A method and device for determining a video playback data source

Abstract

The present disclosure provides a method and device for determining a video playback data source, which are applied to a playback terminal. The method and device perform localized adaptive training on a data source download time prediction model through historical statistical data of data source calls by the playback terminal. The method can obtain the data download prediction duration corresponding to each data source by combining the data source information of the data source under the condition of current operating environment data and historical statistical data of data source calls, and then compare the available data download duration corresponding to the video playback download request with the data download prediction duration corresponding to each data source to determine a suitable data source for video playback, thereby meeting the video viewing needs of users.

Description

Technical Field

[0001] The present disclosure relates to the technical field of streaming media, and in particular to a method and device for determining a video playback data source. Background Art

[0002] With the rapid development of Internet technology and the upgrading of hardware configuration, more and more users can watch videos anytime and anywhere through playback terminals with video playback functions. In order to ensure that users are provided with high-quality video playback functions and experiences, video providers often provide multiple data sources from different sources for the playback terminal to select one of the data sources for video playback.

[0003] However, data sources from different sources have different degrees of data transmission limitations in different scenarios. Therefore, how to determine a suitable data source for video playback has become a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] In view of the above problems, the present disclosure provides a method and device for determining a video playback data source that overcomes the above problems or at least partially solves the above problems. The technical solution is as follows:

[0005] A method for determining a video playback data source, applied to a playback terminal, the method comprising:

[0006] Obtaining a first data source download time prediction model;

[0007] Obtaining historical statistical data of calls of the playback terminal to multiple data sources respectively;

[0008] Using the call history statistical data to train the first data source download time prediction model to obtain a second data source download time prediction model;

[0009] When the playback terminal initiates a video playback download request, obtain the available download duration and operating environment data of the data corresponding to the video playback download request, and obtain the data source information of each of the data sources;

[0010] For any of the data sources: inputting the operating environment data, the call history statistics of the data source and the data source information into the second data source download time prediction model, and obtaining the data download prediction duration corresponding to the data source output by the second data source download time prediction model;

[0011] The video playback data source is determined from among the data sources by using the available download time of the data, the predicted download time of the data corresponding to each of the data sources, and the data source information.

[0012] Optionally, after determining the video playback data source from each of the data sources by using the available data download duration and the predicted data download duration corresponding to each of the data sources, the method further includes:

[0013] Calling the video playback data source to respond to the video playback download request, and obtaining call indicator data of the current call of the video playback data source;

[0014] Using the call indicator data to update the call history statistics of the video playback data source;

[0015] Generate first call log data corresponding to the video playback data source based on the updated call history statistics of the video playback data source, the operating environment data, the data source information of the video playback data source, and the actual duration of data download of the video playback data source under this call;

[0016] The first call log data is transmitted to a server, so that the server uses the first call log data to train the first data source download time prediction model to obtain the trained first data source download time prediction model.

[0017] Optionally, obtaining the first data source download time prediction model includes:

[0018] Obtain the first data source download time prediction model trained by the server, wherein the first data source download time prediction model is trained by the server based on second call log data transmitted by different playback terminals and corresponding to each data source.

[0019] Optionally, the video playback download request includes a requested data time point and a data usage time point, and the available data download duration is the time interval from the requested data time point to the data usage time point.

[0020] Optionally, the data source information includes a cost level of the data source.

[0021] Optionally, using the available download duration of the data, the predicted download duration of the data corresponding to each of the data sources, and the data source information to determine a video playback data source from each of the data sources includes:

[0022] Determine at least one first data source whose predicted downloading time of the data is not greater than the available downloading time of the data;

[0023] Among the at least one first data source, one of the first data sources with the lowest cost level is determined as the video playback data source.

[0024] Optionally, using the available download duration of the data, the predicted download duration of the data corresponding to each of the data sources, and the data source information to determine a video playback data source from each of the data sources includes:

[0025] In a case where the predicted data download durations corresponding to the data sources are all greater than the available data download durations, determining at least one second data source with the lowest cost level among the data sources;

[0026] Among the at least one second data source, the second data source having the shortest predicted data downloading duration is determined as the video playback data source.

[0027] Optionally, the operating environment data includes network connection mode, operator, maximum bandwidth, operating system, hardware configuration parameters, time period, region and data length.

[0028] Optionally, the call history statistics include the total number of requests of the data source, the number of errors, the error rate, the average download speed, the exponentially weighted average speed, the number of slow speeds, the slow speed ratio, the weighted slow speed ratio, the number of downloads of requested data related resources, the most recent download time interval of requested data related resources, and the number of successful downloads of requested data related resources.

