Video playback method and device

Abstract

An embodiment of the present specification provides a video playback method and device, which includes: first, obtaining the playback behavior data of the visiting user of the live broadcast platform, encoding the playback behavior data to obtain the playback behavior characteristics, and then inputting the playback behavior characteristics into the playback behavior prediction model to perform playback layout prediction and playback content prediction, outputting a first playback probability distribution, and then inputting the first playback probability distribution and the video playback index into the adjustment model to perform probability distribution adjustment, outputting a second playback probability distribution, and finally sending the second playback probability distribution to the user terminal, so as to generate the playback window layout parameters and playback content parameters for video playback in the application program based on the second playback probability distribution at the user terminal, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while reducing the complexity of the visiting user manually configuring the playback window on the live broadcast platform.

Description

Technical Field

[0001] This document relates to the field of video technology, and in particular to a video playback method and device. Background Art

[0002] With the continuous development of Internet technology, live video has become an important way for people to obtain information, entertainment and socialize. As users' demands for video content become increasingly diversified, they not only want to watch high-quality videos, but also want to be able to watch multiple different videos at the same time to get a richer viewing experience.

[0003] At present, many live broadcast platforms provide the function of multi-window video playback. For example, some live broadcast platforms allow users to manually adjust the number and size of playback windows, or play videos based on preset window layout templates. However, fixed window layouts have some shortcomings in user experience. For example, fixed window layouts may not meet the window layout habits of different users when playing videos, resulting in users needing to frequently manually adjust the layout, affecting the viewing experience. Therefore, how to effectively display video content during multi-window video playback and optimize the playback window layout according to the user's personalized needs has become an urgent problem to be solved in the current video technology field. Summary of the invention

[0004] An object of one embodiment of the present specification is to provide a video playback method and device to improve the efficiency of video playback window configuration and the accuracy of video playback content recommendation.

[0005] To solve the above technical problems, an embodiment of this specification is implemented as follows:

[0006] In a first aspect, an embodiment of the present specification provides a video playback method, including:

[0007] Acquire playback behavior data of users accessing the live broadcast platform, and encode the playback behavior data to obtain playback behavior features;

[0008] Inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution;

[0009] Inputting the first playback probability distribution and the video playback index into the adjustment model to adjust the probability distribution, and outputting a second playback probability distribution;

[0010] The second playback probability distribution is sent to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0011] In a second aspect, an embodiment of the present specification provides another video playback method, including:

[0012] Obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application;

[0013] Generate a play window layout parameter and a play content parameter of the application based on the play probability distribution;

[0014] A play window for playing the video in the application is configured according to the play window layout parameters, and the video is played in the play window according to the play content parameters.

[0015] In a third aspect, an embodiment of the present specification provides a video playback device, including:

[0016] A data processing module, which obtains the playback behavior data of the visiting users of the live broadcast platform, and encodes the playback behavior data to obtain playback behavior characteristics;

[0017] A prediction module, inputting the playback behavior features into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution;

[0018] An adjustment module, inputting the first playback probability distribution and the video playback index into an adjustment model to adjust the probability distribution, and outputting a second playback probability distribution;

[0019] The sending module sends the second playback probability distribution to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0020] In a fourth aspect, an embodiment of the present specification provides another video playback device, including:

[0021] A data acquisition module, used to obtain the playback probability distribution of the video played by the accessing user through the application on the live broadcast platform;

[0022] A parameter generation module, used to generate a playback window layout parameter and a playback content parameter of the application based on the playback probability distribution;

[0023] The video playing module is used to configure a playing window for playing videos in the application according to the playing window layout parameters, and to play videos in the playing window according to the playing content parameters.

[0024] In the fifth aspect, another embodiment of the present specification provides a video playback device, comprising: a memory, a processor, and computer executable instructions stored in the memory and executable on the processor, wherein the computer executable instructions, when executed by the processor, implement the steps of the video playback method as described in the first aspect above.

[0025] In the sixth aspect, another embodiment of the present specification provides another video playback device, comprising: a memory, a processor, and computer executable instructions stored in the memory and executable on the processor, wherein the computer executable instructions, when executed by the processor, implement the steps of the video playback method as described in the second aspect above.

[0026] In a seventh aspect, another embodiment of the present specification provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the video playback method described in the first aspect above are implemented.

[0027] In an eighth aspect, another embodiment of the present specification provides another computer-readable storage medium, wherein the computer-readable storage medium is used to store computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the video playback method described in the second aspect above are implemented.

[0028] In a ninth aspect, another embodiment of the present specification provides a computer program product, wherein the computer program product includes a video playback program, and the video playback program is executed by a processor to implement the steps of the video playback method as described in the first aspect above.

[0029] In a tenth aspect, another embodiment of the present specification provides another computer program product, wherein the computer program product includes a video playback program, and the video playback program is executed by a processor to implement the steps of the video playback method described in the second aspect above.

[0030] A video playback method provided in this embodiment first obtains the playback behavior data of the visiting user of the live broadcast platform, and encodes the playback behavior data to obtain the playback behavior characteristics, so as to provide a data basis for the subsequent playback layout prediction and playback content prediction. Secondly, the playback behavior characteristics are input into the playback behavior prediction model for playback layout prediction and playback content prediction, and a first playback probability distribution is output to provide a data basis for the subsequent personalized window layout and playback content recommendation. Then, the first playback probability distribution and the video playback index are input into the adjustment model for probability distribution adjustment, and a second playback probability distribution is output. The first playback probability distribution is optimized and adjusted by combining the video playback index to generate a more accurate playback probability distribution. Finally, the second playback probability distribution is sent to the user terminal. The second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation in the process of video playback on the live broadcast platform, reducing the complexity of the visiting user manually configuring the playback window on the live broadcast platform, and improving the visiting user's experience of watching videos on the live broadcast platform.

[0031] Another video playback method provided by this embodiment first obtains the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then generates the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, thereby improving the visiting user's experience in the video watching process and reducing the complexity of the visiting user's manual configuration of the playback window. Finally, the playback window for video playback in the application is configured according to the playback window layout parameters, and the video is played in the playback window according to the playback content parameters, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while also improving the convenience of the visiting user using multiple windows for video playback on the live broadcast platform and improving the visiting user's experience of watching videos on the live broadcast platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solutions in one or more embodiments of the present specification, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0033] Figure 1 A processing flow chart of a video playback method provided by one or more embodiments of this specification;

[0034] Figure 2A schematic diagram of a model architecture of a playback behavior prediction model provided in one or more embodiments of this specification;

[0035] Figure 3 A schematic diagram of a video playback window layout in a multi-window mode provided by one or more embodiments of this specification;

[0036] Figure 4 A processing flow chart of a video playback method applied to a video playback scenario in a live broadcast platform provided by one or more embodiments of this specification;

[0037] Figure 5 Another video playback method processing flow chart provided for one or more embodiments of this specification;

[0038] Figure 6 A schematic diagram of a video playback device provided for one or more embodiments of this specification;

[0039] Figure 7 A schematic diagram of another video playback device provided for one or more embodiments of this specification;

[0040] Figure 8 A schematic diagram of the structure of a video playback device provided in one or more embodiments of this specification;

[0041] Fig. 9 A schematic diagram of the structure of another video playback device provided in one or more embodiments of this specification. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the following will be combined with the drawings in one or more embodiments of this specification to clearly and completely describe the technical solutions in one or more embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. Based on one or more embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this document.

