Dynamic display adjusting method and system for multimedia interactive display

Abstract

The invention relates to the technical field of multimedia information processing, and discloses a dynamic display adjusting method and system for multimedia interactive display, and the method comprises the steps: obtaining user real-time interaction data, user role attributes, content type preferences and historical interaction data, and generating real-time interest distribution points according to the user real-time interaction data, the user role attributes, the content type preferences and the historical interaction data; predicting a user behavior in combination with historical interaction data, and generating a pre-loaded content triggering condition; generating a preloaded content set by integrating the preloaded content triggering condition, the user role attribute and the content type preference; acquiring physical parameters of equipment, and constructing a coordinate matrix and a transformation matrix of cross-equipment content presentation; the content is distributed to the target device through the transformation matrix, and finally the display content is dynamically adjusted according to the real-time interaction data. The problem that in the multimedia interactive display technology, the content cannot be smoothly presented in real time during multi-screen switching is solved, seamless connection of the content during multi-screen switching is achieved, and real-time smoothness of multi-screen switching is ensured.

Description

Technical Field

[0001] The present application relates to the field of multimedia information processing technology, and in particular to a dynamic display adjustment method and system for multimedia interactive presentation. Background Art

[0002] With the rapid development of digital technology, multimedia interactive display technology has been widely used in exhibitions, education and training, commercial displays, and other fields. By integrating visual presentation, interactive perception, and intelligent analysis, it provides users with an immersive interactive experience, enabling dynamic information transmission and personalized expression. In particular, multimedia interactive display technology is playing an increasingly important role in application scenarios such as smart exhibition halls and digital showrooms.

[0003] Currently, multimedia interactive display systems primarily utilize fixed-program content scheduling and trigger-based interaction. Pre-set scripts control device content playback, while basic sensors enable simple content switching. Some researchers are experimenting with inter-device data synchronization technologies, such as coordinating content playback across multiple terminals through a central server.

[0004] However, these technologies are still mainly based on passive response, and the system cannot dynamically predict the content to be presented based on the user's real-time interactive behavior, which leads to the inability to present content smoothly in real time when switching between multiple screens. Summary of the Invention

[0005] The embodiments of the present application provide a dynamic display adjustment method and system for multimedia interactive display, which solves the problem in multimedia interactive display technology that the content cannot be presented smoothly in real time when switching between multiple screens due to the lack of a real-time collaborative mechanism for cross-device spatial mapping and behavior prediction. This achieves seamless connection of content when switching between multiple screens and ensures the real-time smoothness of multi-screen switching.

[0006] The embodiment of the present application provides a dynamic display adjustment method for multimedia interactive presentation, the dynamic display adjustment method for multimedia interactive presentation comprising:

[0007] Obtaining a user portrait feature set, and generating a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes real-time user interaction data, user role attributes, content type preferences, and historical interaction data;

[0008] Predicting user expected behavior based on the user's real-time interest distribution points and the historical interaction data, and generating a trigger condition for preloading content;

[0009] generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference;

[0010] Obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix;

[0011] The preloaded content set is distributed to the target device through the transformation matrix, and the content presentation order is dynamically adjusted according to the real-time user interaction data to determine the dynamic display content.

[0012] Optionally, the step of predicting the user's expected behavior based on the user's real-time interest distribution points and the historical interaction data and generating a trigger condition for preloading content includes:

[0013] Calculating the semantic similarity between the user's real-time interest distribution points and the semantic tags in the semantic content library;

[0014] Semantically associating the semantic tags whose semantic similarity is greater than a preset similarity threshold, and prioritizing the semantic tags based on the weights corresponding to the semantic tags to generate a semantic association graph;

[0015] The user's expected behavior is predicted based on the semantic association graph and the historical interaction data, and the preloading content triggering condition is generated.

[0016] Optionally, the step of predicting the user's expected behavior based on the semantic association graph and the historical interaction data and generating the trigger condition for preloading content includes:

[0017] Extracting content semantics and content type features corresponding to the semantic tags in the historical interaction data based on the semantic tags in the semantic association graph;

[0018] Calculating a behavior prediction probability based on the content semantics, the content type features, and the content type preferences to generate the user expected behavior;

[0019] The preloading content triggering condition is generated according to the user real-time interaction data and the user expected behavior.

[0020] Optionally, the step of generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference includes:

[0021] Extracting the interactive media material associated with the preloaded content triggering condition, and generating a preliminary material set matching the interactive media material according to a semantic association rule;

[0022] Matching the preliminary material set with the content type preference to generate a material subset;

[0023] The semantic tag of the user role attribute is obtained, and the interactive media material in the material subset is filtered according to the correlation between the material subset and the semantic tag to generate the preloaded content set.

[0024] Optionally, after the step of generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference, the following steps are included:

[0025] If the number of content types in the preloaded content set is greater than or equal to the preset number of content types, sorting the preloaded content set by type priority according to the preference weights of the content type preferences;

[0026] The interactive media materials in the preloaded content set are sorted according to the sorting result and the real-time user interaction data to generate an updated preloaded content set.

[0027] Optionally, the steps of obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix include:

[0028] Acquire the physical parameter set, and generate a mapping ratio matrix from the source device to the target device according to the physical parameter set;

[0029] If the resolution difference of the mapping ratio matrix is ​​greater than a preset difference threshold, smoothing the mapping ratio matrix to generate the coordinate matrix;

[0030] A matrix transformation algorithm is used to calculate the smooth transition parameters corresponding to the coordinate matrix to generate the transformation matrix.

[0031] Optionally, the step of obtaining a user portrait feature set and generating a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes user real-time interaction data, user role attributes, content type preferences, and historical interaction data, comprises:

[0032] Acquiring the user real-time interaction data and preprocessing the user real-time interaction data to obtain interaction data in a unified format;

[0033] sorting the dwell time, sight direction, and operation frequency in the interaction data using a time series analysis algorithm to generate focus point weights for each content area;

[0034] Clustering the user role attributes and the content type preferences, and adjusting the focus weights;

[0035] The focus point weights and the historical interaction data are integrated to generate the user's real-time interest distribution points.

