pre-acquired user operation data, heatmap values ​​are calculated to analyze user behavior and construct a cloud platform click heatmap, including:

[0068] S201. Based on the pre-acquired user operation data, analyze the user's click behavior on interactive content on the cloud platform on different devices to obtain click analysis results;

[0069] S202. On the cloud platform operation interface, calculate the heat values ​​of different locations on the interface based on the click analysis results, and construct a cloud platform click heat map.

[0070] This embodiment analyzes user click behavior on different devices based on pre-acquired user operation data, divides the cloud platform operation interface into a grid, and constructs a cloud platform click heatmap that reflects user click hotspots by calculating heat values ​​at different locations on the cloud platform operation interface. By combining user click behavior on different devices, it can analyze the interactive content areas that users are interested in. The constructed cloud platform click heatmap can intuitively display the hot and cold click areas of users on the cloud platform interface, thereby quickly locating the content areas and functional modules that need optimization, avoiding blind optimization, and improving resource utilization efficiency and content optimization effect.

[0071] In this embodiment, based on the pre-acquired user operation data, the user's click behavior on different devices is analyzed. Different devices have different screen sizes, operation methods, and interactive interfaces, and users' click behavior on interactive content on the cloud platform also differs on different devices. Specifically, devices include mobile phones, tablets, computers, etc. By analyzing users' preferences for content and operating habits in different device environments, the differences in users' click behavior on different devices can be combined to analyze the interactive content that users are interested in, thus avoiding poor user interaction experience caused by device differences.

[0072] Specifically, based on the pre-acquired user operation data, key features of click behavior are extracted for each device type dataset, including click location coordinates, click time, click frequency, and click content type. Click content type includes product images, text links, buttons, etc. The click behavior features of users on different devices are overlaid to obtain click analysis results.

[0073] Preferably, based on the click analysis results, user click behavior on the cloud platform interface is quantified, and heatmap values ​​are calculated. By analyzing user click behavior in different areas of the cloud platform interface, heatmap values ​​are calculated for each area, and these heatmap values ​​are mapped onto the cloud platform interface. Different colors and shades are used to construct a cloud platform click heatmap reflecting user click hotspots. By constructing a cloud platform click heatmap, the distribution of user click hotspots on the cloud platform can be visually displayed, quickly distinguishing areas of high and low user attention. When a low heatmap value is found in an important functional area, the content layout or functional design can be adjusted promptly to increase user attention and usage of that area.

[0074] Furthermore, on the cloud platform's operation interface, heat values ​​for different locations on the interface are calculated based on the click analysis results, constructing a cloud platform click heatmap, including:

[0075] S301. Divide the cloud platform operation interface into grids according to the preset grid size to obtain multiple grid areas;

[0076] S302. For each grid area, the click analysis results are analyzed using a preset thermal analysis model to calculate the thermal value;

[0077] S303. Construct a thermal value matrix based on the thermal values ​​of each grid region;

[0078] S304. Construct a click heatmap for the cloud platform based on the heat value matrix.

[0079] like Figure 2 As shown, this embodiment divides the cloud platform operation interface into grids, calculates the thermal value of each grid area using a preset thermal analysis model, constructs a thermal value matrix, and generates a cloud platform click heatmap. Figure 2 The thickness of lines in the heatmap represents the intensity of color, with thicker lines indicating a darker color and a higher heat value for the corresponding user, indicating more frequent clicks in that area. Through grid division, heat value calculation, and heatmap construction, the system can quantify user interest levels by combining user click behavior, analyze areas of high and low user attention, and provide data support for user intent analysis and content optimization strategy formulation, thereby effectively improving the user experience and operational efficiency of the cloud platform.

[0080] In this embodiment, firstly, the cloud platform operation interface is divided into multiple grid areas according to a preset grid size. Different grid sizes are set according to the size, resolution, and content optimization accuracy requirements of the cloud platform operation interface. For a computer interface with a resolution of 1920×1080, the grid size can be set to 20×20 pixels. According to the set grid size, the interface is divided horizontally and vertically into multiple grid areas of the same size. The grid division makes the statistics and analysis of click data more standardized and avoids differences in analysis results caused by inconsistent interface area sizes.