[0043] This specification provides a video playback method embodiment:

[0044] Reference Figure 1 , which shows a processing flow chart of a video playback method provided in this embodiment. The video playback method provided in this embodiment specifically includes the following steps S102 to S108.

[0045] Step S102, obtaining the playback behavior data of the visiting users of the live broadcast platform, and encoding the playback behavior data to obtain playback behavior features.

[0046] In this embodiment, the playback behavior data refers to the data generated when the accessing user accesses the live broadcast platform through an application to play live video. The playback behavior data may include the user code, timestamp and / or video viewing time of the accessing user. In addition, the playback behavior data may also include the behavior data of the accessing user on the live broadcast platform, such as the behavior data of the accessing user clicking on the video, switching the video, closing the video and / or adjusting the size of the video playback window.

[0047] The playback behavior feature refers to a feature obtained after encoding the playback behavior. The playback behavior feature may be a vector feature and / or a tensor feature. Specifically, the playback behavior feature may be a feature obtained after encoding the user code, timestamp, video viewing time, and / or the behavior data of the visiting user on the live broadcast platform.

[0048] In the specific implementation, the playback behavior data of the visiting users of the live broadcast platform are first obtained, and then the obtained playback behavior data are encoded to obtain the playback behavior characteristics; specifically, in the process of obtaining the playback behavior data of the visiting users of the live broadcast platform, the playback behavior data of the visiting users can be obtained directly from the live broadcast platform, or the playback behavior data of the visiting users submitted by the user terminal can be obtained.

[0049] In actual applications, after obtaining the playback behavior data of the visiting users of the live broadcast platform, the playback behavior data can be preprocessed before encoding the playback behavior data. The preprocessing can be serializing the playback behavior data in chronological order and / or correcting abnormal playback behavior data through data cleaning. After the playback behavior data is preprocessed, the playback behavior data can be encoded.

[0050] For example, the playback behavior data may be encoded using a one-hot encoding method to obtain playback behavior features.

[0051] It should be noted that, considering that the playback behavior data of the accessing user belongs to the privacy of the accessing user to a certain extent, if you want to obtain the playback behavior data of the accessing user, you can obtain the accessing user's authorization before obtaining the playback behavior data. The authorization method can be to send a reminder to the accessing user to obtain the playback behavior data during the current video playback process. After the accessing user confirms the reminder through the instruction, the authorization to obtain the playback behavior data of the accessing user can be obtained. In addition, the accessing user's authorization can also be obtained through other methods, such as obtaining the authorization to obtain the playback behavior data of the accessing user through the signing of an agreement.

[0052] Step S104: input the playback behavior features into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and output a first playback probability distribution.

[0053] The playback behavior prediction model refers to a model that predicts the playback layout and playback content of videos played by visiting users on a live broadcast platform based on playback behavior characteristics. The input of the playback behavior prediction model is the encoded playback behavior characteristics, and the output of the playback behavior prediction model is the playback probability distribution.

[0054] Among them, playback layout prediction refers to predicting the preferences of visiting users for the layout of each video playback window during video playback on the live broadcast platform based on playback behavior characteristics. Specifically, playback layout prediction may include predicting the visiting user's preference for the number of video playback windows, predicting the visiting user's preference for the position of the video playback window and / or predicting the visiting user's preference for the size of the video playback window; playback content prediction refers to predicting the preferences of visiting users for different types of video content during video playback on the live broadcast platform based on playback behavior characteristics and / or predicting the visiting user's preference for different live videos. Specifically, playback content prediction may include predicting the visiting user's preference for each live content type and / or each live video.

[0055] The playback probability distribution refers to the set of probabilities obtained by predicting the playback layout based on the playback behavior characteristics and / or the set of probabilities obtained by predicting the playback content based on the playback behavior characteristics; specifically, the set of probabilities obtained by predicting the playback layout may include accessing the user's preference probability for each window number, the preference probability for each window position and / or the preference probability for each window size; the set of probabilities obtained by predicting the playback content may include accessing the user's preference probability for each live content type and / or the preference probability for each live video; based on this, the playback probability distribution may include accessing the user's preference probability for each window number, the preference probability for each window position, the preference probability for each window size, the preference probability for each live content type and / or the preference probability for each live video.

[0056] Optionally, the playback behavior prediction model includes a feature extraction module, an attention module and a feature conversion module; specifically, the feature extraction module can be used to extract time series features of the playback behavior features, the attention module can be used to perform weight calculations, and the feature conversion module can be used to perform probability distribution conversions.

[0057] For example, the model architecture of the playback behavior prediction model is as follows Figure 2 As shown, based on Figure 2The playback behavior prediction model shown in FIG. 20 inputs the playback behavior feature of the shape (T, D) obtained by encoding processing into the playback behavior prediction model, and firstly extracts the time series feature of the playback behavior feature (T, D) through the LSTM (Long Short-Term Memory) layer (feature extraction module) shown in 201, and outputs the time series feature of the shape (T, H), where T corresponds to the number of time steps and H is the number of hidden units; then, the time series feature (T, H) can be weighted by the attention layer (attention module) shown in 202, and a weighted behavior vector (i.e., weighted feature) of the shape (H,) can be output; finally, the weighted behavior vector is converted into a first playback probability distribution of the shape (K,) through the output layer (feature conversion module) shown in 203 through the softmax function, where the K value is the number of types of behaviors that the accessing user can perform in the multi-window video playback scenario, and the executable behaviors may include the accessing user clicking on the window, switching the live video, closing the window and / or adjusting the window size, and can also be other behaviors performed by the accessing user during video playback, such as hiding the window and reducing the bit rate.

[0058] In specific implementation, in order to predict the preferences of visiting users for playback window layout and playback content when playing videos on the live broadcast platform, the playback behavior characteristics can be input into the playback behavior prediction model to perform playback layout prediction and playback content prediction, and the output is a first playback probability distribution that reflects the visiting user's preferences for video playback window layout and video playback content.

[0059] In the specific implementation process, in order to provide a personalized video playback experience for the visiting user, the playback behavior of the visiting user's video playback can be predicted based on the playback behavior characteristics, and the playback behavior characteristics can be input into the playback behavior prediction model to perform playback layout prediction and playback content prediction to output a first playback probability distribution; specifically, the playback layout prediction and playback content prediction include:

[0060] The feature extraction module extracts the time series features of the playback behavior features to obtain the time series features;

[0061] The weighted features are obtained by assigning weights to the temporal features through the attention module;

[0062] The weighted features are converted into probability distributions through a feature conversion module to obtain a first playback probability distribution.

[0063] For example, the formula for outputting the first playback probability distribution based on the playback behavior prediction model can be expressed as follows:

[0064] Assume that the playback behavior characteristics after the above encoding process can be expressed as , ,…… Indicates that is the vector representation of the tth access user behavior, then the temporal features output by the feature extraction module can be recorded as ;

[0065] The attention module can focus on the LSTM output at each time step Calculate the weight, then the weighted feature obtained by weight calculation of each time series feature can be recorded as A;

[0066] The feature conversion module converts the weighted feature A into a probability distribution through a fully connected layer and a softmax function to obtain the first playback probability distribution R:

[0067]

[0068] in, and are the weight matrix and bias vector of the fully connected layer respectively.