[0029] A video playback data source determination device is applied to a playback terminal, the device comprising: a first model acquisition unit, a call history statistical data acquisition unit, a second model acquisition unit, a data information acquisition unit, a data download prediction duration acquisition unit and a video playback data source determination unit.

[0030] The first model obtaining unit is used to obtain a first data source download time prediction model;

[0031] The calling history statistics data obtaining unit is used to obtain the calling history statistics data of the playback terminal to the multiple data sources respectively;

[0032] The second model obtaining unit is used to train the first data source download time prediction model by using the call history statistical data to obtain a second data source download time prediction model;

[0033] The data information obtaining unit is used to obtain the available download time and operating environment data of the data corresponding to the video playback download request when the playback terminal initiates a video playback download request, and obtain the data source information of each of the data sources;

[0034] The data download prediction duration obtaining unit is used for: for any of the data sources: inputting the operating environment data, the call history statistics of the data source and the data source information into the second data source download time prediction model, and obtaining the data download prediction duration corresponding to the data source output by the second data source download time prediction model;

[0035] The video play data source determination unit is used to determine a video play data source from each of the data sources by using the available data download duration, the predicted data download duration corresponding to each of the data sources, and the data source information.

[0036] By means of the above technical scheme, a method and device for determining a video playback data source provided by the present disclosure are applied to a playback terminal, and can obtain a first data source download time prediction model; obtain historical statistical data of calls of the playback terminal to multiple data sources respectively; use the historical statistical data to train the first data source download time prediction model to obtain a second data source download time prediction model; when the playback terminal initiates a video playback download request, obtain the data available download time and operating environment data corresponding to the video playback download request, and obtain the data source information of each data source; for any data source: input the operating environment data, the historical statistical data of calls of the data source, and the data source information into the second data source download time prediction model, and obtain the data download prediction time corresponding to the data source output by the second data source download time prediction model; use the data available download time and the data download prediction time corresponding to each data source to determine the video playback data source in each data source. The present invention performs localized adaptive training on a data source download time prediction model by using historical statistical data of calls to data sources by playback terminals. In the case of current operating environment data and historical statistical data of data source calls, combined with data source information of the data source, the trained data source download time prediction model can be used to obtain the data download prediction duration corresponding to each data source. The data available download duration corresponding to the video playback download request is then compared with the data download prediction duration corresponding to each data source to determine a suitable data source for video playback, thereby satisfying the user's video viewing needs.

[0037] The above description is only an overview of the technical solution of the present disclosure. In order to more clearly understand the technical means of the present disclosure, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present disclosure more obvious and easy to understand, the specific implementation methods of the present disclosure are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present disclosure. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0039] Figure 1 A logical diagram showing an existing data source selection method;

[0040] Figure 2 A schematic diagram showing a flow chart of an implementation of a method for determining a video playback data source provided by an embodiment of the present disclosure;

[0041] Figure 3 A schematic diagram illustrating the predicted duration of data downloading provided by an embodiment of the present disclosure is shown;

[0042] Figure 4 A flow chart showing another implementation of the method for determining a video playback data source provided by an embodiment of the present disclosure;

[0043] Figure 5 A logical schematic diagram of a method for determining a video playback data source provided by an embodiment of the present disclosure is shown;

[0044] Figure 6 A schematic diagram showing the structure of a device for determining a video playback data source provided by an embodiment of the present disclosure is shown;

[0045] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0046] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0047] Figure 1The figure shows a logic diagram of the existing data source selection method. The existing data source selection method selects data sources based on the water level of the existing data in the cache pool and the current time segment corresponding to the data to be downloaded, using different data source selection strategies corresponding to different time segments. For example: the amount of data to be downloaded is determined based on different business scenarios (normal playback speed, multiples, etc.) and playback bitrates, and then the 60-second backward preparation playback buffer is divided into three different time periods with the current playback time as the origin. Time period one includes 0 to 20 seconds, time period two includes 20 to 40 seconds, and time period three includes 40 to 60 seconds. Different data sources are selected based on the time period where the current data fills the cache area. A high-quality and high-cost strategy is adopted in time period one, a high-quality and low-cost competition strategy is adopted in time period two, and a low-cost competition strategy is adopted in time period three for downloading.