[0036] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention further provides a dynamic display adjustment system for multimedia interactive presentation, characterized in that the dynamic display adjustment system for user multimedia interactive presentation includes:

[0037] User portrait analysis module, used to collect and fuse multi-source data in real time to generate user portrait feature sets;

[0038] A behavior prediction module, configured to generate a preloaded content set based on the user portrait feature set;

[0039] A cross-device content rendering module, configured to generate a transformation matrix based on a set of physical parameters of a source device and a target device;

[0040] The content distribution and dynamic adjustment module is used to distribute the preloaded content set to the target device through the transformation matrix, and dynamically adjust the content presentation order according to the real-time user interaction data to determine the dynamic display content.

[0041] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention also provides a terminal device, including a memory, a processor, and a dynamic display adjustment program for multimedia interactive display stored on the memory and runnable on the processor. When the processor executes the dynamic display adjustment program for multimedia interactive display, the method described above is implemented.

[0042] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention also provides a computer-readable storage medium, on which a dynamic display adjustment program for multimedia interactive display is stored. When the dynamic display adjustment program for multimedia interactive display is executed by a processor, the method described above is implemented.

[0043] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0044] Obtain a user portrait feature set consisting of real-time user interaction data, role attributes, content preferences, and historical interaction data, and generate real-time interest distribution points based on this. Combine historical data to predict user behavior, generate trigger conditions for preloading content, and generate a preloaded content set based on the trigger conditions, role attributes, and content preferences. Obtain device parameters to construct a coordinate matrix and transformation matrix for cross-device content presentation, distribute the content in the preloaded content set to the target device through the transformation matrix, and finally dynamically adjust the display order based on the user's real-time interactive behavior. This effectively solves the problem in multimedia interactive display technology that the content cannot be presented smoothly and in real time when switching between multiple screens due to the lack of a real-time collaborative mechanism for cross-device spatial mapping and behavior prediction. This achieves seamless content connection when switching between multiple screens, ensuring real-time smoothness of multi-screen switching. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of a first embodiment of a dynamic display adjustment method for multimedia interactive presentation according to the present application;

[0046] Figure 2 This is a flow chart of a second embodiment of a dynamic display adjustment method for multimedia interactive presentation according to the present application;

[0047] Figure 3 This is a schematic diagram of the terminal structure of the hardware operating environment involved in an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to solve the problem of insufficient timeliness of content when switching between multiple screens in multimedia interactive display technology, this solution obtains a user portrait feature set, generates real-time interest distribution points based on the user portrait feature set, predicts user behavior based on historical interaction data, generates preloaded content trigger conditions, and generates preloaded content sets based on preloaded content trigger conditions, user role attributes, and content type preferences. The physical parameter sets of the source and target devices are obtained, and the coordinate matrix and transformation matrix for cross-device content presentation are constructed. The preloaded content set is allocated to the target device through the transformation matrix. Finally, the display order of the preloaded content is dynamically adjusted according to the user's real-time interaction data to obtain the final output of the multimedia interactive display. This achieves seamless content connection when switching between multiple screens, ensuring real-time smoothness of multi-screen switching.

[0049] To better understand the above technical solutions, exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0050] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0051] Example 1

[0052] In this embodiment, a dynamic display adjustment method for multimedia interactive presentation is provided.

[0053] Reference Figure 1 The dynamic display method for multimedia interactive presentation of this embodiment includes the following steps:

[0054] Step S100: Obtain a user portrait feature set, and generate a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes user real-time interaction data, user role attributes, content type preferences, and historical interaction data;

[0055] In this embodiment, the user portrait feature set includes real-time user interaction data, user role attributes, content type preferences, and historical interaction data. User interaction data includes the user's dwell time on a certain interface, the user's line of sight, and the frequency of operation. User role attributes are classification identifiers that describe user characteristics, behavior patterns, and needs. User role attributes can be obtained based on the information filled out by the user when the user first registers, or they can be modeled based on the user's behavioral characteristics, or through a dynamic role calibration mechanism, the current user role attributes are calibrated according to a certain ratio based on the user's registration information, behavior patterns, and social graph. Content type preference refers to the user's preference for different types of content, such as learning and education, news and entertainment, etc., and can be obtained through questionnaires or historical behavior.

[0056] As an optional implementation method, after obtaining the real-time user interaction data, it is pre-processed to unify the format of the interaction data for easy subsequent processing. The dwell time, sight direction, and operation frequency in the unified format of the real-time user interaction data are sorted by a time series analysis algorithm to generate the focus point weights for each content area; the user role attributes and content type preferences are then clustered, and the focus point weights are adjusted according to the clustering results. The focus point weights are fused with the historical interaction data to generate the user's real-time interest distribution points. Data preprocessing ensures data quality, time series analysis and clustering optimize weight calculation, and data fusion and visualization intuitively present the results, which improves the intelligence level of the interactive interface and also provides reliable support for content recommendation and user behavior analysis.

[0057] Exemplarily, the touch position and operation frequency are extracted from the sensor, and the sight track is extracted from the camera, and these data are cleaned to remove noise. If there is missing data, an interpolation algorithm can be used to fill it in and obtain interaction data in a unified format. Based on the interaction data, a time series analysis algorithm is used to arrange the dwell time, sight direction and operation frequency in chronological order, and the focus weight of each content area is calculated. If the dwell time exceeds the preset dwell threshold, the weight value is increased to obtain a quantified focus weight. The user role attributes and content type preferences are obtained, and the user role attributes and content type preferences are matched using a clustering algorithm. The focus weight is adjusted according to the matching results. If the user role prefers a specific type of content type, the weight of the relevant content type is increased to determine a personalized weight distribution. Through a data fusion algorithm, the personalized weight distribution is combined with the user's real-time interaction data to generate the user's real-time interest distribution points. A visualization tool is used to draw an interest heat map to obtain a dynamically updated point of interest distribution.