[0069] Step S106: input the first playback probability distribution and the video playback index into an adjustment model to adjust the probability distribution, and output a second playback probability distribution.

[0070] The adjustment model refers to a model that adjusts the playback probability distribution based on the video playback index obtained in real time. The input of the adjustment model is the playback probability distribution and the video playback index, and the output is the adjusted playback probability distribution, that is, the output of the adjustment model is the second playback probability distribution.

[0071] Among them, video playback indicators refer to indicators used to evaluate video playback quality and visiting users' video playback experience. Video playback indicators may include network status indicators, hardware performance indicators and / or visiting user feedback indicators. Specifically, network status indicators may include network transmission rate, network delay and / or packet loss rate, and may also be other indicators reflecting data transmission conditions. Hardware performance indicators may include CPU (Central Processing Unit) occupancy rate and / or memory occupancy rate, and may also be other indicators reflecting resource usage. Visiting user feedback indicators may include visiting user satisfaction and / or visiting user suggestions.

[0072] In specific implementation, after obtaining the first playback probability distribution, the first playback probability distribution and the video playback index can be input into the adjustment model, and the first playback probability distribution is further adjusted based on the video playback index, and the adjusted playback probability distribution is output, that is: the first playback probability distribution is adjusted based on the video playback index, and the second playback probability distribution is output.

[0073] During the specific implementation process, in order to improve the accuracy of the subsequent configuration of the playback window layout and recommended playback content on the live broadcast platform and determine the playback window layout and playback content that are more suitable for the current video playback scene, the first playback probability distribution and the video playback index can be input into the adjustment model to adjust the probability distribution and output the second playback probability distribution; specifically, the probability distribution adjustment includes:

[0074] The basic weight of the video playback index and the index influencing factor are input into the weight adjustment algorithm for calculation to obtain the target weight;

[0075] The first playback probability distribution is adjusted according to the target weight to obtain a second playback probability distribution.

[0076] The basic weight refers to the weight pre-set for each video playback indicator, and the basic weight can be obtained by evaluating each video playback indicator without additional real-time data; the indicator influencing factor is a numerical value calculated based on the video playback indicator obtained in real time, and is used to measure the change in the importance of a certain video playback indicator at a specific moment.

[0077] Specifically, the basic weight of the video playback index and the index influencing factor are input into the weight adjustment algorithm for calculation to obtain the target weight, including:

[0078] Calculate the target basic weight based on the basic weight, indicator impact factor and adjustment coefficient;

[0079] The basic weights are summed up to obtain the total basic weights, and the target weight is obtained by calculating the ratio of the target basic weight to the total basic weights.

[0080] For example, for n video playback indicators, each video playback indicator has a basic weight pre-set according to the actual situation of different dimensions. is the basic weight of the i-th video playback index, where i=1, 2...n; the index impact factor can be calculated based on the real-time acquired video playback index , It can represent the influence of real-time playback behavior data on the i-th video playback index. The target weight calculated by the linear combination method can be expressed as follows:

[0081]

[0082] in, is the basic weight of the i-th video playback index, is the adjustment factor, is the index influencing factor of the i-th video playback index.

[0083] Step S108: sending the second playback probability distribution to the user terminal.

[0084] In a specific implementation, after obtaining the second playback probability distribution adjusted by the adjustment model, the second playback probability distribution can be sent to the user terminal, and after the second playback probability is sent, the user terminal can obtain the sent second playback probability distribution. Optionally, the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0085] Specifically, after the user terminal obtains the second playback probability distribution sent, it can first generate the playback window layout parameters and playback content parameters of the application based on the second playback probability distribution, and then configure the playback window for video playback in the application according to the playback window layout parameters, and play the video in the playback window according to the playback content parameters.

[0086] In a specific implementation process, in the process of generating the playback window layout parameters and playback content parameters of the application based on the second playback probability distribution, the user terminal generates the playback window layout parameters and playback content parameters of the application based on the second playback probability distribution, including:

[0087] Determining a playback mode for playing the video in the application based on the second playback probability distribution;

[0088] If the playback mode is a multi-window mode, determine the window layout parameters and playback content parameters of each playback window;

[0089] If the playback mode is single-window mode, determine the window layout parameters and playback content parameters of the playback window.

[0090] Optionally, the playback window layout parameters may include the number of playback windows, the playback window size and / or the playback window position.

[0091] Specifically, if the second playback probability distribution indicates that the accessing user has a higher probability of watching multiple videos at the same time, the playback mode is determined to be multi-window mode. In the multi-window mode, the window layout parameters and playback content parameters of each playback window can be further determined; if the second playback probability distribution indicates that the accessing user prefers to watch a single video, the playback mode is determined to be single-window mode. In the single-window mode, the window layout parameters of the playback window can be further determined, and a specific video content can be selected as the main playback content. Based on this, the playback content parameters of the playback window can be determined.

[0092] In actual applications, in order to meet the diverse needs of visiting users for video playback on the live broadcast platform and improve the user experience of visiting users, visiting users can also adjust the playback window size, hide or display specific playback windows and / or drag the playback window in multi-window mode or single-window mode during video playback on the live broadcast platform based on the user terminal.

[0093] In addition, the behavior data of the visiting user adjusting the playback window size, hiding or displaying a specific playback window and / or dragging the playback window during video playback can be obtained. After encoding the obtained behavior data, it can be input into the above-mentioned playback behavior prediction model to fine-tune the playback behavior prediction model and continuously optimize the accuracy of the playback behavior prediction model in predicting the playback behavior preferred by the visiting user.

[0094] Furthermore, in the case where the playback mode is determined to be a multi-window mode, in order to enable the live broadcast content preferred by the accessing user to be displayed preferentially and easily noticed by the accessing user, thereby improving the accessing user's satisfaction with video playback on the live broadcast platform, if the playback mode is a multi-window mode, the playback content parameters of each playback window are determined, including:

[0095] Sorting each playback content in descending order according to the playback probability of each playback content in the playback probability distribution to obtain a ranking of each playback content;

[0096] The playback content parameters of each playback window are determined according to the ranking.

[0097] Specifically, each playback content is sorted in descending order according to the playback probability of each playback content in the playback probability distribution. After obtaining the ranking position of each playback content, the preference of the visiting user for each playback content can be determined. For playback content with a high playback probability, it can be allocated to a larger playback window for display, and / or, allocated to a playback window at the center of the screen or at a position that the visiting user is accustomed to for display; for playback content with a low playback probability, it can be allocated to a smaller playback window for display, and / or, allocated to a playback window at the edge of the screen for display.