[0048] In order to avoid affecting the current playback behavior, the existing data source selection method often tends to be conservative in the division of the cache area. For example, when the cache water level is low, low-cost bandwidth cannot be used, and the division of this cache area cannot be based on different environmental changes, and the low-cost bandwidth cannot be maximized. At the same time, the existing data source selection method lacks an effective feedback mechanism. It is impossible to know whether the cache water level area division effect is ideal and whether the strategy within the water level area is efficient. Adjustment requires manual intervention.

[0049] In order to solve the above problem, it is necessary to obtain the data source download time required by each data source when requesting data from each data source. If the data source download time can be accurately predicted, the data source that meets the needs can be further selected.

[0050] The method for determining a video playback data source provided in the embodiments of the present disclosure may be applied to a playback terminal, wherein the playback terminal may be a device with a video playback function, such as a smart phone, a tablet computer, and a laptop computer.

[0051] like Figure 2 As shown, a flow chart of an implementation of a method for determining a video playback data source provided by an embodiment of the present disclosure is provided. The method for determining a video playback data source may include:

[0052] S100: Obtain a first data source download time prediction model.

[0053] The data source download time prediction model may be a model built based on a machine learning algorithm. Optionally, the data source download time prediction model may be a LightGBM (Light Gradient Boosting Machine) algorithm model.

[0054] Optionally, the disclosed embodiment can obtain a first data source download time prediction model trained by the server, wherein the first data source download time prediction model is trained by the server based on second call log data corresponding to each data source and transmitted by different playback terminals.

[0055] The server may be a server that provides video playback services to the playback terminal.

[0056] Specifically, the server can collect call log data transmitted by different playback terminals and corresponding to each data source, use these call log data as training data, perform machine learning on the first data source download time prediction model, and obtain a trained first data source download time prediction model.

[0057] Among them, the call log data may include, when the playback terminal calls a data source for video playback: the call history statistics of the playback terminal's calls to the data source, the operating environment data when the playback terminal initiates a video playback download request, the data source information of the data source, and the actual data download duration of the data source under this call.

[0058] Optionally, the call history statistics include the total number of requests for the data source, the number of errors, the error rate, the average download speed, the exponentially weighted average speed, the number of slow times, the slow ratio, the weighted slow ratio, the number of downloads of resources related to the requested data, the time interval between the most recent downloads of resources related to the requested data, and the number of successful downloads of resources related to the requested data.

[0059] The total number of requests is the total number of times the playback terminal calls the data source within a first preset time period. Optionally, the first preset time period may be 3 days.

[0060] The number of errors is the number of errors that occur when the playback terminal calls the data source. Optionally, the error may include that the data source cannot obtain correct data, and / or that the data source cannot normally complete the data acquisition process.

[0061] The error rate is the ratio of the number of errors to the total number of requests.

[0062] The average download speed is the average data download speed counted after the playback terminal successfully calls the data source each time.

[0063] The exponentially weighted average speed is the download speed corresponding to each successful call to the data source by the playback terminal in history, and is a speed weightedly calculated in the order of the call time. The later the download speed, the greater the weight corresponding to it.

[0064] The number of slow times is the total number of times the actual download speed of data from the data source is lower than the predicted download speed.

[0065] The slow ratio is the ratio of the number of slow requests to the total number of requests.

[0066] Among them, the weighted slow ratio is the ratio of the weighted slow times to the total number of requests, and the weighted slow times are the values ​​calculated by multiplying the slow times calculated in different time periods by the corresponding weights. The disclosed embodiment can divide the time periods according to the second preset time period in the order of the calling time for each slow time of the playback terminal in the history of calling the data source, and weightedly calculate the slow times corresponding to each time period to obtain the weighted slow times. For example: The disclosed embodiment can divide the second preset time period into 10 time periods, linearly assign weights of 0 to 2 to each time period, multiply the slow times of different time periods by the weights to obtain the weighted slow times, and obtain the weighted slow ratio by calculating the ratio of the weighted slow times to the total number of requests.

[0067] The number of downloads of the requested data-related resources is the number of downloads of the related resources that are coherent with the video data downloaded from the data source. For example, the related resources may include the front and back segment information of the video data.

[0068] The time interval between the most recent downloads of the requested data-related resources is the time interval between the most recent downloads of the relevant resources by the data source.

[0069] The number of successful downloads of requested data-related resources is the number of times the data source completely downloads the related resources.

[0070] Optionally, the operating environment data includes network connection mode, operator, maximum bandwidth, operating system, hardware configuration parameters, time period, region and data length.

[0071] The network connection mode refers to the network mode used by the playback terminal when initiating a video playback download request, for example: WI-FI network, 4G network, and 5G network.