[0058] It's important to note that personalized weight distribution relies on interaction data, user role attributes, and content type preferences. Interaction data here refers to long-term historical data, which is used to calculate weights and build user preference models and is updated daily or weekly. After generating the personalized weight distribution, real-time user interaction data is combined to generate a real-time user interest distribution. Real-time user interaction data here refers to the current session's real-time data and is used to dynamically modify the interest distribution, updating every second or minute.

[0059] As another alternative implementation, after sensors collect raw data such as dwell time, gaze direction, and operation frequency, they use a time series analysis algorithm to sort the raw data and calculate the weight of the attention points. This quantitative method can accurately reflect user attention and optimize content recommendation results.

[0060] For example, a time series analysis algorithm can use a sliding window approach to chronologically arrange dwell time, gaze direction, and operation frequency. Assuming a window size of 5 seconds, if the user spends 4 seconds in a certain content area, exceeding the preset threshold of 3 seconds, the weight is increased by 0.2, and the base weight of 0.5 is adjusted to 0.7. If the gaze direction is concentrated in this area, the weight is increased by another 0.1, for a final weight of 0.8.

[0061] Optionally, user role attributes and content type preferences can also adjust focus weights. For example, if a user is a student with a preference for educational content, a clustering algorithm can be used to categorize users into groups like education and entertainment, and their weights adjusted accordingly. If a student user focuses on educational content, the weight increases by 20%, from 0.8 to 0.96. This personalized adjustment ensures that the resulting preloaded content set is more tailored to the user's needs.

[0062] As another optional implementation, in order to make the user interest distribution result more intuitive, the user interests may be presented by drawing a visual interest heat map.

[0063] For example, the weight of operation frequency can be set to 40%, and the weight of gaze retention can be set to 60%. The comprehensive score of each area can be calculated to generate an interest heat map. Then, a visualization tool can be used to draw the heat map. Red marks high-attention areas, and green marks low-attention areas. Dynamic updates are refreshed every 5 seconds.

[0064] Step S200: predicting user expected behavior based on the user's real-time interest distribution points and the historical interaction data, and generating a trigger condition for preloading content;

[0065] In this embodiment, a semantic analysis algorithm is used to extract semantic tags matching the user's points of interest from the semantic content library based on the user's real-time distribution of points of interest. This generates a content semantic association graph and determines the priority sequence for content presentation. Using this content semantic association graph, a deep learning model is used to predict the probability of the user's expected behavior. Combined with historical interaction data, a user behavior prediction model is generated to determine the trigger conditions for preloading content. The user's expected behavior refers to the predicted next interaction behavior of the user, and the preloading content trigger conditions refer to the trigger conditions for preloading content.

[0066] As an optional implementation, when generating the trigger conditions for preloading content, the semantic similarity between the user's real-time interest distribution points and the semantic tags in the semantic content library is first calculated, and semantic tags with semantic similarity greater than a preset similarity threshold are semantically associated. The semantic tags are prioritized based on the weights corresponding to the semantic tags to generate a semantic association graph. The user's expected behavior is predicted based on the semantic association graph and the historical interaction data to generate the trigger conditions for preloading content. Semantic analysis ensures accurate content matching, and the semantic association graph can clearly map the content relationship. The sorting algorithm optimizes the presentation order, making the final prediction of the user's expected behavior more accurate.

[0067] For example, a semantic analysis tool calculates semantic similarity between semantic tags associated with the distribution of user interest points in a semantic content library, determining their similarity with the semantic tags associated with the distribution of user interest points. Semantic tags with similarity greater than a preset similarity threshold are then connected and weighted using an association graph generation algorithm to generate a content semantic association graph. A sorting algorithm is then used to sort the weights between the nodes in descending order to determine the priority sequence for content presentation. Finally, historical interaction data is combined to predict user expected behavior and generate trigger conditions for preloading content. For example, a semantic content library contains a variety of interactive media assets, including videos, images, and text. Each asset is labeled with semantic tags, such as "algebra," "algorithm," "basic programming," "culinary tutorials," and "healthy diet." Using word embedding technology, the semantic analysis tool calculates the similarity between the user interest keyword "mathematics" and the semantic tag "algebra." Assume the score is 0.85, exceeding the preset similarity threshold of 0.7, while the similarity with "algorithm" is 0.6, falling below the threshold. By screening out highly similar tags, the predicted content is ensured to be relevant to user needs. Furthermore, through semantic association, relevant resources can be precisely targeted.

[0068] Optionally, the semantic association graph generation algorithm constructs semantic labels as nodes, with connections between nodes based on label co-occurrence frequency. For example, if "algebra" and "math tutorial" co-occur in multiple articles, their connection weight is 0.9; whereas "algebra" and "basic programming" co-occur less frequently, their weight is 0.3. The resulting content semantic association graph displays label relationships using nodes and weighted edges, intuitively reflecting the strength of associations between content. This structured representation facilitates subsequent sorting and recommendation.

[0069] As another optional implementation, after generating a semantic association graph, based on the semantic tags in the semantic association graph, the content semantics and content type features corresponding to the semantic tags in the historical interaction data are extracted. The behavior prediction probability is calculated based on the content semantics, the content type features and the content type preferences, and the user's expected behavior is generated. The pre-loaded content trigger condition is generated based on the user's real-time interaction data and the user's expected behavior. In particular, when extracting content semantics and content type features through a semantic association graph, it can be implemented based on semantic tags and semantic analysis tools. The semantic association graph represents semantic tags through nodes, and edges represent semantic relationships between tags. In addition, the system also generates a semantic vector for each content. For example, the vector of a food video may be close to the vector of the "healthy diet" tag, indicating a high degree of semantic association.