[0098] For example, the obtained playback probability distribution may be: P (single window) = 15%, P (double windows) = 10%, P (triple windows) = 20%, P (quadruple windows) = 40%, P (five windows) = 10%, P (six windows) = 5%. Based on this, the playback mode may be determined to be the multi-window mode, and the playback window layout may be further determined to be Figure 3 The four windows shown in the figure are: P1 (sports) = 35%, P2 (games) = 25%, P3 (education) = 5%, P4 (technology) = 20%, and P5 (entertainment) = 15%. After sorting the playback probabilities of the various playback contents, we can obtain P1>P2>P4>P5>P3. Based on this, Figure 3 Among the four playback windows 301, 302, 303, and 304 shown, playback window 301 can play sports videos, playback window 302 can play game videos, playback window 303 can play science and technology videos, and playback window 304 can play entertainment videos. Science and technology videos may not be played in the playback window temporarily.

[0099] In actual applications, based on the video playback in the above-mentioned multi-window mode, in order to optimize the effect of video playback on the user terminal and provide a better video playback experience for visiting users, adaptation can be performed according to the different operating systems, browsers and device terminals used by visiting users; for example, CSS (Cascading Style Sheets) media queries can be used to obtain the resolutions and page sizes of different clients, and Flexbox layout (FlexibleBox Layout) or Grid layout (CSS Grid Layout) can be used to ensure that the layout can automatically adapt to different screen sizes and resolutions, such as the screen size and resolution of mobile phones, tablets and / or computers; for another example, interface elements can be modularized, and multi-window components can be encapsulated using Web component technology to ensure the consistent performance of multi-window components on different browsers and live broadcast platforms; for another example, an automated test suite can be established to adapt to mainstream operating systems, browsers and device terminal types to ensure that the functions and performance after each update meet expectations.

[0100] In summary, the video playback method provided in this embodiment first obtains the playback behavior data of the visiting user of the live broadcast platform, and encodes the playback behavior data to obtain the playback behavior characteristics, so as to provide a data basis for the subsequent playback layout prediction and playback content prediction. Secondly, the playback behavior characteristics are input into the playback behavior prediction model for playback layout prediction and playback content prediction, and a first playback probability distribution is output to provide a data basis for the subsequent personalized window layout and playback content recommendation. Then, the first playback probability distribution and the video playback index are input into the adjustment model for probability distribution adjustment, and a second playback probability distribution is output. The first playback probability distribution is optimized and adjusted by combining the video playback index to generate a more accurate playback probability distribution. Finally, the second playback probability distribution is sent to the user terminal. The second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation during video playback on the live broadcast platform, reducing the complexity of the visiting user manually configuring the playback window on the live broadcast platform, and improving the visiting user's experience of watching videos on the live broadcast platform.

[0101] The following combination Figure 4 , taking the application of the video playback method provided by this embodiment in the video playback scenario in the live broadcast platform as an example, the video playback method provided by this embodiment is further described. Figure 4 The video playback method applied to the video playback scenario in the live broadcast platform specifically includes steps S402 to S430.

[0102] Step S402, obtaining the playback behavior data of the visiting users of the live broadcast platform.

[0103] Step S404: Encode the playback behavior data to obtain playback behavior features.

[0104] Step S406: Input the playback behavior features into the playback behavior prediction model to perform playback layout prediction and playback content prediction, and output a first playback probability distribution.

[0105] Step S408: input the first playback probability distribution and the video playback index into the adjustment model to adjust the probability distribution, and output a second playback probability distribution.

[0106] Step S410: sending the second playback probability distribution to the user terminal.

[0107] It should be noted that any one of steps S402 to S410 or any combination of multiple steps can be selected from steps S102 to S108 to form a new implementation method according to the needs of implementation deployment; and any one or multiple technical features in the technical scheme composed of steps S402 to S410 can be selected from steps S102 to S108 to form a new implementation method according to the needs of actual deployment, or the technical features in one or more optional implementation methods provided by steps S102 to S108 can be selected to form a new implementation method, which will not be described one by one here.

[0108] Another video playback method embodiment provided in this specification:

[0109] Reference Figure 5 , which shows a processing flow chart of another video playback method provided by this embodiment. The video playback method provided by this embodiment specifically includes the following steps S502 to S506.

[0110] Step S502, obtaining the playback probability distribution of the video played by the accessing user on the live broadcast platform through the application.

[0111] The playback probability distribution in this embodiment refers to a set of probabilities after probability distribution adjustment for a set of probabilities obtained by predicting the playback layout based on playback behavior characteristics and / or a set of probabilities obtained by predicting the playback content based on playback behavior characteristics; specifically, the set of probabilities obtained by predicting the playback layout may include accessing the user's preference probability for each window number, the preference probability for each window position, and / or the preference probability for each window size; the set of probabilities obtained by predicting the playback content may include accessing the user's preference probability for each live video and / or the preference probability for each live content type; based on this, the playback probability distribution may include accessing the user's preference probability for each window number, the preference probability for each window position, the preference probability for each window size, the preference probability for each live video, and / or the preference probability for each live content type.

[0112] Here, the obtained playback probability distribution may be the second playback probability distribution obtained after the above-mentioned data processing of the playback behavior data of the visiting users of the live broadcast platform; specifically, the process of performing data processing on the playback behavior data to obtain the second playback probability distribution may refer to the above-mentioned embodiment, which will not be repeated in this embodiment.

[0113] In specific implementation, the server processes the playback behavior data of the users visiting the live broadcast platform to obtain the playback probability distribution. After sending the playback probability distribution, the playback probability distribution of the videos played by the visiting users on the live broadcast platform through the application sent by the server can be obtained.

[0114] Optionally, the playback probability distribution is obtained by inputting the first playback probability distribution and the video playback index into an adjustment model to perform probability distribution adjustment; the first playback probability distribution is obtained by inputting playback behavior characteristics into a playback behavior prediction model to perform playback layout prediction and playback content prediction; the playback behavior characteristics are obtained by encoding the playback behavior data of the access users of the live broadcast platform.

[0115] Among them, the playback behavior prediction model refers to a model that predicts the playback layout and playback content for the video playback of visiting users on the live broadcast platform based on the playback behavior characteristics. The input of the playback behavior prediction model is the encoded playback behavior characteristics, and the output of the playback behavior prediction model is the playback probability distribution.

[0116] The playback behavior feature refers to the feature obtained after encoding the playback behavior data. The playback behavior feature can be a vector feature and / or a tensor feature. Specifically, the playback behavior feature can be a feature obtained after encoding the user code, timestamp, video viewing time and / or the behavior data of the visiting user on the live broadcast platform.

[0117] Playback behavior data refers to the data generated when the accessing user accesses the live broadcast platform through an application to play live videos. The playback behavior data may include the user code, timestamp and / or video viewing duration of the accessing user. In addition, the playback behavior data may also include the behavior data of the accessing user on the live broadcast platform, such as the behavior data of the accessing user clicking on videos, switching videos, closing videos and / or adjusting the size of the video playback window.

[0118] Playback layout prediction refers to predicting the preferences of visiting users for the layout of each video playback window during video playback on the live broadcast platform based on playback behavior characteristics. Specifically, playback layout prediction may include predicting the preferences of visiting users for the number of video playback windows, predicting the preferences of visiting users for the positions of video playback windows, and / or predicting the preferences of visiting users for the size of video playback windows. Playback content prediction refers to predicting the preferences of visiting users for different types of video content during video playback on the live broadcast platform based on playback behavior characteristics. Specifically, playback content prediction may include predicting the preferences of visiting users for each live video and / or predicting the preferences of visiting users for each type of live content.