[0072] The operator is the provider of the communication network used by the playback terminal when initiating a video playback download request, for example: a mobile communication operator, a Unicom communication operator, and a Telecom communication operator.

[0073] The maximum bandwidth is the network connection mode used by the playback terminal when initiating a video playback download request and / or the maximum bandwidth historically provided by the operator.

[0074] The operating system is the operating environment currently running on the playback terminal, for example, Android system, IOS system, Windows system, MacOS system, and Linux system.

[0075] The hardware configuration parameters are parameters of the basic configuration hardware of the playback terminal, for example: CPU (main frequency, instruction set), memory size, and available memory size.

[0076] The time period is the current time in the region where the playback terminal is located.

[0077] The region is a parameter corresponding to the geographical location of the playback terminal.

[0078] The data length is the data length requested to be downloaded in the video playback download request initiated by the playback terminal.

[0079] Optionally, the data source information may include attribute information of the data source and system load information.

[0080] The attribute information includes the requesting IP region of the data source, the IP operator, the data source type (CDN / PCDN / P2P), and a predefined cost level. The lower the cost level, the lower the cost of the data source.

[0081] The system load information includes the total number of download threads and the total download bandwidth provided by the data source.

[0082] Optionally, the server may use call history statistics, operating environment data and data source information as feature data, and the actual data download duration as a label of the feature data, to train a first data source download time prediction model to obtain a trained first data source download time prediction model.

[0083] Optionally, the disclosed embodiment may obtain a trained first data source download time prediction model from the server when the daily playback terminal is started for the first time.

[0084] The disclosed embodiment obtains a unified data source download time prediction model provided by the server for different playback terminals, which is beneficial for the playback terminal to use the first data source download time prediction model to predict the data download duration of the data source, thereby saving resources of the video playback service provider.

[0085] S110: Obtain historical statistical data of calls made by the playback terminal to multiple data sources.

[0086] S120: Train the first data source download time prediction model by calling historical statistical data to obtain a second data source download time prediction model.

[0087] After obtaining the first data source download time prediction model trained by the server, the playback terminal provided by the embodiment of the present disclosure can use the call history statistical data corresponding to multiple data sources stored locally by the playback terminal to retrain the first data source download time prediction model using the call history statistical data to obtain the second data source download time prediction model.

[0088] The disclosed embodiment retrains the data source download time prediction model by using the call history statistical data stored locally on the playback terminal. It can perform localized adaptive transformation on the basis of the data source download time prediction model uniformly trained on the server side to obtain a data source download time prediction model dedicated to the playback terminal.

[0089] The disclosed embodiment can provide exclusive customized data source download time prediction models for different playback terminals by retraining the data source download time prediction model locally on the playback terminal, thereby improving the adaptability of the data source download time prediction model on the playback terminal, thereby being able to more accurately predict the data download duration of the data source.

[0090] S130: When the playback terminal initiates a video playback download request, obtain the available download duration and operating environment data of the data corresponding to the video playback download request, and obtain data source information of each data source.

[0091] The video playback download request is a download request for video data corresponding to the target video.

[0092] Optionally, the video playback download request includes a data request time point and a data usage time point, and the data available download time length is the time interval from the data request time point to the data usage time point.

[0093] The disclosed embodiment can utilize the request data time point and the data usage time point to automatically calculate the available download time of the data and obtain the operating environment data of the playback terminal when initiating a video playback download request.

[0094] The disclosed embodiment can send a video playback download request to a scheduling server, which obtains scheduling addresses of multiple data sources corresponding to the video data from a content distribution network according to its own strategy, and obtains data source information of the data source through the scheduling address of the data source.

[0095] S140. For any data source: input the operating environment data, the call history statistics of the data source and the data source information into the second data source download time prediction model to obtain the data download prediction duration corresponding to the data source output by the second data source download time prediction model.

[0096] For ease of understanding, here we combine Figure 3The following diagram is used to illustrate the predicted duration of data download: Figure 3 It can be seen that the available data download time is the time interval from the time point of requesting data to the time point of data use. The predicted data download time corresponding to data source G and data source H is within the available data download time, while the predicted data download time corresponding to data source G is shorter than the predicted data download time corresponding to data source H, and the predicted data download time corresponding to data source I exceeds the available data download time.

[0097] S150: Determine a video playback data source from among the data sources by using the available data download duration, the predicted data download duration corresponding to each data source, and the data source information.

[0098] Optionally, the disclosed embodiment may determine at least one first data source whose predicted data download duration is not greater than the available data download duration; and among the at least one first data source, determine a first data source with the lowest cost level as the video playback data source.