[0070] For example, after extracting content semantics and content type features, we combine interaction frequency and user preferences with time series analysis to generate a multidimensional feature vector that comprehensively characterizes user behavior patterns. For example, if a user watches food videos an average of 10 times per week and prefers "Chinese cooking," we integrate the interaction frequency, preference label, and time series features into a feature vector [10, Chinese cooking, periodic peaks]. This multidimensional feature vector is then fed into a convolutional neural network, where the convolutional layer extracts local features, such as the user's preference for "cooking tips" videos. Combining the time series and interaction frequency, we calculate the behavior prediction probability and determine the user's expected behavior. Preloading content trigger conditions are generated based on real-time user interaction data and expected user behavior.

[0071] As another optional implementation, after determining the user's expected behavior, the prediction accuracy of the prediction result is determined. If it exceeds a preset accuracy threshold, the trigger conditions for preloading content are determined based on the behavior prediction probability and user preferences, resulting in a trigger condition set. After obtaining the trigger condition set, the content in the trigger condition set can also be prioritized based on the user's real-time interactive behavior to determine the final content loading instructions. By predicting and preloading content that users are interested in, the allocation efficiency of platform resources can be optimized, while also reducing user waiting time and improving the smoothness of content acquisition.

[0072] For example, if the feature vector [10, Chinese cooking, periodic peak] indicates that the user's expected behavior is more likely to watch "Chinese cooking" videos on weekends with a probability of 0.85, and the prediction accuracy exceeds the preset accuracy threshold of 0.8, the prediction result is reliable. Combined with the user's preference for Chinese recipes, the resulting trigger condition might be "preload Chinese cooking videos on Friday night." If multiple trigger conditions exist, a set of trigger conditions is generated, prioritized based on the user's real-time interactive behavior, and an execution sequence is generated to determine the final content loading instruction. If Chinese cooking videos have a higher priority than dessert videos, the execution sequence "load Chinese cooking videos first, then dessert videos" is generated, resulting in the final loading instruction "preload video ID 123 at 8:00 PM on Friday."

[0073] It should be noted that the trigger condition set needs to take into account device performance and network status to ensure smooth loading. For example, the system usually triggers loading when the user's device is idle and has a wireless network connection.

[0074] Step S300: generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference;

[0075] In this embodiment, based on the preloading content triggering conditions, a distributed computing framework is used to extract interactive media materials that meet semantic associations from the semantic content library, and a preloading content set is generated by combining the semantic tags of content type preferences and user role attributes to determine a candidate list of content presented across devices.

[0076] As an optional implementation, interactive media materials associated with the preloaded content triggering conditions are extracted, and a preliminary material set matching the interactive media materials is generated according to semantic association rules. The preliminary material set is matched with the content type preference for similarity to generate a material subset, and the semantic label of the user role attribute is obtained. The interactive media materials in the material subset are filtered according to the association between the material subset and the semantic label to generate the preloaded content set.

[0077] Exemplarily, a distributed computing framework is used to extract the interactive media materials associated with the preloaded content triggering conditions from the semantic content library. In order to extract more comprehensive material content, the materials associated with the interactive media materials can be further extracted according to the semantic association rules to generate a preliminary material set. Since the preliminary material set generated according to the semantic association may contain interactive media materials that do not conform to the user's preferences, such as PDF type files in the preliminary material set, while 80% of the user's historical operations are video files, it is necessary to match the preliminary material set with the content type preference by correlation, extract interactive media materials with a correlation higher than a threshold, and generate a material subset. The preloaded content set is generated based on the material subset and the user role attributes.

[0078] Optionally, when matching semantic associations to generate a preliminary material set, the system analyzes the relationship between materials based on a semantic association graph. The semantic association graph uses materials as nodes, and the edge weights between nodes represent the semantic similarity of the materials. For example, the video "Red Braised Pork Tutorial" has a close semantic distance because its content theme is highly relevant to the label "Chinese Home Cooking". On this basis, the pre-trained semantic model is used to convert the materials into vector representations, and the semantic association strength is quantified by calculating the distance between vectors. Finally, materials with a semantic association with the target theme higher than a set threshold are screened out, such as materials with a semantic association with "Chinese cooking" higher than a threshold of 0.8, forming a preliminary material set to provide high-quality basic data support for subsequent precise recommendations or content processing.

[0079] As another optional implementation, the resources of the preloaded content set are limited, and it is necessary to prioritize loading the materials most relevant to the user. The material subset may contain content that is incompatible with the user's role attributes. Therefore, it is necessary to adapt the correlation between the material subset and the semantic labels of the user's role attributes, filter out the final interactive media materials, and generate the preloaded content set.

[0080] For example, if the material subset is a collection associated with "Chinese cooking", the final interactive media materials are filtered out through the semantic tags of the user role attributes. If the user role is "newcomer in the workplace", the semantic tags can be used to filter out "quick cooking skills" materials instead of "complex banquet recipes", thereby reducing the interference of irrelevant content and improving the accuracy of recommended content.

[0081] Step S400: obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix;

[0082] In this embodiment, screen size, resolution, and pixel density parameters are obtained from the source device and the target device, and the physical parameter set is calculated using a standardized formula.

[0083] As an optional implementation, after obtaining the physical parameter set, a mapping ratio matrix from the source device to the target device is generated based on the physical parameter set. If the resolution difference of the mapping ratio matrix is ​​greater than a preset difference threshold, the mapping ratio matrix is ​​smoothed to generate a smoothed coordinate matrix. The final physical coordinates of the content presentation are calculated using a matrix multiplication algorithm to obtain a coordinate matrix for cross-device content presentation. When smoothing, bilinear interpolation can be used to adjust the coordinate values ​​by weighted averaging of adjacent pixels. For example, when the resolution difference of a device causes jagged edges on the content, the color values ​​of the four adjacent pixels are taken, and new pixel values ​​are calculated based on the interpolation weights to make the transition more natural, which can ensure the clarity of the content on high-resolution devices while avoiding distortion on low-resolution devices.