[0081] Secondly, after dividing the cloud platform's interface into grids, a preset heatmap analysis model is used to analyze the click analysis results within each grid area, calculating the heatmap value for that area. A higher heatmap value indicates greater user attention and more frequent user activity within that grid area. Based on user behavior, factors influencing the heatmap value calculation and their corresponding weights are determined. These factors include click count, click duration, and user dwell time, with each factor weighted at 0.4, 0.3, and 0.3 respectively. The heatmap analysis model is a weighted model, combining the influencing factors and their corresponding weights to calculate the heatmap value. Click analysis results are input into the heatmap analysis model to analyze the factors influencing the heatmap value calculation, and a weighted sum is performed to calculate the heatmap value for each grid area. By calculating the heatmap value, different user behavior factors can be comprehensively considered, accurately reflecting the degree of user attention to different areas.

[0082] Preferably, the heat values ​​calculated for each grid area are arranged according to the row and column order of the grid on the interface to construct a heat value matrix. The heat value matrix neatly arranges the heat value information of each grid area of ​​the operation interface, providing standard data support for the construction of heat maps and improving the efficiency of heat map construction. Based on the constructed heat value matrix, the values ​​in the heat value matrix are mapped to image elements of different colors and shades according to the set color mapping rules, generating a cloud platform click heat map reflecting the distribution of user click hotspots. The darker the color, the higher the heat value and the higher the corresponding user attention. The correspondence between heat values ​​and colors is set, with heat values ​​of 0-10 corresponding to light blue, 11-20 to blue, 21-30 to dark blue, and 31 and above to red, and the color depth increases with the increase of heat value. By constructing a cloud platform click heat map, the distribution of user click hotspots on the cloud platform operation interface can be intuitively reflected. When updating interactive content, the interface area that needs to be optimized can be quickly located, improving the efficiency of interactive content optimization.

[0083] Furthermore, by using a pre-defined intent analysis model, the user's operational intent is analyzed from the click heatmap on the cloud platform to construct an intent feature vector, including:

[0084] S401. Based on the cloud platform click heatmap, perform gradient division on the cloud platform operation interface and calculate the corresponding gradient value, then filter out the first gradient region whose gradient value is greater than the preset gradient threshold.

[0085] S402. Analyze the user's operational intent within the first gradient region using a preset intent analysis model, and construct an intent feature vector.

[0086] This embodiment performs gradient analysis on the cloud platform's click heatmap. Based on the gradient of pixel values ​​in the heatmap, it locates hotspot areas where users frequently operate. Using a preset intent analysis model, it analyzes the user's operational intent in these hotspot areas and constructs intent feature vectors. Through gradient analysis and intent feature calculation, it can accurately identify the user's operational intent on the cloud platform, improving the accuracy and reliability of intent analysis. The intent feature vectors provide direction for optimizing interactive content on the cloud platform, enabling personalized content updates that better match the user's intent, thereby enhancing user experience and platform competitiveness.

[0087] In this embodiment, based on the constructed cloud platform click heatmap, the gradient changes of pixel values ​​in the heatmap are analyzed. The gradient of pixel values ​​in the heatmap reflects the intensity of heat value changes. In areas with strong user operation intent, the heat value changes more strongly, corresponding to higher gradient values, forming high gradient regions. The Sobel operator is used to perform a convolution operation on the cloud platform click heatmap to calculate the gradient value of each pixel in the horizontal and vertical directions. The gradient values ​​in the horizontal and vertical directions are superimposed to obtain the gradient value of the grid region. By statistically analyzing the gradient value distribution of all pixels, the gradient threshold is set to the mean plus 2 standard deviations. At this point, the selected region is the area with the most significant heat value changes. The gradient value of each pixel in the heatmap is traversed, and pixels with gradient values ​​greater than the preset gradient threshold are selected to form the first gradient region. Through gradient analysis and threshold selection, the key areas of user operation can be selected, improving the targeting and accuracy of intent analysis.