[0099] The disclosed embodiments can use the lowest-cost data source as the video playback data source while ensuring that the video data can be downloaded in a timely manner, and provide a video data download service to improve the cost-effectiveness of the data source, thereby reducing the cost of the video playback provider without affecting the user's normal video viewing.

[0100] Optionally, in an embodiment of the present disclosure, when the predicted data download duration corresponding to each data source is greater than the available data download duration, at least one second data source with the lowest cost level may be determined among each data source; and among at least one second data source, a second data source with the shortest predicted data download duration may be determined as the video playback data source.

[0101] In the case where all data sources are unable to download video data within the available data download time, the disclosed embodiment can preferentially select a data source with the shortest predicted data download time among the lowest-cost data sources as the video playback data source, provide video data download services, and reduce the cost of the video playback provider while reducing the impact on users' video viewing.

[0102] It is understandable that the disclosed embodiment can also select the video playback data source based on at least one of the attribute information of the data source in the data source information, combined with the available data download time and the predicted data download time corresponding to each data source. For example: the disclosed embodiment can preferentially select a data source with a data source type of PCDN as the video playback data source among the data sources whose predicted data download time is not greater than the available data download time, followed by a data source with a data source type of P2P, and finally a data source with a data source type of CDN.

[0103] Optional, based on Figure 2 The method shown, such as Figure 4 As shown, a flow chart of another implementation of the method for determining a video playback data source provided by an embodiment of the present disclosure is provided. After step S150, the method for determining a video playback data source may further include:

[0104] S160: Call the video playback data source to respond to the video playback download request, and obtain the calling index data of the current call of the video playback data source.

[0105] The disclosed embodiment can call the playback data source to download the video data of the target video requested in the video playback download request, and collect the calling index data during the download process.

[0106] The calling index data is the corresponding data collected for the preset calling index items during the video data download process of the video playback data source. The setting of the calling index items is related to the data of the data source to be counted in the calling historical statistical data.

[0107] Optionally, the call index items may include whether an error occurs in calling the data source, the download speed of the video data, whether related resources that are coherent with the video data are downloaded, and whether the download of the related resources is successful.

[0108] S170: Use the call index data to update the call history statistics of the video playback data source.

[0109] It is understandable that the call index data generated during the video playback data source downloading process will cause the playback terminal to change the historical statistical data of the video playback data source. For example, if an error occurs in the video playback data source downloading, the error count of the video playback data source should be increased by 1, and the error rate of the video playback data source should be updated according to the updated error count.

[0110] The disclosed embodiment can update the call history statistics of the video playback data source in real time by collecting the call index data of the video playback data source, so as to subsequently optimize the data source download time prediction model.

[0111] S180, generating first call log data corresponding to the video playback data source based on the updated call history statistics of the video playback data source, the operating environment data, the data source information of the video playback data source, and the actual duration of data downloading of the video playback data source under this call.

[0112] S190. Transmit the first call log data to the server, so that the server uses the first call log data to train the first data source download time prediction model to obtain the trained first data source download time prediction model.

[0113] The server can collect the call log data uploaded by the playback terminal, and use the call log data as training data to train the data source download time prediction model, so that the server can use the latest data uploaded by the playback terminal to optimize the data source download time prediction model, so that the playback terminal can subsequently obtain a data source download time prediction model with better prediction effect on the server.

[0114] In order to facilitate understanding of the overall solution of the embodiment of the present disclosure, Figure 5 The logic diagram of the method for determining the video playback data source shown is used to illustrate: the embodiment of the present disclosure collects the actual parameters of data requested by all playback terminals from each data source as features and the final speed as a label to make training data, uses lightGBM as a training model for training, and sends the trained first data source download time prediction model to each playback terminal every day. After the playback terminal obtains the first data source download time prediction model, it uses the local recent historical data for further adaptive training to obtain the second data source download time prediction model. When data is needed, a maximum waiting time (deadline) for using the data is provided, that is, the available download time of the data. Different data sources use the second data source download time prediction model to predict the download time, compare the available download time of the data, and select the data source with the lowest cost from all data sources that are less than the available download time of the data for download. The actual download speed is reported as training data and used as training data for the next round of the first data source download time prediction model and the second data source download time prediction model.

