WO2017161744A1 - 一种视频推荐方法及装置 - Google Patents

一种视频推荐方法及装置 Download PDF

Info

Publication number
WO2017161744A1
WO2017161744A1 PCT/CN2016/088488 CN2016088488W WO2017161744A1 WO 2017161744 A1 WO2017161744 A1 WO 2017161744A1 CN 2016088488 W CN2016088488 W CN 2016088488W WO 2017161744 A1 WO2017161744 A1 WO 2017161744A1
Authority
WO
WIPO (PCT)
Prior art keywords
video
user
played
track
playing
Prior art date
Application number
PCT/CN2016/088488
Other languages
English (en)
French (fr)
Inventor
刘维娟
Original Assignee
乐视控股（北京）有限公司
乐视网信息技术（北京）股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 乐视控股（北京）有限公司, 乐视网信息技术（北京）股份有限公司 filed Critical 乐视控股（北京）有限公司
Publication of WO2017161744A1 publication Critical patent/WO2017161744A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H04N21/4668—Learning process for intelligent management, e.g. learning user preferences for recommending movies for recommending content, e.g. movies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
    • H04N21/25866—Management of end-user data
    • H04N21/25891—Management of end-user data being end-user preferences
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H04N21/4667—Processing of monitored end-user data, e.g. trend analysis based on the log file of viewer selections