[0088] Specifically, within the selected first-gradient region, the user's operational intent is analyzed using a pre-defined intent analysis model to extract the user's intent features and construct an intent feature vector. The intent feature vector can transform the user's abstract operational intent into a specific feature vector. When optimizing interactive content, targeted content optimization can be performed based on the intent feature vector to improve the effectiveness of interactive content optimization.

[0089] Furthermore, using a pre-defined intent analysis model, the user's operational intent within the first gradient region is analyzed to construct an intent feature vector, including:

[0090] S501. Within the first gradient region, the user's operational intent is analyzed using a preset intent analysis model to obtain a first intent feature vector.

[0091] S502. According to the preset feature mapping rules, the first intent feature vector is mapped to the cloud platform operation interface to obtain the second intent feature vector.

[0092] S503. Combine the first intention feature vector and the second intention feature vector to construct the intention feature vector.

[0093] In this embodiment, within the first gradient region, the user's operational intent is analyzed using a preset intent analysis model to obtain a first intent feature vector. The first gradient region is an area where user operations are active and intents are obvious. The intent analysis model analyzes user click behavior data within this region, including click count, click duration, frequency, etc., to analyze the user's operational intent in this region and convert the intent into a feature vector. First, user operation data within the first gradient region is extracted. The intent analysis model includes models such as support vector machines and random forests. In this embodiment, the intent analysis model is a random forest model. The random forest model is trained using a large amount of historical user operation data and corresponding intent annotations to obtain a pre-trained random forest model. The user operation data from the first gradient region is input into the pre-trained intent analysis model, and the model outputs the user operation intent category and the confidence score of each intent category. By combining the confidence scores of each intent category, the first intent feature vector is obtained.

[0094] For example, if the intent analysis model outputs a purchase intent confidence score of 0.7, a view details intent confidence score of 0.2, and a share intent confidence score of 0.1, then the first intent feature vector is [0.7, 0.2, 0.1]. Analyzing user intent in the first gradient region allows for intent analysis based on the core area of ​​user behavior, reducing data interference, improving the accuracy and efficiency of intent recognition, and quickly obtaining preliminary intent information of users in key areas.

[0095] Specifically, according to preset feature mapping rules, the first intent feature vector is mapped onto the overall operation interface of the cloud platform to obtain the second intent feature vector. Different cloud platform operation interface layouts and functional module distributions will affect the user's operation intent. Through vector mapping, the first intent feature vector is associated with the position, function, and other information of the operation interface, mapping the intent feature to a specific interface scenario. Actual operation scenario information is added to the intent feature to supplement the interface dimension information missing in the first intent feature vector. Based on the layout and function of the cloud platform operation interface, feature mapping rules are set. The feature mapping rules set in this embodiment include: when the first gradient area is located in the top navigation bar of the interface and the search intent confidence in the first intent feature vector is high, the value of the feature dimension related to the navigation bar search function in the second intent feature vector is increased; when the first gradient area is located in the product display area, the feature dimension values ​​of functions such as viewing product details and adding to the shopping cart are adjusted according to the first intent feature vector; the specific position and functional module of the first gradient area in the cloud platform operation interface are determined, and the first intent feature vector is adjusted and expanded according to the feature mapping rules.

[0096] For example, the first intent feature vector is [0.7, 0.2, 0.1], representing a purchase intent confidence level of 0.7, a view details intent confidence level of 0.2, and a share intent confidence level of 0.1. The first gradient region is located in the product display area. According to the feature mapping rules, the values ​​of dimensions related to functions such as adding to the shopping cart and viewing product details are weighted and calculated based on the original intent confidence levels. By adding new dimensions to represent intent features related to the functions of the product display area, the second intent feature vector is obtained. By combining intent features with the cloud platform operation interface scenario, more interface dimension information can be added to the intent feature vector, making the intent analysis results more consistent with the user's true intent in actual interface operations, thereby improving the accuracy and optimization effect of interactive content recommendation and interface optimization.

[0097] Preferably, the first intent feature vector reflects the user's operational intent within the first gradient region, while the second intent feature vector supplements the platform interface scene information. By combining the first and second intent feature vectors through direct concatenation of vector elements, multi-dimensional information about the user's operational intent can be integrated, avoiding the limitations of a single vector, improving the completeness and accuracy of the intent feature vector, and providing rich data support for optimizing interactive content on the cloud platform.