[0115] The present invention provides a method for determining a video playback data source, which is applied to a playback terminal and can obtain a first data source download time prediction model; obtain historical statistical data of calls to multiple data sources by the playback terminal respectively; train the first data source download time prediction model using the historical statistical data of calls to obtain a second data source download time prediction model; when the playback terminal initiates a video playback download request, obtain the data available download time and operating environment data corresponding to the video playback download request, and obtain the data source information of each data source; for any data source: input the operating environment data, the historical statistical data of calls to the data source, and the data source information into the second data source download time prediction model, and obtain the data download prediction time corresponding to the data source output by the second data source download time prediction model; and determine the video playback data source from each data source using the data available download time and the data download prediction time corresponding to each data source. The present invention performs localized adaptive training on a data source download time prediction model by using historical statistical data of calls to data sources by playback terminals. In the case of current operating environment data and historical statistical data of data source calls, combined with data source information of the data source, the trained data source download time prediction model can be used to obtain the data download prediction duration corresponding to each data source. The data available download duration corresponding to the video playback download request is then compared with the data download prediction duration corresponding to each data source to determine a suitable data source for video playback, thereby satisfying the user's video viewing needs.

[0116] Although operations are depicted in a particular order, this should not be understood as requiring that the operations be performed in the particular order shown or in a sequential order.Multitasking and parallel processing may be advantageous under certain circumstances.

[0117] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0118] Corresponding to the above method embodiment, the present disclosure embodiment also provides a video playback data source determination device, whose structure is as follows: Figure 6 As shown, applied to a playback terminal, the video playback data source determination device may include: a first model acquisition unit 01, a call history statistical data acquisition unit 02, a second model acquisition unit 03, a data information acquisition unit 04, a data download prediction duration acquisition unit 05 and a video playback data source determination unit 06.

[0119] The first model obtaining unit 01 is used to obtain a first data source download time prediction model.

[0120] The calling history statistics data obtaining unit 02 is used to obtain the calling history statistics data of the playback terminal for multiple data sources respectively.

[0121] The second model obtaining unit 03 is used to train the first data source download time prediction model by calling historical statistical data to obtain the second data source download time prediction model.

[0122] The data information obtaining unit 04 is used to obtain the available download time and operating environment data of the data corresponding to the video playback download request when the playback terminal initiates a video playback download request, and obtain the data source information of each data source.

[0123] The data download prediction duration acquisition unit 05 is used for any data source: inputting the operating environment data, the call history statistics data of the data source and the data source information into the second data source download time prediction model, and obtaining the data download prediction duration corresponding to the data source output by the second data source download time prediction model.

[0124] The video play data source determination unit 06 is used to determine the video play data source from each data source by using the available data download time, the predicted data download time corresponding to each data source and the data source information.

[0125] Optionally, the video playback data source determination device may further include: a calling indicator data acquisition unit, a calling historical statistical data updating unit, a calling log data generation unit, and a calling log data transmission unit.

[0126] The calling indicator data obtaining unit is used to use the available data download time, the predicted data download time corresponding to each data source and the data source information in the video playback data source determination unit 06 to determine the video playback data source in each data source, and then call the video playback data source to respond to the video playback download request to obtain the calling indicator data of the current call of the video playback data source.

[0127] The calling history statistics data updating unit is used to update the calling history statistics data of the video playback data source by using the calling index data.

[0128] The call log data generation unit is used to generate the first call log data corresponding to the video playback data source based on the call history statistics data, operating environment data, data source information of the video playback data source and the actual data download duration of the video playback data source under this call.

[0129] The call log data transmission unit is used to transmit the first call log data to the server, so that the server uses the first call log data to train the first data source download time prediction model to obtain the trained first data source download time prediction model.

[0130] Optionally, the first model acquisition unit 01 can be specifically used to obtain a first data source download time prediction model trained by the server, wherein the first data source download time prediction model is trained by the server based on second call log data transmitted by different playback terminals and corresponding to each data source.

[0131] Optionally, the video playback download request includes a data request time point and a data usage time point, and the data available download time length is the time interval from the data request time point to the data usage time point.

[0132] Optionally, the data source information includes a cost level of the data source.

[0133] Optionally, the video playback data source determination unit 06 can be specifically used to determine at least one first data source whose predicted data download duration is not greater than the available data download duration; among the at least one first data source, determine the first data source with the lowest cost level as the video playback data source.

[0134] Optionally, the video playback data source determination unit 06 can be specifically used to determine at least one second data source with the lowest cost level among each data source when the predicted data download duration corresponding to each data source is greater than the available data download duration; and among at least one second data source, determine a second data source with the shortest predicted data download duration as the video playback data source.