[0084] For example, the screen size, resolution and pixel density parameters of the source device and the target device are obtained using the standardized formula S x =W / D,S y =H / D, where W represents the screen width in pixels, H represents the screen height in pixels, D represents the pixel density, and S x and S y are the physical inches corresponding to each logical pixel in the horizontal and vertical directions, respectively, and the physical parameter set of the device is obtained. According to the preset logical coordinate system, the linear transformation formula M=(P x / L x , P y / L y ), where P x 、P y Indicates the horizontal and vertical values ​​of the physical coordinates in inches, L x 、L yThe horizontal and vertical values ​​representing the logical coordinates are normalized values ​​(0 to 1), and the mapping ratio M is in inches / logical units, for example, M = (50 inches, 30 inches). The mapping ratio between the physical parameter set and the logical position is calculated to obtain the mapping ratio matrix. If the resolution difference of the mapping ratio matrix exceeds the preset threshold, the mapping ratio matrix is ​​smoothed using the bilinear interpolation formula V = V1*(1-a)*(1-b) + V2*a*(1-b) + V3*(1-a)*b + V4*a*b, where V1, V2, V3, and V4 represent neighboring pixel values ​​and a and b represent interpolation weights, to obtain the smoothed coordinate matrix. Through the smoothed coordinate matrix, the matrix multiplication formula R = C*T is used, where C represents the content logical coordinates and T represents the smoothed coordinate matrix, to calculate the final physical coordinates of the content presentation and obtain the coordinate matrix for cross-device content presentation.

[0085] It's important to note that the coordinate matrix used to present content across multiple devices supports dynamic adjustment to accommodate device changes. For example, if a user switches from a mobile phone to a tablet, the system recalculates the coordinate set based on the new device's screen parameters, ensuring the content layout remains unchanged. This dynamic adaptation enhances a seamless user experience across multiple devices.

[0086] Optionally, the acquisition of physical device parameters can also be combined with the device orientation. In landscape mode, the width and height parameters are interchangeable, and the mapping ratio needs to be recalculated to ensure that the content can be displayed correctly in different usage scenarios.

[0087] As another optional implementation, after obtaining the coordinate matrix, a matrix transformation algorithm is used to calculate the smooth transition parameters corresponding to the coordinate matrix to generate the transformation matrix.

[0088] Exemplarily, the coordinate matrix of each device screen is first obtained, and the matrix transformation algorithm is used to calculate the mapping parameters of the content between different devices to generate an initial transformation matrix to determine the mapping relationship of the content between physically separated devices. According to the initial transformation matrix, the transition parameters are adjusted for the inter-device switching scenario. If the parameter deviation exceeds the preset threshold, the transition parameters are optimized by a linear interpolation algorithm to obtain a smooth transition matrix. The coordinate offset of the content presentation is obtained from the smooth transition matrix, and the display position of the content on the target device is calculated in combination with the device adaptation rules to determine the content consistency. Using the smooth transition matrix and the display position, a convolutional neural network algorithm is used to render the content to generate a logically coherent presentation effect and obtain the final content presentation result.

[0089] Step S500: allocating the preloaded content set to the target device through the transformation matrix, and dynamically adjusting the content presentation order according to the real-time user interaction data to determine the dynamically displayed content.

[0090] In this embodiment, a transformation matrix and a real-time rendering algorithm are used to distribute preloaded content sets to each device screen. The content presentation order is dynamically adjusted based on the user's real-time interaction pattern to produce the final output of the multimedia interactive display. Cross-device adaptation is achieved through the spatial transformation matrix. Combined with real-time rendering and intelligent interaction algorithms, the dynamic display of the multimedia interactive display is made real-time.

[0091] As an optional implementation, based on the preloaded content set, a transformation matrix is ​​used to generate the initial content distribution layout of each device screen, the mapping relationship between the content and the screen is obtained, and a content distribution table for multiple devices is obtained. If the user inputs interactive behavior through the device screen, the user operation is classified according to the content distribution table and the interaction mode, and the multimedia content is generated using a real-time rendering algorithm to determine the visual effect of the content presentation. Through the presentation order rules, combined with the visual effect of the content presentation, the arrangement order of the multimedia content is dynamically adjusted, the updated content presentation sequence is obtained, and the dynamic display output on the screen is obtained. When user interaction continues to occur, a real-time feedback mechanism is used to obtain user operation data from the dynamic display output, update the transformation matrix and the distribution rules, and determine the final output effect of the multimedia display. Among them, the content distribution table provides a stable content mapping basis, real-time rendering and presentation order adjustment enhance the dynamic adaptability of the interaction, and the feedback mechanism further optimizes the effect of long-term use. These solutions jointly support the integrity of multi-device collaborative display.

[0092] For example, the core logic of content distribution can be used to generate the initial content distribution layout for each device screen for a preloaded content set. Preloaded content sets typically contain multimedia resources such as images, videos, or text. These resources need to be appropriately distributed based on the device screen resolution and size. For example, consider three devices with screen resolutions of 1920x1080, 1280x720, and 2560x1440. By parsing the content set's metadata, such as resource dimensions and priority, and combining it with device screen parameters, an initial transformation matrix can be generated. This transformation matrix defines the scaling and position offset of content on different screens. High-priority video content may be assigned to the highest-resolution 2560x1440 screen to ensure optimal visual quality, while text content may be assigned to a 720p screen to conserve resources. This distribution method uses matrix mapping to clearly define the correspondence between content and screens, generating a content distribution table that provides a foundation for subsequent interactions. After a user enters an interactive action on the device screen, the user action is classified based on the content distribution table and interaction mode. Interactions may include tapping, swiping, or zooming. For example, if a user scrolls through video content on a 1280x720 screen, the system identifies the content as a video stream based on the allocation table. Combining the scrolling direction and speed, it classifies it as a "fast forward" operation. The real-time rendering algorithm then intervenes and adjusts the order in which the video frames are loaded based on this classification.