[0119] Optionally, the playback behavior prediction model includes a feature extraction module, an attention module and a feature conversion module; specifically, the feature extraction module can be used to extract time series features of the playback behavior features, the attention module can be used to perform weight calculations, and the feature conversion module can be used to perform probability distribution conversions.

[0120] The adjustment model refers to a model that adjusts the playback probability distribution based on the video playback indicators obtained in real time. The input of the adjustment model is the playback probability distribution and the video playback indicators, and the output is the adjusted playback probability distribution, that is, the output of the adjustment model is the second playback probability distribution.

[0121] Video playback indicators refer to indicators used to evaluate video playback quality and visiting users' video playback experience. Video playback indicators may include network status indicators, hardware performance indicators and / or visiting user feedback indicators. Specifically, network status indicators may include network transmission rate, network delay and / or packet loss rate, hardware performance indicators may include CPU (Central Processing Unit) occupancy rate and / or memory occupancy rate, and visiting user feedback indicators may include visiting user satisfaction and / or visiting user suggestions.

[0122] It should be noted that, considering that the playback behavior data of the accessing user belongs to the privacy of the accessing user to a certain extent, if you want to obtain the playback behavior data of the accessing user, you can obtain the accessing user's authorization before obtaining the playback behavior data. The authorization method can be to send a reminder to the accessing user to obtain the playback behavior data during the current video playback process. After the accessing user confirms the reminder through the instruction, the authorization to obtain the playback behavior data of the accessing user can be obtained. In addition, the accessing user's authorization can also be obtained through other methods, such as obtaining the authorization to obtain the playback behavior data of the accessing user through the signing of an agreement.

[0123] Step S504: generating a play window layout parameter and a play content parameter of the application based on the play probability distribution.

[0124] In specific implementation, in order to improve the efficiency of video playback window configuration and the accuracy of playback content recommendation, and reduce the complexity of manually configuring the playback window when visiting users play videos on the live broadcast platform, the playback window layout parameters and playback content parameters can be generated according to the obtained playback probability distribution, so as to provide a data basis for subsequent configuration of the playback window and video playback in the application.

[0125] In the specific implementation process, in order to improve the experience of visiting users in playing videos on the live broadcast platform and reduce the complexity of visiting users manually configuring the playback window, the playback window layout parameters and playback content parameters can be generated based on the playback probability distribution; the playback window layout parameters and playback content parameters of the application are generated based on the playback probability distribution, including:

[0126] Determine a playback mode for playing the video in the application based on the playback probability distribution;

[0127] If the playback mode is a multi-window mode, determine the window layout parameters and playback content parameters of each playback window;

[0128] If the playback mode is single-window mode, determine the window layout parameters and playback content parameters of the playback window.

[0129] Optionally, the playback window layout parameters may include the number of playback windows, the playback window size and / or the playback window position.

[0130] Specifically, if the playback probability distribution indicates that the accessing user has a higher probability of watching multiple videos at the same time, the playback mode is determined to be multi-window mode. In the multi-window mode, the window layout parameters and playback content parameters of each playback window are further determined; if the playback probability distribution indicates that the accessing user prefers to watch a single video, the playback mode is determined to be single-window mode. In the single-window mode, the window layout parameters of the playback window can be further determined, and the playback content parameters of the playback window can be determined based on the selected specific video content.

[0131] In addition, in order to meet the diverse needs of visiting users for video playback on the live broadcast platform and improve the user experience of visiting users, visiting users can also adjust the playback window size, hide or display specific playback windows and / or drag the playback window in multi-window mode or single-window mode during video playback on the live broadcast platform based on the user terminal.

[0132] Furthermore, in the case where the playback mode is determined to be a multi-window mode, in order to enable the live broadcast content preferred by the accessing user to be displayed preferentially and easily noticed by the accessing user, thereby improving the accessing user's satisfaction with video playback on the live broadcast platform, if the playback mode is a multi-window mode, the playback content parameters of each playback window are determined, including:

[0133] Sorting each playback content in descending order according to the playback probability of each playback content in the playback probability distribution to obtain a ranking of each playback content;

[0134] The playback content parameters of each playback window are determined according to the ranking.

[0135] Specifically, each playback content is sorted in descending order according to the playback probability of each playback content in the playback probability distribution. After obtaining the ranking position of each playback content, the preference of the visiting user for each playback content can be determined. For playback content with a high playback probability, it can be allocated to a larger playback window for display, and / or, allocated to a playback window at the center of the screen or a position that the visiting user is accustomed to for display; for playback content with a low playback probability, it can be allocated to a smaller playback window for display, and / or, allocated to a playback window at the edge of the screen for display. After determining the playback content assigned to each playback window, the playback content parameters of each playback window can be determined based on each playback content.

[0136] For example, after determining the playback content assigned to each playback window, the live video identification, playback content type parameters, playback content label parameters, live video display priority parameters and / or video quality parameters of each playback content can be obtained. In addition, other attribute parameters related to the playback content can also be obtained, and the playback content parameters of each playback window can be determined based on the obtained live video identification, type parameters, label parameters, live video display priority parameters and / or video quality parameters of each playback content.

[0137] Step S506: configuring a play window for playing the video in the application according to the play window layout parameters, and playing the video in the play window according to the play content parameters.

[0138] During specific implementation, the playback mode is determined according to the playback probability distribution, and after further generating the playback window layout parameters and playback content parameters of the application, the playback window for video playback in the application can be configured according to the playback window layout parameters, and the video can be played in the playback window according to the playback content parameters, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendations, while also improving the convenience for visiting users to use multiple windows for video playback on the live broadcast platform, and improving the experience of visiting users watching videos on the live broadcast platform.

[0139] In actual applications, based on the video playback in the above-mentioned multi-window mode, in order to optimize the effect of video playback on the user terminal and provide a better video playback experience for visiting users, adaptation can be performed according to the different operating systems, browsers and device terminals used by visiting users; for example, CSS (Cascading Style Sheets) media queries can be used to obtain the resolutions and page sizes of different clients, and Flexbox layout (FlexibleBox Layout) or Grid layout (CSS Grid Layout) can be used to ensure that the layout can automatically adapt to different screen sizes and resolutions, such as the screen size and resolution of mobile phones, tablets and / or computers; for another example, interface elements can be modularized, and multi-window components can be encapsulated using Web component technology to ensure the consistent performance of multi-window components on different browsers and live broadcast platforms; for another example, an automated test suite can be established to adapt to mainstream operating systems, browsers and device terminal types to ensure that the functions and performance after each update meet expectations.

[0140] To summarize, another video playback method provided in this embodiment first obtains the playback probability distribution of the video played by the visiting user through the application on the live broadcast platform to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then generates the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, thereby improving the visiting user's experience in the video watching process and reducing the complexity of the visiting user's manual configuration of the playback window. Finally, the playback window for video playback in the application is configured according to the playback window layout parameters, and the video is played in the playback window according to the playback content parameters, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while also improving the convenience of the visiting user using multiple windows for video playback on the live broadcast platform and improving the visiting user's experience of watching videos on the live broadcast platform.