[0135] Optionally, the operating environment data includes network connection mode, operator, maximum bandwidth, operating system, hardware configuration parameters, time period, region and data length.

[0136] Optionally, the call history statistics include the total number of requests for the data source, the number of errors, the error rate, the average download speed, the exponentially weighted average speed, the number of slow times, the slow ratio, the weighted slow ratio, the number of downloads of resources related to the requested data, the time interval between the most recent downloads of resources related to the requested data, and the number of successful downloads of resources related to the requested data.

[0137] The present disclosure provides a video playback data source determination device, which is applied to a playback terminal and can obtain a first data source download time prediction model; obtain historical statistical data of calls to multiple data sources by the playback terminal respectively; train the first data source download time prediction model using the historical statistical data of calls to obtain a second data source download time prediction model; when the playback terminal initiates a video playback download request, obtain the data available download time and operating environment data corresponding to the video playback download request, and obtain the data source information of each data source; for any data source: input the operating environment data, the historical statistical data of calls to the data source, and the data source information into the second data source download time prediction model, and obtain the data download prediction time corresponding to the data source output by the second data source download time prediction model; and determine the video playback data source from each data source using the data available download time and the data download prediction time corresponding to each data source. The present invention performs localized adaptive training on a data source download time prediction model by using historical statistical data of calls to data sources by playback terminals. In the case of current operating environment data and historical statistical data of data source calls, combined with data source information of the data source, the trained data source download time prediction model can be used to obtain the data download prediction duration corresponding to each data source. The data available download duration corresponding to the video playback download request is then compared with the data download prediction duration corresponding to each data source to determine a suitable data source for video playback, thereby satisfying the user's video viewing needs.

[0138] Regarding the device in the above embodiment, the specific manner in which each unit performs the operation has been described in detail in the embodiment of the method, and will not be elaborated here.

[0139] The video playback data source determination device includes a processor and a memory. The above-mentioned first model acquisition unit 01, call historical statistical data acquisition unit 02, second model acquisition unit 03, data information acquisition unit 04, data download prediction duration acquisition unit 05 and video playback data source determination unit 06 are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0140] The processor includes a kernel, and the kernel calls the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters can be adjusted to perform localized adaptive training on the data source download time prediction model through the historical statistical data of the playback terminal calling the data source. Under the current operating environment data and the historical statistical data of the data source calling, combined with the data source information of the data source, the trained data source download time prediction model can be used to obtain the data download prediction duration corresponding to each data source, and then the data available download duration corresponding to the video playback download request is compared with the data download prediction duration corresponding to each data source, so as to determine the appropriate data source for video playback and meet the user's video viewing needs.

[0141] An embodiment of the present disclosure provides a computer-readable storage medium having a program stored thereon, and when the program is executed by a processor, the method for determining a video playback data source is implemented.

[0142] An embodiment of the present disclosure provides a processor, which is used to run a program, wherein the method for determining a video playback data source is executed when the program is running.

[0143] like Figure 7 As shown, the embodiment of the present disclosure provides an electronic device 1000, which includes at least one processor 1001, at least one memory 1002 connected to the processor 1001, and a bus 1003; wherein the processor 1001 and the memory 1002 communicate with each other through the bus 1003; the processor 1001 is used to call the program instructions in the memory 1002 to execute the above-mentioned video playback data source determination method. The electronic device in this article can be a server, a PC, a PAD, a mobile phone, etc.

[0144] The present disclosure also provides a computer program product, which, when executed on an electronic device, is suitable for executing a program that initializes the steps of a method for determining a video playback data source.

[0145] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, apparatuses, electronic devices (systems), and computer program products according to embodiments of the present disclosure. 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 device to generate a machine, so that the instructions executed by the processor of the computer or other programmable 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.

[0146] In a typical configuration, an electronic device includes one or more processors (CPU), a memory, and a bus. The electronic device may also include an input / output interface, a network interface, and the like.

[0147] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip. The memory is an example of a computer-readable medium.

[0148] Computer readable 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 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 magnetic 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 media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0149] In the description of the present disclosure, it should be understood that the terms "up", "down", "front", "back", "left" and "right" etc. indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the positions or elements referred to must have specific directions, be constructed and operate in specific directions. Therefore, they should not be understood as limitations of the present disclosure.

[0150] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. It should also be noted that the terms "include", "comprise" or any other variants 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 further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0151] It will be appreciated by those skilled in the art that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present disclosure 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 code.