[0098] Furthermore, based on the intent feature vector, user intents are categorized using a pre-defined content optimization model. Then, the interactive content on the cloud platform is layered and optimized according to the categorized user intents to obtain the first interactive content, including:

[0099] S601. Based on the intent feature vector, analyze the intensity of the user's intent through a preset content optimization model, classify the user's intent, and obtain multi-level user intent.

[0100] S602. Based on multi-level user intents, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content. The multi-level user intents include high-level user intents, intermediate-level user intents, and low-level user intents.

[0101] This embodiment analyzes the intent feature vector using a content optimization model, classifies user intent into intensity levels, and optimizes the display format and recommendation strategy of interactive content for different levels of user intent. This achieves precise matching between content and user needs, enabling users to quickly obtain content that meets their needs, thereby improving user experience and platform operational efficiency.

[0102] In this embodiment, based on the intent feature vector, the intensity of the user's intent is analyzed using a preset content optimization model, and the user intent is graded to obtain multi-level user intent. The intent feature vector includes multi-dimensional information about the user's operation intent. First, the intent feature vector is standardized using the Z-score standardization method to eliminate the influence of differences in the dimensions. The content optimization model includes models such as logistic regression, decision trees, and neural networks. In this embodiment, the content optimization model uses a decision tree model. The decision tree model is trained using a large number of historical intent feature vectors and their corresponding labeled intent levels to obtain a pre-trained decision tree model. The preprocessed intent feature vector is input into a pre-trained decision tree model. The model outputs the predicted probability for each intent level. The criteria for classifying user intents are as follows: when the intent probability is greater than or equal to 0.7, it is classified as a high-level user intent; when the intent probability is greater than 0.3 and less than 0.7, it is classified as a medium-level user intent; and when the intent probability is less than or equal to 0.3, it is classified as a low-level user intent. Classifying user intents by level allows for the development of different content optimization strategies for different intent levels. This enables the platform to update content to different degrees for different intent levels, reducing resource waste while achieving content update optimization.

[0103] Specifically, based on the different levels of user intent, the interactive content on the cloud platform is layered and optimized to obtain the first interactive content. Different levels of user intent correspond to different levels of content optimization needs. By layering the interactive content on the cloud platform and setting different interactive content display and interaction methods for users with different intent levels, the matching degree between content and user needs can be improved. This avoids the waste of resources caused by using a single fixed content optimization method, achieves accurate matching between interactive content and user intent, improves content conversion rate and user satisfaction, and enhances resource utilization.

[0104] Furthermore, such as Figure 3 As shown, based on multi-level user intent, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content, including:

[0105] S701. For advanced user intents, generate the first optimized instruction by generating a real-time interactive scenario;

[0106] S702. For intermediate user intent, a second optimization instruction is generated by dynamically reorganizing and enhancing the interactive content.

[0107] S703. For low-level user intent, generate a third optimization instruction by adjusting the dynamic effects of the interactive content;

[0108] S704. Combining the first optimization instruction, the second optimization instruction, and the third optimization instruction, the interactive content of the cloud platform is optimized in layers to obtain the first interactive content.

[0109] In this embodiment, for advanced user intents, a first optimized instruction is generated by generating a real-time interactive scenario. Advanced user intents are strong and clear, and their needs can be quickly met by generating a real-time interactive scenario, which can simulate a real interactive environment. Based on the cloud platform business type and the direction of the advanced user intent, the type of real-time interactive scenario is determined. According to the determined scenario type, corresponding interface interactive elements are set, including interactive interface layout, prompt information, operation buttons, etc. The designed real-time interactive scenario is transformed into a first optimized instruction, which includes specific parameters of the scenario elements, display rules, interaction logic, and other information.

[0110] For example, in an e-commerce live streaming platform, the first optimization instruction is set as follows: display a countdown module for a limited-time flash sale in the upper right corner of the product details page, and hide the module after the countdown ends; clicking the "Buy Now" button will redirect to the payment page; refresh the remaining inventory information every 10 seconds; for advanced user intent, provide users with an efficient operation path through real-time interactive scenarios, shorten the conversion process, improve the conversion rate, and enhance the user experience and satisfaction on the platform.