[0093] Optionally, presentation order rules are used to dynamically adjust the order of multimedia content. For example, in a multi-device collaborative presentation, for example, if a user clicks an image on a 1920x1080 screen, the system prioritizes that image based on the rules and adjusts the content sequence so that it is displayed larger and synchronized across all devices. The updated content presentation sequence might place images first, videos second, and text last. This dynamic adjustment generates a presentation output that better aligns with user intent by analyzing user interaction frequency and content type.

[0094] It's important to note that the real-time feedback mechanism operates while user interaction continues, capturing user action data from the dynamic display output and updating the transformation matrix and allocation rules. For example, if a user continuously zooms in and out of an image on a 2560x1440 screen, the system records the zoom ratio and frequency, and updates the zoom parameters in the transformation matrix to ensure that subsequent content loading is more consistent with user habits. The final output quality is determined based on the visual consistency and response speed of the content. If the image displayed on all devices has the same zoom ratio and the response time is less than 200 milliseconds, the output quality is considered satisfactory. This mechanism ensures display coordination across multiple devices by continuously optimizing the matrix and rules.

[0095] In this embodiment, real-time interest distribution points are generated based on user profile features. User behavior is predicted based on historical interaction data, and preloaded content trigger conditions are generated. Furthermore, a preloaded content set is generated based on user role attributes and content type preferences. A transformation matrix from the source device to the target device is generated based on the physical parameter sets of the source and target devices. The preloaded content set is then distributed to each target device using this transformation matrix. Finally, the display order of the preloaded content is dynamically adjusted based on the real-time user interaction data acquired in real time, resulting in the final output of the multimedia interactive display. This achieves seamless content integration when switching between multiple screens. Furthermore, because user interaction behavior is monitored in real time and preloaded content is adjusted based on real-time data, the real-time nature of the content can be ensured.

[0096] Example 2

[0097] Based on the first embodiment, another embodiment of the present application is proposed, referring to Figure 2 After the step of generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference, the following steps are included:

[0098] Step S310: If the number of content types in the preloaded content set is greater than or equal to the preset number of content types, the preloaded content set is prioritized according to the preference weights of the content type preferences;

[0099] Step S320: sorting the interactive media materials in the preloaded content set according to the sorting result and the real-time user interaction data to generate an updated preloaded content set.

[0100] In this embodiment, the priority sorting of preloaded content sets is dynamically optimized by dual driving forces of content type preference weights and real-time user interaction data, which can ensure that media materials with high relevance and high interaction tendencies are presented in an orderly manner and improve the accuracy of content matching.

[0101] As an optional implementation, priority is calculated by preference weights, using the weighted average formula Get the sorted subset of interactive media materials, where W i represents the preference weight of the i-th content type, S i Indicates the semantic relevance of this type of material, Q i Represents the quality correction factor, usually with a default value of 1, which is dynamically adjusted according to the clarity of the material and the interactivity rate of the material. Based on the subset of interactive media materials, a judgment is made on cross-device compatibility. If the format of the interactive content material on the target device is adapted, the material is retained to obtain a preloaded content set after compatibility screening. Among them, preference weight calculation and presentation priority are the core links of the content recommendation system, which aims to sort materials according to user preferences. Preference weights reflect the user's tendency towards different content types, such as video, graphics or audio. The weighted average formula ensures that the priority reflects the user's interests by combining weights and semantic relevance.

[0102] For example, a user's preference weight for short videos is 0.7, for text and images is 0.5, and for audio is 0.3. The semantic relevance of short videos is 0.9, for text and images is 0.6, and for audio is 0.4. The quality correction factor for short videos is 1.0, for text and images is 0.8, and for audio is 1.2. The calculated priority for short videos is 0.6, for text and images is 0.23, and for audio is 0.14. Therefore, the priority ranking is short videos > text and images > audio. In practice, the system can dynamically adjust the weights based on user historical behavior. For example, if a user frequently watches short videos, the weight will automatically increase, thereby improving recommendation accuracy.

[0103] As another optional implementation, after priority sorting, real-time user interaction data is continuously obtained. If a change in the user role is detected, interactive media materials that match the semantic label of the new role are obtained from the preloaded content set, and the cosine similarity is calculated using a vector space model to obtain an updated preloaded content set. Through the updated preloaded content set, a sorting algorithm is used to generate a content candidate list according to presentation priority and semantic relevance, and the final content candidate list is determined. Among them, the application of the vector space model needs to be combined with a semantic label system. For example, the label system includes categories such as "food" and "travel", and each category is refined into specific labels, such as "Sichuan cuisine" and "sea view". The system generates labels for materials and users through natural language processing technology to ensure accurate matching. In practice, a user is labeled "Sichuan cuisine lover", and the system gives priority to recommending Sichuan cuisine-related videos. This refined label system improves the pertinence of recommendations.

[0104] For example, after a user's role changes, material matching relies on a vector space model, which calculates the semantic fit between the new role and the material through cosine similarity. For example, if a user switches from a "tourist" to a "local resident," materials related to local culture, such as local food recommendations, are extracted. The vector space model converts user tags and material tags into vectors and calculates the cosine value of the angle between them. A higher value indicates a higher degree of match. Assuming the cosine similarity between the local food tag vector of a certain material and the new role tag is 0.9, it will be retained first. This method ensures that the content dynamically adapts to changes in user identity.

[0105] Optionally, when generating the candidate content list, the sorting algorithm comprehensively considers priority and semantic relevance. A weighted Top-K algorithm can be used to select the top 10 assets from the preloaded content set. In a real-world scenario, a short video with a priority score of 0.85 and a semantic relevance of 0.9 has the highest overall score and is ranked high. Sorting can also consider diversity, for example, avoiding recommending the same type of content repeatedly. This approach ensures a highly accurate and diverse candidate list.