[0141] The following combination Figure 4 , taking the application of the video playback method provided by this embodiment in the video playback scenario in the live broadcast platform as an example, the video playback method provided by this embodiment is further described. Figure 4 The video playback method applied to the video playback scenario in the live broadcast platform specifically includes steps S402 to S430.

[0142] Step S412, obtaining the playback probability distribution of the video played by the accessing user on the live broadcast platform through the application.

[0143] Step S414: determining a playback mode for playing the video in the application based on the playback probability distribution.

[0144] Step S416: When the playback mode is the single-window mode, determine the window layout parameters and playback content parameters of the playback window.

[0145] Step S418: configuring a play window for playing the video in the application according to the play window layout parameters.

[0146] Step S420: Play the video in the play window according to the play content parameters.

[0147] In addition, after executing step S414 and determining the playback mode of the video playback in the application based on the playback probability distribution, if the playback mode is the multi-window mode, the following steps S422 to S430 may be executed.

[0148] Step S422: when the playback mode is a multi-window mode, determine the window layout parameters of each playback window.

[0149] Step S424: sort the playback contents in descending order according to the playback probability of each playback content in the playback probability distribution to obtain the ranking of each playback content.

[0150] Step S426, determining the playback content parameters of each playback window according to the ranking.

[0151] Step S428: multiple playback windows for video playback are configured in the application according to the layout parameters of the playback windows.

[0152] Step S430: Play the video in each play window according to the play content parameters of each play window.

[0153] It should be noted that after executing step S414, steps S416 to S420 or steps S422 to S430 can be executed. Steps S416 to S420 and steps S422 to S430 are two implementation processes of generating the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, and configuring the playback window for video playback in the application according to the playback window layout parameters, and performing video playback in the playback window according to the playback content parameters. Steps S416 to S420 are the process of determining the window layout parameters and playback content parameters of the playback window in single-window mode, and performing video playback based on the window layout parameters and playback content parameters. Steps S422 to S430 are the process of determining the window layout parameters and playback content parameters of the playback window in multi-window mode, and performing video playback based on the window layout parameters and playback content parameters.

[0154] It should be noted that any one of steps S412 to S420 or any combination of multiple steps can be combined with any one of steps S502 to S506 to form a new implementation method according to the needs of implementation deployment; and any one or multiple technical features in the technical scheme composed of steps S412 to S420 can be combined with any one or multiple technical features in the technical scheme composed of steps S502 to S506 to form a new implementation method according to the needs of actual deployment, or the technical features in one or more optional implementation methods provided by steps S502 to S506 can be combined with a new implementation method, which will not be elaborated here one by one.

[0155] It should be noted that any one or any combination of steps S412 to S414 and steps S422 to S430 can select any one or any combination of steps S502 to S506 to form a new implementation method according to the needs of implementation deployment; and any one or any combination of technical features in the technical scheme composed of steps S412 to S414 and steps S422 to S430 can also select any one or more technical features in the technical scheme composed of steps S502 to S506 to form a new implementation method according to the needs of actual deployment, or select technical features in one or more optional implementation methods provided by steps S502 to S506 to form a new implementation method, which will not be repeated here.

[0156] Figure 6 A schematic diagram of a video playback device provided by an embodiment of the present invention is shown in FIG. Figure 6 As shown, the device comprises:

[0157] The data processing module 602 obtains the playback behavior data of the visiting user of the live broadcast platform, and encodes the playback behavior data to obtain playback behavior characteristics;

[0158] Prediction module 604, inputs the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputs a first playback probability distribution;

[0159] An adjustment module 606 inputs the first playback probability distribution and the video playback index into an adjustment model to adjust the probability distribution and output a second playback probability distribution;

[0160] The sending module 608 sends the second playback probability distribution to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0161] A video playback device provided in this embodiment first runs a data processing module 602 to obtain playback behavior data of a visiting user of a live broadcast platform, and encodes and processes the playback behavior data to obtain playback behavior features, so as to provide a data basis for subsequent playback layout prediction and playback content prediction. Then, a prediction module 604 is run to input the playback behavior features into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and output a first playback probability distribution, so as to provide a data basis for subsequent personalized window layout and playback content recommendation. Then, an adjustment module 606 is run to input the first playback probability distribution and the video playback index into the adjustment model to perform probability distribution adjustment, and output a second playback probability distribution. The first playback probability distribution is optimized and adjusted by combining the video playback index to generate a more accurate playback probability distribution. Finally, a sending module 608 is run to send the second playback probability distribution to a user terminal. The second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in an application program, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation during video playback on the live broadcast platform, reducing the complexity of manually configuring the playback window for visiting users on the live broadcast platform, and improving the experience of visiting users in watching videos on the live broadcast platform.

[0162] A video playback device provided in one embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0163] Figure 7 A schematic diagram of another video playback device provided by an embodiment of the present invention is shown in FIG. Figure 7 As shown, the device comprises:

[0164] The data acquisition module 702 is used to obtain the playback probability distribution of the video played by the accessing user through the application on the live broadcast platform;

[0165] A parameter generation module 704, configured to generate a playback window layout parameter and a playback content parameter of the application based on the playback probability distribution;

[0166] The video playing module 706 is used to configure a playing window for playing videos in the application according to the playing window layout parameters, and to play videos in the playing window according to the playing content parameters.

[0167] Another video playback device provided in this embodiment first runs the data acquisition module 702 to obtain the playback probability distribution of the video played by the visiting user through the application on the live broadcast platform, so as to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then runs the parameter generation module 704 to generate the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, thereby improving the user experience in the video watching process and reducing the complexity of the user manually configuring the playback window. Finally, the video playback module 706 is run to configure the playback window for video playback in the application according to the playback window layout parameters, and play the video in the playback window according to the playback content parameters, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while also improving the convenience of the visiting user using multiple windows for video playback on the live broadcast platform and improving the visiting user's experience of watching videos on the live broadcast platform.

[0168] Another video playback device provided in one embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0169] Furthermore, an embodiment of the present specification also provides a video playback device, Figure 8 A schematic diagram of the structure of a video playback device provided in one embodiment of this specification is shown in FIG. Figure 8 As shown, the device includes: a memory 801, a processor 802, a bus 803 and a communication interface 804. The memory 801, the processor 802 and the communication interface 804 communicate through the bus 803, and the communication interface 804 may include an input and output interface, which includes but is not limited to a keyboard, a mouse, a display, a microphone, a loudspeaker, etc.

[0170] Figure 8 In the embodiment, the memory 801 stores computer executable instructions that can be run on the processor 802, and when the computer executable instructions are executed by the processor 802, the following process is implemented:

[0171] Acquire playback behavior data of users accessing the live broadcast platform, and encode the playback behavior data to obtain playback behavior features;

[0172] Inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution;

[0173] Inputting the first playback probability distribution and the video playback index into the adjustment model to adjust the probability distribution, and outputting a second playback probability distribution;

[0174] The second playback probability distribution is sent to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0175] A video playback device provided in this embodiment first obtains the playback behavior data of the visiting user of the live broadcast platform through the cooperation of the memory 801, the processor 802, the bus 803 and the communication interface 804, and encodes the playback behavior data to obtain the playback behavior characteristics, so as to provide a data basis for the subsequent playback layout prediction and playback content prediction. Secondly, the playback behavior characteristics are input into the playback behavior prediction model for playback layout prediction and playback content prediction, and a first playback probability distribution is output to provide a data basis for the subsequent personalized window layout and playback content recommendation. Then, the first playback probability distribution and the video playback index are input into the adjustment model for probability distribution adjustment, and a second playback probability distribution is output. The first playback probability distribution is optimized and adjusted by combining the video playback index to generate a more accurate playback probability distribution. Finally, the second playback probability distribution is sent to the user terminal. The second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application program, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation in the process of video playback on the live broadcast platform, reducing the complexity of the visiting user manually configuring the playback window on the live broadcast platform, and improving the visiting user's experience of watching videos on the live broadcast platform.