[0152] The above are only embodiments of the present disclosure and are not intended to limit the present disclosure. For those skilled in the art, the present disclosure may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the scope of the claims of the present disclosure.

Claims

1. A method for determining a video playback data source, characterized in that: Applied to a playback terminal, the method includes: Obtaining a first data source download time prediction model, wherein the first data source download time prediction model is trained by the server based on second call log data corresponding to each data source and transmitted by different playback terminals; Obtaining historical statistical data of calls of the playback terminal to multiple data sources respectively; Using the call history statistical data to train the first data source download time prediction model to obtain a second data source download time prediction model; When the playback terminal initiates a video playback download request, obtain the available download duration and operating environment data of the data corresponding to the video playback download request, and obtain the data source information of each of the data sources; For any of the data sources: inputting the operating environment data, the call history statistics of the data source and the data source information into the second data source download time prediction model, and obtaining the data download prediction duration corresponding to the data source output by the second data source download time prediction model; The video playback data source is determined from among the data sources by using the available download time of the data, the predicted download time of the data corresponding to each of the data sources, and the data source information.

2. The method according to claim 1, characterized in that: After determining the video playback data source from each of the data sources by using the available data download duration and the predicted data download duration corresponding to each of the data sources, the method further includes: Calling the video playback data source to respond to the video playback download request, and obtaining call indicator data of the current call of the video playback data source; Using the call indicator data to update the call history statistics of the video playback data source; Generate first call log data corresponding to the video playback data source based on the updated call history statistics of the video playback data source, the operating environment data, the data source information of the video playback data source, and the actual duration of data download of the video playback data source under this call; The first call log data is transmitted to a server, so that the server uses the first call log data to train the first data source download time prediction model to obtain the trained first data source download time prediction model.

3. The method according to claim 1, characterized in that The video playback download request includes a data request time point and a data usage time point, and the data available download time length is the time interval from the data request time point to the data usage time point.

4. The method according to claim 1, characterized in that The data source information includes a cost level of the data source.

5. The method according to claim 4, characterized in that The using the available download duration of the data, the predicted download duration of the data corresponding to each of the data sources, and the data source information to determine a video playback data source from each of the data sources includes: Determine at least one first data source whose predicted downloading time of the data is not greater than the available downloading time of the data; Among the at least one first data source, one of the first data sources with the lowest cost level is determined as the video playback data source.

6. The method according to claim 4, characterized in that The using the available download duration of the data, the predicted download duration of the data corresponding to each of the data sources, and the data source information to determine a video playback data source from each of the data sources includes: In a case where the predicted data download durations corresponding to the data sources are all greater than the available data download durations, determining at least one second data source with the lowest cost level among the data sources; Among the at least one second data source, the second data source having the shortest predicted data downloading duration is determined as the video playback data source.

7. The method according to claim 1, characterized in that The operating environment data includes network connection mode, operator, maximum bandwidth, operating system, hardware configuration parameters, time period, region and data length.

8. The method according to claim 1, characterized in that: The call history statistics include the total number of requests of the data source, the number of errors, the error rate, the average download speed, the exponentially weighted average speed, the number of slow times, the slow ratio, the weighted slow ratio, the number of downloads of requested data related resources, the time interval between the most recent downloads of requested data related resources, and the number of successful downloads of requested data related resources.

9. A device for determining a video playback data source, characterized in that: Applied to a playback terminal, the device comprises: a first model acquisition unit, a call history statistics data acquisition unit, a second model acquisition unit, a data information acquisition unit, a data download prediction duration acquisition unit, and a video playback data source determination unit. The first model obtaining unit is used to obtain a first data source download time prediction model, wherein the first data source download time prediction model is obtained by training the server based on second call log data transmitted by different playback terminals and corresponding to each data source respectively; The calling history statistics data obtaining unit is used to obtain the calling history statistics data of the playback terminal to the multiple data sources respectively; The second model obtaining unit is used to train the first data source download time prediction model by using the call history statistical data to obtain a second data source download time prediction model; The data information obtaining unit is used to obtain the available download time and operating environment data of the data corresponding to the video playback download request when the playback terminal initiates a video playback download request, and obtain the data source information of each of the data sources; The data download prediction duration obtaining unit is used for: for any of the data sources: inputting the operating environment data, the call history statistics of the data source and the data source information into the second data source download time prediction model, and obtaining the data download prediction duration corresponding to the data source output by the second data source download time prediction model; The video play data source determination unit is used to determine a video play data source from each of the data sources by using the available data download duration, the predicted data download duration corresponding to each of the data sources, and the data source information.

Images