[0106] Optionally, the ranking algorithm can also incorporate a time decay factor to prioritize the latest content. For example, if a food video was published today, its ranking weight will increase by 0.1, while videos from a week ago will remain unchanged. This approach ensures content freshness and meets users' demand for real-time information.

[0107] As yet another optional implementation, the generation of the preloaded content set relies on a joint screening of compatibility and role matching.

[0108] For example, the system selects 50 compatible assets from 100 preliminary assets, then eliminates 20 irrelevant assets based on the user's role, resulting in a total of 30 preloaded assets. This layered filtering ensures efficient and adaptable content. In practice, if a user is browsing on a tablet, the system prioritizes assets that support landscape display and match their "foodie" tag.

[0109] In this embodiment, when the preloaded content set contains a wide variety of content types, the preloaded content is prioritized based on the user's type preferences, ensuring that highly compatible content is loaded and presented first. Furthermore, the preloaded content set can dynamically change based on user role attributes. When changes in the user role are detected, adjustments can be made based on real-time data, improving the accuracy of content matching.

[0110] Example 3

[0111] In an embodiment of the present application, a dynamic display adjustment device for multimedia interactive presentation is proposed.

[0112] Reference Figure 3 , Figure 3 This is a schematic diagram of the terminal structure of the hardware operating environment involved in an embodiment of the present application.

[0113] like Figure 3 As shown, the control terminal may include: a processor 1001, such as a CPU, a network interface 1003, a memory 1004, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The network interface 1003 may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface). The memory 1004 may be a high-speed RAM memory or a stable memory (non-volatile memory), such as a disk memory. The memory 1004 may also be a storage device independent of the aforementioned processor 1001.

[0114] Those skilled in the art will understand that Figure 3 The terminal structure shown in the figure does not constitute a limitation to the terminal, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0115] like Figure 3 As shown, the memory 1004 as a computer storage medium may include an operating system, a network communication module, and a dynamic display adjustment program for multimedia interactive presentation.

[0116] exist Figure 3In the hardware structure of the dynamic display adjustment device for multimedia interactive presentation shown, the processor 1001 may call the dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004 and perform the following operations:

[0117] Obtaining a user portrait feature set, and generating a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes real-time user interaction data, user role attributes, content type preferences, and historical interaction data;

[0118] Predicting user expected behavior based on the user's real-time interest distribution points and the historical interaction data, and generating a trigger condition for preloading content;

[0119] generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference;

[0120] Obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix;

[0121] The preloaded content set is distributed to the target device through the transformation matrix, and the content presentation order is dynamically adjusted according to the real-time user interaction data to determine the dynamic display content.

[0122] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0123] Calculating the semantic similarity between the user's real-time interest distribution points and the semantic tags in the semantic content library;

[0124] Semantically associating the semantic tags whose semantic similarity is greater than a preset similarity threshold, and prioritizing the semantic tags based on the weights corresponding to the semantic tags to generate a semantic association graph;

[0125] The user's expected behavior is predicted based on the semantic association graph and the historical interaction data, and the preloading content triggering condition is generated.

[0126] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0127] Extracting content semantics and content type features corresponding to the semantic tags in the historical interaction data based on the semantic tags in the semantic association graph;

[0128] Calculating a behavior prediction probability based on the content semantics, the content type features, and the content type preferences to generate the user expected behavior;

[0129] The preloading content triggering condition is generated according to the user real-time interaction data and the user expected behavior.

[0130] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0131] Extracting the interactive media material associated with the preloaded content triggering condition, and generating a preliminary material set matching the interactive media material according to a semantic association rule;

[0132] Matching the preliminary material set with the content type preference to generate a material subset;

[0133] The semantic tag of the user role attribute is obtained, and the interactive media material in the material subset is filtered according to the correlation between the material subset and the semantic tag to generate the preloaded content set.

[0134] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0135] If the number of content types in the preloaded content set is greater than or equal to the preset number of content types, sorting the preloaded content set by type priority according to the preference weights of the content type preferences;

[0136] The interactive media materials in the preloaded content set are sorted according to the sorting result and the real-time user interaction data to generate an updated preloaded content set.

[0137] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0138] Acquire the physical parameter set, and generate a mapping ratio matrix from the source device to the target device according to the physical parameter set;

[0139] If the resolution difference of the mapping ratio matrix is ​​greater than a preset difference threshold, smoothing the mapping ratio matrix to generate the coordinate matrix;

[0140] A matrix transformation algorithm is used to calculate the smooth transition parameters corresponding to the coordinate matrix to generate the transformation matrix.

[0141] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0142] Acquiring the user real-time interaction data and preprocessing the user real-time interaction data to obtain interaction data in a unified format;

[0143] sorting the dwell time, sight direction, and operation frequency in the interaction data using a time series analysis algorithm to generate focus point weights for each content area;

[0144] Clustering the user role attributes and the content type preferences, and adjusting the focus weights;

[0145] The focus point weights and the historical interaction data are integrated to generate the user's real-time interest distribution points.

[0146] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention further provides a dynamic display adjustment system for multimedia interactive presentation, wherein the dynamic display adjustment system for user multimedia interactive presentation comprises:

[0147] User portrait analysis module, used to collect and fuse multi-source data in real time to generate user portrait feature sets;

[0148] A behavior prediction module, configured to generate a preloaded content set based on the user portrait feature set;

[0149] A cross-device content rendering module, configured to generate a transformation matrix based on a set of physical parameters of a source device and a target device;

[0150] The content distribution and dynamic adjustment module is used to distribute the preloaded content set to the target device through the transformation matrix, and dynamically adjust the content presentation order according to the real-time user interaction data to determine the dynamic display content.