[0176] The video playback device provided in an embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0177] Furthermore, an embodiment of the present specification also provides another video playback device, Fig. 9 A schematic diagram of the structure of another video playback device provided in one embodiment of this specification, such as Fig. 9 As shown, the device includes: a memory 901, a processor 902, a bus 903 and a communication interface 904. The memory 901, the processor 902 and the communication interface 904 communicate through the bus 903, and the communication interface 904 may include an input and output interface, which includes but is not limited to a keyboard, a mouse, a display, a microphone, a loudspeaker, etc.

[0178] Fig. 9 In the embodiment, the memory 901 stores computer executable instructions that can be run on the processor 902. When the computer executable instructions are executed by the processor 902, the following process is implemented:

[0179] Obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application;

[0180] Generate a play window layout parameter and a play content parameter of the application based on the play probability distribution;

[0181] A play window for playing the video in the application is configured according to the play window layout parameters, and the video is played in the play window according to the play content parameters.

[0182] Another video playback device provided in this embodiment, through the cooperation of memory 901, processor 902, bus 903 and communication interface 904, first obtains the playback probability distribution of the video played by the visiting user through the application on the live broadcast platform, so as to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then generates the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, so as to improve the user experience in the video watching process and reduce the complexity of the user manually configuring the playback window. Finally, the playback window for video playback in the application is configured according to the playback window layout parameters, and the video is played in the playback window according to the playback content parameters, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation, and also improving the convenience of the visiting user to use multiple windows for video playback on the live broadcast platform, and improving the visiting user's experience of watching videos on the live broadcast platform.

[0183] The video playback device provided in an embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0184] Furthermore, another embodiment of the present specification provides a computer-readable storage medium, wherein the computer-readable storage medium is used to store computer-executable instructions, and when the computer-executable instructions are executed by a processor, the following process is implemented:

[0185] Acquire playback behavior data of users accessing the live broadcast platform, and encode the playback behavior data to obtain playback behavior features;

[0186] Inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution;

[0187] Inputting the first playback probability distribution and the video playback index into the adjustment model to adjust the probability distribution, and outputting a second playback probability distribution;

[0188] The second playback probability distribution is sent to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0189] The computer-readable storage medium provided in this embodiment first obtains the playback behavior data of the visiting users of the live broadcast platform, and encodes the playback behavior data to obtain playback behavior characteristics, so as to provide a data basis for subsequent playback layout prediction and playback content prediction. Secondly, the playback behavior characteristics are input into the playback behavior prediction model for playback layout prediction and playback content prediction, and a first playback probability distribution is output to provide a data basis for subsequent personalized window layout and playback content recommendation. Then, the first playback probability distribution and the video playback index are input into the adjustment model for probability distribution adjustment, and a second playback probability distribution is output. The first playback probability distribution is optimized and adjusted by combining the video playback index to generate a more accurate playback probability distribution. Finally, the second playback probability distribution is sent to the user terminal. The second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation during video playback on the live broadcast platform, reducing the complexity of manual configuration of playback windows by visiting users on the live broadcast platform, and improving the experience of visiting users in watching videos on the live broadcast platform.

[0190] The computer-readable storage medium includes a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0191] The computer-readable storage medium provided in an embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0192] Furthermore, another embodiment of the present specification provides another computer-readable storage medium, wherein the computer-readable storage medium is used to store computer-executable instructions, and when the computer-executable instructions are executed by a processor, the following process is implemented:

[0193] Obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application;

[0194] Generate a play window layout parameter and a play content parameter of the application based on the play probability distribution;

[0195] A play window for playing the video in the application is configured according to the play window layout parameters, and the video is played in the play window according to the play content parameters.

[0196] Another computer-readable storage medium provided in this embodiment first obtains the playback probability distribution of the video played by the visiting user through the application on the live broadcast platform to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then generates the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, so as to improve the user experience in the video watching process and reduce the complexity of the user manually configuring the playback window. Finally, the playback window for video playback in the application is configured according to the playback window layout parameters, and the video is played in the playback window according to the playback content parameters, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while also improving the convenience of the visiting user using multiple windows for video playback on the live broadcast platform and improving the visiting user's experience of watching videos on the live broadcast platform.

[0197] The computer-readable storage medium includes a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0198] The computer-readable storage medium provided in an embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0199] Furthermore, another embodiment of the present specification also provides a computer program product, which implements the following process when executed by a processor:

[0200] Acquire playback behavior data of users accessing the live broadcast platform, and encode the playback behavior data to obtain playback behavior features;

[0201] Inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution;

[0202] Inputting the first playback probability distribution and the video playback index into the adjustment model to adjust the probability distribution, and outputting a second playback probability distribution;

[0203] The second playback probability distribution is sent to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters for video playback in the application.

[0204] The computer program product provided in this embodiment first obtains the playback behavior data of the visiting user of the live broadcast platform, and encodes the playback behavior data to obtain the playback behavior characteristics, so as to provide a data basis for the subsequent playback layout prediction and playback content prediction, and then inputs the playback behavior characteristics into the playback behavior prediction model for playback layout prediction and playback content prediction, and outputs the first playback probability distribution, so as to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then inputs the first playback probability distribution and the video playback index into the adjustment model for probability distribution adjustment, and outputs the second playback probability distribution, and optimizes and adjusts the first playback probability distribution by combining the video playback index to generate a more accurate playback probability distribution, and finally sends the second playback probability distribution to the user terminal, and the second playback probability distribution is used to generate the playback window layout parameters and playback content parameters for video playback in the application program, thereby improving the efficiency of video playback window configuration and the accuracy of playback content recommendation in the process of video playback on the live broadcast platform, reducing the complexity of the visiting user manually configuring the playback window on the live broadcast platform, and improving the visiting user's experience of watching videos on the live broadcast platform. The computer program product provided in one embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0205] Furthermore, another embodiment of the present specification provides another computer program product, which implements the following process when executed by a processor:

[0206] Obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application;

[0207] Generate a play window layout parameter and a play content parameter of the application based on the play probability distribution;

[0208] A play window for playing the video in the application is configured according to the play window layout parameters, and the video is played in the play window according to the play content parameters.