[0111] For intermediate-level user intent, a second optimization instruction is generated by dynamically reorganizing and enhancing interactive content. Intermediate-level users have some interest in the content but their needs are not yet clear. This is achieved by dynamically reorganizing and enhancing interactive content, refining it, highlighting key information, and optimizing its effectiveness. Existing interactive content is analyzed to extract information relevant to intermediate-level user intent. Based on the relevance and importance of this information, the interactive content is rearranged and combined. Visual design and copywriting optimization can be used to enhance the content's presentation, specifically adjusting text layout, color schemes, image size, and clarity to make the content more attractive. The content reorganization and enhancement scheme is then used to generate a second optimization instruction, including the specific methods of content reorganization and parameter settings for enhancement. Through content reorganization and enhancement, more attractive content can be presented to intermediate-level users, increasing user dwell time and engagement on the platform.

[0112] For low-level user intent, the third optimization command is generated by adjusting the dynamic effects of interactive content. Low-level user intent is weak, indicating insufficient interest in platform content. Adjusting the dynamic effects of interactive content, specifically adding animations, special effects, and sound effects, attracts user attention and guides them to explore platform content. Based on the cloud platform's style and elements of user interest, appropriate dynamic effects are selected, including flashing, bouncing, and explosion animations. Relevant parameters for these dynamic effects are designed, including animation speed, duration, frequency of occurrence, and sound volume. The selected dynamic effects and parameter settings are then used to generate the third optimization command. Generating dynamic effects attracts users, lowers the initial cognitive threshold for platform content, stimulates user curiosity and exploration, and encourages further browsing and use of the platform.

[0113] Preferably, the interactive content of the cloud platform is optimized in layers by combining the first optimization instruction, the second optimization instruction, and the third optimization instruction to obtain the first interactive content; the optimization instructions generated for different user intent levels are integrated, and the optimization instructions are applied to the corresponding interactive content of the cloud platform according to the user intent layers, so as to realize personalized content display and service for users with different intents, and enable the platform content to match the needs of users at different levels to achieve the best optimization effect.

[0114] Furthermore, within a preset time period, the cloud platform click heatmap is updated. The updated heatmap is then used to analyze user interaction with the primary interactive content, which is then updated accordingly. This intelligent updating of the cloud platform's interactive content includes:

[0115] S801. Within a preset time period, update the cloud platform click heatmap based on real-time user operation data to obtain an updated cloud platform click heatmap.

[0116] S802. Based on the updated cloud platform click heatmap, analyze user interaction with the first interactive content and construct an interaction effect vector.

[0117] S803. Update the first interactive content through interactive effect vectors to intelligently update the interactive content on the cloud platform.

[0118] This embodiment updates the cloud platform click heatmap within a preset time period, analyzes the constantly changing behavioral trends and needs of users, analyzes the user's interaction with the first interactive content based on the updated cloud platform click heatmap, constructs an interaction effect vector, and updates the first interactive content based on the interaction effect vector, realizing real-time intelligent updates of the cloud platform's interactive content, so that the platform's interactive content continuously matches user needs and is dynamically optimized.

[0119] In this embodiment, within a preset time period, the cloud platform click heatmap is updated based on real-time user operation data to obtain an updated cloud platform click heatmap. Within the time period, the changes in user operation behavior and interest preferences are analyzed, and the cloud platform click heatmap is updated based on real-time user operation data, updating the distribution range and color changes of hotspot areas in the heatmap. By updating the heatmap, the latest user click hotspots and behavioral trends can be reflected in a timely manner, providing accurate data support for the optimization of the first interactive content.