[0151] Optionally, the processor 1001 may call a dynamic display adjustment program for multimedia interactive presentation stored in the memory 1004, and further perform the following operations:

[0152] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention also provides a terminal device, including a memory, a processor, and a dynamic display adjustment program for user multimedia interactive display stored on the memory and runnable on the processor. When the processor executes the dynamic display adjustment program for user multimedia interactive display, the dynamic display adjustment method for user multimedia interactive display as described above is implemented.

[0153] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention also provides a computer-readable storage medium, on which a dynamic display adjustment program for user multimedia interactive display is stored. When the dynamic display adjustment program for user multimedia interactive display is executed by a processor, the dynamic display adjustment method for user multimedia interactive display as described above is implemented.

[0154] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0155] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0156] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0157] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0158] It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The present application may be implemented by means of hardware comprising several distinct components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second and third etc. does not indicate any order. These words may be interpreted as names.

[0159] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0160] Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present application fall within the scope of the claims and their equivalents, the present application is intended to include such modifications and variations.

Claims

1. A dynamic display adjustment method for multimedia interactive display, characterized in that: The dynamic display adjustment method for multimedia interactive display includes: Obtaining a user portrait feature set, and generating a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes real-time user interaction data, user role attributes, content type preferences, and historical interaction data; Predicting user expected behavior based on the user's real-time interest distribution points and the historical interaction data, and generating a trigger condition for preloading content; generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference; Obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix; The preloaded content set is distributed to the target device through the transformation matrix, and the content presentation order is dynamically adjusted according to the real-time user interaction data to determine the dynamic display content.

2. The dynamic display adjustment method for multimedia interactive presentation according to claim 1, characterized in that: The step of predicting the user's expected behavior based on the user's real-time interest distribution points and the historical interaction data and generating a trigger condition for preloading content includes: Calculating the semantic similarity between the user's real-time interest distribution points and the semantic tags in the semantic content library; Semantically associating the semantic tags whose semantic similarity is greater than a preset similarity threshold, and prioritizing the semantic tags based on the weights corresponding to the semantic tags to generate a semantic association graph; The user's expected behavior is predicted based on the semantic association graph and the historical interaction data, and the preloading content triggering condition is generated.

3. The dynamic display adjustment method for multimedia interactive presentation according to claim 2, characterized in that: The step of predicting the user's expected behavior based on the semantic association graph and the historical interaction data and generating the trigger condition for preloading content includes: Extracting content semantics and content type features corresponding to the semantic tags in the historical interaction data based on the semantic tags in the semantic association graph; Calculating a behavior prediction probability based on the content semantics, the content type features, and the content type preferences to generate the user expected behavior; The preloading content triggering condition is generated according to the user real-time interaction data and the user expected behavior.

4. The dynamic display adjustment system for multimedia interactive display according to claim 1, characterized in that: The step of generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference includes: Extracting the interactive media material associated with the preloaded content triggering condition, and generating a preliminary material set matching the interactive media material according to a semantic association rule; Matching the preliminary material set with the content type preference to generate a material subset; The semantic tag of the user role attribute is obtained, and the interactive media material in the material subset is filtered according to the correlation between the material subset and the semantic tag to generate the preloaded content set.

5. The dynamic display adjustment system for multimedia interactive display according to claim 1, characterized in that: After the step of generating a preloaded content set according to the preloaded content triggering condition, the user role attribute, and the content type preference, the method includes: If the number of content types in the preloaded content set is greater than or equal to the preset number of content types, sorting the preloaded content set by type priority according to the preference weights of the content type preferences; The interactive media materials in the preloaded content set are sorted according to the sorting result and the real-time user interaction data to generate an updated preloaded content set.

6. The dynamic display adjustment system for multimedia interactive display according to claim 1, characterized in that: The steps of obtaining a set of physical parameters of a source device and a target device, constructing a coordinate matrix for cross-device content presentation corresponding to the preloaded content set, and generating a transformation matrix corresponding to the coordinate matrix include: Acquire the physical parameter set, and generate a mapping ratio matrix from the source device to the target device according to the physical parameter set; If the resolution difference of the mapping ratio matrix is ​​greater than a preset difference threshold, smoothing the mapping ratio matrix to generate the coordinate matrix; A matrix transformation algorithm is used to calculate the smooth transition parameters corresponding to the coordinate matrix to generate the transformation matrix.

7. The dynamic display adjustment system for multimedia interactive display according to claim 1, characterized in that: The steps of obtaining a user portrait feature set and generating a user's real-time interest distribution points based on the user portrait feature set, wherein the user portrait feature set includes user real-time interaction data, user role attributes, content type preferences, and historical interaction data, include: Acquiring the user real-time interaction data and preprocessing the user real-time interaction data to obtain interaction data in a unified format; sorting the dwell time, sight direction, and operation frequency in the interaction data using a time series analysis algorithm to generate focus point weights for each content area; Clustering the user role attributes and the content type preferences, and adjusting the focus weights; The focus point weights and the historical interaction data are integrated to generate the user's real-time interest distribution points.

8. A dynamic display adjustment system for multimedia interactive display, characterized in that: The dynamic display adjustment system for user multimedia interactive display includes: User portrait analysis module, used to collect and fuse multi-source data in real time to generate user portrait feature sets; A behavior prediction module, configured to generate a preloaded content set based on the user portrait feature set; A cross-device content rendering module, configured to generate a transformation matrix based on a set of physical parameters of a source device and a target device; The content distribution and dynamic adjustment module is used to distribute the preloaded content set to the target device through the transformation matrix, and dynamically adjust the content presentation order according to the real-time user interaction data to determine the dynamic display content.

9. A terminal device, characterized in that: It includes a memory, a processor, and a dynamic display adjustment program for multimedia interactive display stored in the memory and runnable on the processor. When the processor executes the dynamic display adjustment program for multimedia interactive display, it implements the method described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a dynamic display adjustment program for multimedia interactive presentation. When the dynamic display adjustment program for multimedia interactive presentation is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images