[0209] Another computer program product provided by this embodiment first obtains the playback probability distribution of the video played by the visiting user through the application on the live broadcast platform to provide a data basis for the subsequent personalized window layout and playback content recommendation, and then generates the playback window layout parameters and playback content parameters of the application based on the playback probability distribution, so as to improve the user experience in the video watching process and reduce the complexity of the user manually configuring the playback window. Finally, the playback window for video playback in the application is configured according to the playback window layout parameters, and the video is played in the playback window according to the playback content parameters, thereby improving the efficiency of the video playback window configuration and the accuracy of the playback content recommendation, while also improving the convenience of the visiting user using multiple windows for video playback on the live broadcast platform and improving the visiting user's experience of watching videos on the live broadcast platform.

[0210] The computer program product provided in one embodiment of this specification can implement each process in the aforementioned method embodiment and achieve the same functions and effects, which will not be repeated here.

[0211] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0212] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks 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.

[0213] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.

[0214] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0215] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0216] Memory may include non-permanent storage in a computer-readable storage medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable storage medium.

[0217] Computer-readable storage media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer-readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable storage media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0218] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0219] 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. Moreover, 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 disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0220] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A video playback method, characterized in that: The method comprises: Acquire the playback behavior data of the visiting user of the live broadcast platform, and encode the playback behavior data to obtain playback behavior characteristics; the playback behavior data includes the behavior data of the visiting user's video viewing time, video clicking, video switching, video closing and / or video playback window adjustment; Inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution; the first playback probability distribution includes a set of probabilities obtained by performing the playback layout prediction and the playback content prediction based on the playback behavior feature; Inputting the first playback probability distribution, the network status index, and the hardware performance index into the adjustment model to adjust the probability distribution, and outputting a second playback probability distribution; The second playback probability distribution is sent to the user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters of at least one playback window for video playback in the application.

2. According to the video playback method of claim 1, the playback behavior prediction model comprises a feature extraction module, an attention module and a feature conversion module; The playback layout prediction and playback content prediction include: Extracting the timing features of the playing behavior features by the feature extraction module to obtain timing features; Obtaining weighted features by weighting the temporal features through the attention module; The first playback probability distribution is obtained by performing probability distribution conversion on the weighted feature through the feature conversion module.

3. The video playback method according to claim 1, wherein the probability distribution adjustment comprises: The basic weight of the video playback index and the index influencing factor are input into the weight adjustment algorithm for calculation to obtain the target weight; The video playback index includes the network status index and the hardware performance index; Probability distribution adjustment is performed on the first playback probability distribution according to the target weight to obtain the second playback probability distribution.

4. The video playback method according to claim 3, wherein the step of inputting the basic weight of the video playback index and the index influencing factor into a weight adjustment algorithm for calculation to obtain a target weight comprises: Calculate the target basic weight according to the basic weight, the indicator influencing factor and the adjustment coefficient; The basic weights are summed to obtain a basic weight sum, and the target weight is obtained by calculating a ratio of the target basic weight to the basic weight sum.

5. A video playback method, characterized in that: The method comprises: Obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application; the playback probability distribution is obtained by inputting the first playback probability distribution, network status indicators and hardware performance indicators into the adjustment model to adjust the probability distribution; the first playback probability distribution is obtained by inputting the playback behavior characteristics into the playback behavior prediction model to predict the playback layout and the playback content; the playback behavior characteristics are obtained by encoding the playback behavior data of the visiting user; the first playback probability distribution includes a set of probabilities obtained by performing the playback layout prediction and the playback content prediction based on the playback behavior characteristics; the playback behavior data includes the behavior data of the visiting user on video viewing time, clicking on videos, switching videos, closing videos and / or adjusting the size of the video playback window; Generate a play window layout parameter and a play content parameter of at least one play window of the application based on the play probability distribution; The at least one playback window for playing the video in the application is configured according to the playback window layout parameters, and the video is played in the at least one playback window according to the playback content parameters.

6. The video playback method according to claim 5, wherein generating the playback window layout parameters and playback content parameters of the application based on the playback probability distribution comprises: Determining a playback mode for playing the video in the application based on the playback probability distribution; If the playback mode is a multi-window mode, determining window layout parameters and playback content parameters of each playback window; If the playback mode is a single-window mode, the window layout parameters and playback content parameters of the playback window are determined.

7. The video playback method according to claim 6, wherein determining playback content parameters of each playback window comprises: Sorting the playback contents in descending order according to the probability of each playback content in the playback probability distribution to obtain a ranking of the playback contents; The playback content parameters of each playback window are determined according to the ranking.

8. A video playback device, characterized in that: The device comprises: A data processing module is used to obtain the playback behavior data of the visiting user of the live broadcast platform, and encode the playback behavior data to obtain playback behavior characteristics; the playback behavior data includes the video viewing time, video clicking, video switching, video closing and / or video playback window size adjustment behavior data of the visiting user; A prediction module, used for inputting the playback behavior feature into a playback behavior prediction model to perform playback layout prediction and playback content prediction, and outputting a first playback probability distribution; the first playback probability distribution includes a set of probabilities obtained by performing the playback layout prediction and the playback content prediction based on the playback behavior feature; An adjustment module, configured to input the first playback probability distribution, the network status index, and the hardware performance index into an adjustment model to adjust the probability distribution, and output a second playback probability distribution; A sending module is used to send the second playback probability distribution to a user terminal; the second playback probability distribution is used to generate playback window layout parameters and playback content parameters of at least one playback window for video playback in an application.

9. A video playback device, characterized in that: The device comprises: A data acquisition module is used to obtain the playback probability distribution of the video played by the visiting user on the live broadcast platform through the application; the playback probability distribution is obtained by inputting the first playback probability distribution, network status indicators and hardware performance indicators into the adjustment model to adjust the probability distribution; the first playback probability distribution is obtained by inputting the playback behavior characteristics into the playback behavior prediction model to predict the playback layout and the playback content; the playback behavior characteristics are obtained by encoding the playback behavior data of the visiting user; the first playback probability distribution includes a set of probabilities obtained by performing the playback layout prediction and the playback content prediction based on the playback behavior characteristics; the playback behavior data includes the behavior data of the visiting user's video viewing time, clicking on videos, switching videos, closing videos and / or adjusting the size of the video playback window; A parameter generation module, used to generate a playback window layout parameter and a playback content parameter of at least one playback window of the application based on the playback probability distribution; The video playing module is used to configure the at least one playing window for playing the video in the application according to the playing window layout parameters, and to play the video in the at least one playing window according to the playing content parameters.

10. A video playback device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer executable instructions, and when the computer executable instructions are executed on the processor, the steps of the method described in any one of claims 1 to 4 can be implemented.

11. A video playback device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer executable instructions, and when the computer executable instructions are executed on the processor, the steps of the method described in any one of claims 5 to 7 can be implemented.

12. A computer-readable storage medium having computer-executable instructions stored therein, characterized in that: When the computer executable instructions are executed by a processor, the steps of the method described in any one of claims 1 to 4 can be implemented.

13. A computer-readable storage medium having computer-executable instructions stored therein, characterized in that: When the computer executable instructions are executed by a processor, the steps of the method described in any one of claims 5 to 7 can be implemented.

14. A computer program product, characterized in that The computer program product includes a video player program, and when the video player program is executed by a processor, the steps of the method described in any one of claims 1 to 4 can be implemented.

15. A computer program product, characterized in that The computer program product includes a video playback program, and when the video playback program is executed by a processor, the steps of the method described in any one of claims 5 to 7 can be implemented.

Images