[0120] Specifically, based on the updated cloud platform click heatmap, user interaction with the first interactive content is analyzed to construct an interaction effect vector. The updated cloud platform click heatmap includes the distribution of user clicks on the first interactive content. By analyzing the heat values ​​and click behavior patterns in different areas of the heatmap, key features are extracted. Based on the cloud platform's business objectives and content characteristics, indicators for analyzing interaction are set, including click count, click conversion rate, average dwell time, and the proportion of heat values ​​in different areas. Data corresponding to each indicator is extracted from the updated cloud platform click heatmap and real-time user operation data. The values ​​corresponding to each indicator are arranged in order to obtain the interaction effect vector. By constructing the interaction effect vector, the effectiveness of the first interactive content can be quickly evaluated, and existing problems and corresponding optimization directions can be analyzed.

[0121] Preferably, the first interactive content is updated based on the calculated interactive effect vector to intelligently update the interactive content on the cloud platform. The interactive effect vector includes user feedback information on the first interactive content. Analyzing the values ​​of each indicator in the vector can determine the strengths and weaknesses of the interactive content, formulate corresponding optimization strategies, adjust and update the first interactive content to better meet user needs, optimize the content based on real-time user feedback, and enable the interactive content on the cloud platform to dynamically adapt to changes in user needs, improve the quality of interactive content and user experience, enhance the platform's attractiveness and stickiness to users, and improve the platform's operational effectiveness.

[0122] Furthermore, by using interactive effect vectors, the first interactive content is updated to intelligently update the interactive content on the cloud platform, including:

[0123] S901. Based on the interaction effect vector, the first interactive content is decomposed to obtain the first content, the second content, and the third content.

[0124] S902. Using a preset content update model, update the corresponding content in the first content, the second content, and the third content respectively to construct an updated content sequence;

[0125] S903. Based on the update content sequence, replace and update the corresponding interface elements in the first interactive content to intelligently update the interactive content on the cloud platform.

[0126] In this embodiment, the first interactive content is decomposed into first content, second content, and third content based on the interactive effect vector. Content decomposition rules are set according to the structure and functional characteristics of the cloud platform's interactive content. For the interactive content of the e-commerce product details page, it is functionally divided into a product display area, a user review area, and a purchase operation area. According to the decomposition criteria, the first interactive content is decomposed into first content, second content, and third content. Content decomposition allows for individual analysis and optimization of each part of the content, avoiding direct optimization of the entire content and improving the accuracy and efficiency of interactive content optimization.

[0127] Specifically, for the decomposed first, second, and third content, the content is updated separately using a preset content update model, and the updated content is combined to construct an updated content sequence. The content update model includes decision trees, neural networks, and other models. In this embodiment, the content update model is a neural network model. The neural network model is trained using a large amount of historical content data and corresponding user feedback data to obtain a pre-trained neural network model. Each part of the content is input into the trained content update model, and the model analyzes each part of the content and user feedback data to output corresponding update strategies and updated content. According to the display order or logical relationship of the content in the interface, the updated first, second, and third content are arranged in sequence to obtain the updated content sequence. By updating different content separately, corresponding optimization solutions can be provided for each part of the content, making the updated content more in line with user needs and behavioral habits.

[0128] Preferably, based on the updated content sequence, the corresponding original interface elements in the cloud platform interface are replaced with the updated content. After the interface element replacement is completed, the interface functionality is tested to check whether the updated content displays correctly and whether the functions are usable, thus achieving real-time updates of interactive content. The optimized content is then quickly replaced on the cloud platform operation interface, enabling the platform to respond promptly to user needs and behavioral changes, improving user experience and satisfaction. Content testing ensures the stability and reliability of the updated content, avoiding functional failures and user churn caused by updates.

[0129] Example 2:

[0130] In this embodiment, as Figure 4 This document provides a cloud platform interactive content intelligent update system, which implements a cloud platform interactive content intelligent update method, including:

[0131] The cloud platform click heatmap construction module analyzes and calculates heat values ​​based on pre-acquired user operation data to construct a cloud platform click heatmap.

[0132] The intent feature vector construction module analyzes the user's operational intent from the click heatmap of the cloud platform using a preset intent analysis model, and constructs an intent feature vector.

[0133] The interactive content optimization module classifies user intents based on intent feature vectors using a preset content optimization model, and then optimizes the interactive content on the cloud platform in layers according to the classified user intents to obtain the first interactive content.

[0134] The interactive content update module updates the cloud platform click heatmap within a preset time period. It analyzes user interaction with the first interactive content based on the updated cloud platform click heatmap and updates the first interactive content accordingly, thus intelligently updating the interactive content on the cloud platform.

[0135] In this embodiment, the cloud platform click heatmap construction module collects user operation data, specifically including user click locations, dwell time, operation frequency, and other behavioral information on the cloud platform. The cloud platform interface is divided into a grid, and a heat value is calculated for each area based on the user operation data. A higher heat value indicates a higher level of user attention to the interactive content in that area. A cloud platform click heatmap is constructed based on these heat values. By constructing the cloud platform click heatmap, user behavior is visualized, allowing for quick identification of key areas of user attention and neglected areas, providing data support for analyzing user intent and optimizing platform interface layout and content display. The intent feature vector construction module analyzes user operation intent based on the cloud platform click heatmap, selecting the first gradient area with strong user operation intent. Using a preset intent analysis model, features of user operation behavior within the first gradient area are extracted, including click count and click location distribution. The intent analysis model is used to analyze user operation intent and construct intent feature vectors. Combining the cloud platform click heatmap analysis with the construction of intent feature vectors transforms user intent into calculable and processable feature vectors, providing a basis for optimizing platform interactive content based on user intent.

[0136] Specifically, the interactive content optimization module analyzes user intent strength based on intent feature vectors using a pre-defined content optimization model, categorizing user intent into multiple levels such as high, medium, and low. Differentiated content optimization strategies are developed for different levels of user intent. For example, real-time interactive scenarios are generated for high-level user intent, content effects are dynamically reorganized and enhanced for medium-level user intent, and dynamic content effects are adjusted for low-level user intent. This layered optimization of cloud platform interactive content yields the first interactive content. Layered optimization allows for different degrees of optimization based on varying user intent strengths, improving the platform's interactive content optimization efficiency and enhancing user engagement and satisfaction. The interactive content update module updates the cloud platform's click heatmap within a pre-defined time period based on real-time user operation data. By analyzing the updated heatmap, it assesses user feedback on the first interactive content and uses a pre-defined content update model to update each part of the content, constructing an update content sequence to replace and update the cloud platform's interactive content. Real-time updates of interactive content allow for adaptation to evolving user needs and behaviors, preventing content from becoming outdated and improving user experience.

[0137] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligently updating interactive content on a cloud platform, characterized in that, include: Based on the pre-acquired user operation data, analyze and calculate the heat value of user operation behavior, and construct a click heat map of the cloud platform. Based on the cloud platform click heatmap, the cloud platform operation interface is divided into gradients and the corresponding gradient values ​​are calculated. The first gradient region with the gradient value greater than the preset gradient threshold is then selected. Within the first gradient region, the user's operational intent is analyzed using a pre-defined intent analysis model to obtain the first intent feature vector; According to preset feature mapping rules, the first intent feature vector is mapped to the cloud platform operation interface to obtain the second intent feature vector. The feature mapping rules include: if the first gradient region is located in the top navigation bar of the interface and the search intent confidence in the first intent feature vector is high, the value of the feature dimension related to the navigation bar search function in the second intent feature vector is increased; if the first gradient region is located in the product display area, the feature dimension values ​​of the product details viewing and adding to cart functions are adjusted according to the first intent feature vector; the specific location and functional module to which the first gradient region belongs in the cloud platform operation interface are determined, and the first intent feature vector is adjusted and expanded accordingly. Combine the first intent feature vector and the second intent feature vector to construct an intent feature vector; Based on the intent feature vector, user intent is classified using a preset content optimization model, and the interactive content of the cloud platform is layered and optimized based on the classified user intent to obtain the first interactive content. Within a preset time period, the cloud platform click heatmap is updated, and the user's interaction with the first interactive content is analyzed in combination with the updated cloud platform click heatmap. The first interactive content is then updated accordingly, so as to intelligently update the cloud platform interactive content.

2. The intelligent update method for interactive content on a cloud platform according to claim 1, characterized in that, The step of analyzing and calculating heatmap values ​​for user actions based on pre-acquired user action data, and constructing a cloud platform click heatmap, includes: Based on the pre-acquired user operation data, analyze the user's click behavior on interactive content on the cloud platform on different devices to obtain click analysis results; On the cloud platform operation interface, the heat values ​​of different locations on the interface are calculated based on the click analysis results to construct a cloud platform click heat map.

3. The intelligent update method for interactive content on a cloud platform according to claim 2, characterized in that, The process of calculating heat values ​​at different locations on the cloud platform operation interface based on the click analysis results and constructing a cloud platform click heatmap includes: The cloud platform operation interface is divided into multiple grid areas according to the preset grid size. For each grid area, the click analysis results are analyzed using a preset thermal analysis model to calculate the thermal value; Construct a thermal value matrix based on the thermal values ​​of each grid region; Based on the aforementioned heat value matrix, a click heatmap for the cloud platform is constructed.

4. The intelligent update method for interactive content on a cloud platform according to claim 1, characterized in that, Based on the intent feature vector, user intents are classified using a preset content optimization model, and the interactive content on the cloud platform is then layered and optimized according to the classified user intents to obtain the first interactive content, including: Based on the intent feature vector, the intensity of user intent is analyzed through a preset content optimization model, and user intent is classified into levels to obtain multi-level user intent; Based on the multi-level user intents, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content, wherein the multi-level user intents include high-level user intents, intermediate-level user intents, and low-level user intents.

5. The intelligent update method for interactive content on a cloud platform according to claim 4, characterized in that, Based on the multi-level user intent, the interactive content of the cloud platform is layered and optimized to obtain the first interactive content, including: For advanced user intent, generate the first optimized instruction by creating a real-time interactive scenario; For intermediate user intent, a second optimized instruction is generated by dynamically reorganizing and enhancing the interactive content. For basic user intent, a third optimization instruction is generated by adjusting the dynamic effects of the interactive content; By combining the first optimization instruction, the second optimization instruction, and the third optimization instruction, the interactive content of the cloud platform is optimized in layers to obtain the first interactive content.

6. The intelligent update method for interactive content on a cloud platform according to claim 1, characterized in that, The process of updating the cloud platform click heatmap within a preset time period, analyzing user interaction with the first interactive content based on the updated cloud platform click heatmap, and updating the first interactive content accordingly, to achieve intelligent updating of the cloud platform interactive content, includes: Within a preset time period, the cloud platform click heatmap is updated based on real-time user operation data to obtain an updated cloud platform click heatmap. Based on the updated cloud platform click heatmap, analyze user interaction with the first interactive content and construct an interaction effect vector; The first interactive content is updated using the interactive effect vector to intelligently update the interactive content on the cloud platform.

7. The intelligent update method for interactive content on a cloud platform according to claim 6, characterized in that, The first interactive content is updated using the aforementioned interactive effect vector to intelligently update the interactive content on the cloud platform, including: Based on the interaction effect vector, the first interaction content is decomposed to obtain the first content, the second content, and the third content. Using a pre-defined content update model, the corresponding content in the first, second, and third content is updated respectively to construct an updated content sequence; Based on the updated content sequence, the corresponding interface elements in the first interactive content are replaced and updated to intelligently update the interactive content on the cloud platform.

8. A cloud platform interactive content intelligent update system, characterized in that, A method for intelligently updating interactive content on a cloud platform as described in any one of claims 1 to 7, comprising: The cloud platform click heatmap construction module analyzes and calculates heat values ​​based on pre-acquired user operation data to construct a cloud platform click heatmap. The intent feature vector construction module analyzes the user's operation intent from the click heatmap of the cloud platform using a preset intent analysis model, and constructs an intent feature vector. The interactive content optimization module classifies user intents according to the intent feature vectors using a preset content optimization model, and then performs layered optimization of the interactive content on the cloud platform based on the classified user intents to obtain the first interactive content. The interactive content update module updates the cloud platform click heatmap within a preset time period, analyzes user interaction with the first interactive content based on the updated cloud platform click heatmap, and updates the first interactive content accordingly, thereby intelligently updating the cloud platform interactive content.

Images