Definitions

  • the present invention relates to the field of communications technologies, and in particular, to a video recommendation method and apparatus.
  • Embodiments of the present invention provide a video recommendation method and apparatus for recommending a video that is of interest to a user, thereby improving a viewing experience of the user.
  • a video recommendation method including:
  • the step of calculating the viewing probability of the user when the video to be played is recommended includes:
  • the quotient of the number of times the user views the recommended video to be played and the number of times the video to be played is recommended is calculated, and the viewing probability of the user when the video to be played is recommended is obtained.
  • the step of calculating the viewing probability of the user when the video to be played is recommended includes:
  • the product of the similarity and the user's viewing probability is calculated to obtain the probability of recommending the video to be played.
  • the method for calculating the similarity with the user video playback track includes:
  • a quotient of the number of elements of the third video set and the number of elements of the fourth video set is calculated, and the similarity between the target play track and the video play track is obtained.
  • the step of determining, in the database, the target play track that is similar to the user video play track with a similarity greater than a preset threshold includes:
  • a video recommendation apparatus including:
  • Receiving module receiving a user video playing track and a currently playing video
  • a first determining module configured to determine, in the database, a target playing track that is similar to a user video playing track with a similarity greater than a preset threshold, where the target playing track includes a currently playing video;
  • a second determining module configured to determine a to-be-played video after the currently playing video in the target playing track
  • a calculation module configured to calculate a viewing probability of the user when the video to be played is recommended
  • a third determining module configured to determine a to-be-played video whose viewing probability is greater than the first threshold as a recommended video
  • a sending module configured to send the recommended video to the terminal.
  • the calculation module includes:
  • a first obtaining sub-module configured to acquire a number of times the user views when the video to be played is recommended
  • a second obtaining submodule configured to obtain a number of times the video to be played is recommended
  • the first calculation sub-module is configured to calculate a quotient of the number of times the user views the recommended video to be played and the number of times the video to be played is recommended, and obtain the viewing probability of the user when the video to be played is recommended.
  • the calculation module includes:
  • a second calculation sub-module configured to calculate a user viewing probability when the to-be-played video is recommended during the playing of the currently played video
  • the third calculation sub-module is configured to calculate a product of the similarity and the user's viewing probability to obtain a probability of recommending the video to be played.
  • the first determining module includes:
  • a third acquiring submodule configured to acquire a first video set of the target playing track and a second video set of the user video playing track
  • a first determining submodule configured to determine an intersection of the first video set and the second video set as a third video set
  • a second determining submodule configured to determine a union of the first video set and the second video set as a fourth video set
  • a fourth calculating sub-module configured to calculate a quotient of the number of elements of the third video set and the number of elements of the fourth video set, to obtain a similarity between the target playing track and the user video playing track.
  • the first determining module includes:
  • a third determining submodule configured to determine, in the database, a play track having a currently played video
  • a determining submodule configured to determine whether the number of playing trajectories is greater than a second threshold
  • a server comprising: the video recommendation device of the second aspect.
  • an embodiment of the present invention provides a computer storage medium, wherein the computer storage medium can store a program, and when the program is executed, each of the video recommendation methods provided by the first aspect of the present invention can be implemented. Some or all of the steps in the implementation.
  • the database when receiving the user video playing track and the currently playing video sent by the terminal, it is indicated that the user who uses the terminal is currently watching the video, and in order to be able to push the video of interest for the user, the database needs to be first in the database. Determining the target play track with the video, and removing some low-scoring play tracks by preset thresholds, so as to know the videos to be played that other users may watch after watching the video. The to-be-played video whose viewing probability is greater than the first threshold when the user is recommended is determined as the recommended video, and finally the recommended video that the user is most likely to watch is sent to the terminal.
  • the present invention determines the video that the user is most likely to watch by analyzing the viewing behavior of other users and according to the viewing probability of the user. Therefore, the solution provided by the present invention can push the video of interest to the user to improve the viewing experience of the user.
  • FIG. 1 is a flowchart of a video recommendation method according to an embodiment of the present invention.
  • FIG. 2 is a flowchart of another video recommendation method according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a video recommendation apparatus according to an embodiment of the present invention.
  • FIG. 1 is a flowchart of a video recommendation method according to an embodiment of the present invention.
  • the video recommendation method provided by the present invention can push a video of interest to a user to improve the viewing experience of the user.
  • the method includes the following steps.
  • Step S11 Receive a video playback track of the user and the currently played video, and determine, in the database, a target play track whose similarity with the user video play track is greater than a preset threshold, where the target play track includes the currently played video.
  • the method provided in this embodiment may be applied to a server, and the server may be a computer or the like, and the terminal may be a smart TV, a smart phone, a tablet computer, and a home computer.
  • the user video playing track is a video sequence that the user watches in a certain period of time. For example, if a user watches an A movie, a B movie, a C movie, and a D movie in the last 5 days, then the user's video playback tracks are A movie, B movie, C movie, and D movie.
  • the database is set on the server side, and the database is used to store the playback trajectory of all users.
  • the playback trajectory of the user X is A movie, B movie, C movie, and D movie
  • the playback trajectory of the user Y is A movie, B movie, E movie, and F movie
  • the playback trajectory of the user Z is A movie, B movie, H movies and I movies.
  • the terminal When the user views the video using the terminal, the terminal sends the user's video playback track and the currently played video to the server, so that the server can analyze and push the appropriate video.
  • the server Upon receiving the video play track sent by the terminal and the currently played video, the server first determines in the database that the video with the user currently playing and the similarity with the user video play track is greater than a preset threshold, so as to facilitate Through these target playback tracks, it is known which videos other users will watch after watching the video currently played by the user, and at the same time, through the preset threshold, some playback trajectories with lower similarity can be removed.
  • the terminal first determines that the video playback track of the user U is the video X, the video Y, and the video A. And the currently playing video A, and then the terminal sends the video playback track XYA of the user U and the currently played video A to the server, so that the server can analyze and push the appropriate video.
  • the server receives the video play track XYA sent by the terminal and the currently played video A
  • the server first determines in the database that the video A with the user currently playing and the similarity with the video play track XYA is greater than a preset threshold. Tracks, in order to use these target playback tracks to know which videos other users will watch after watching the video currently played by the user, and remove some of the less similar playback tracks.
  • step S11 There are many ways to calculate the similarity between the target play track and the user video play track mentioned in step S11, and a manner is provided below.
  • a first video set of the target play track and a second video set of the user video play track are obtained.
  • the intersection of the first video set and the second video set is determined as the third video set.
  • the union of the first video set and the second video set is determined as a fourth video set.
  • the quotient of the number of elements of the third video set and the number of elements of the fourth video set is calculated, and the similarity between the target play track and the user video play track is obtained.
  • the target play track is XYAB
  • the first video set of the target play track is XYAB
  • the user video play track is XYA
  • the second video set of the user video play track is XYA.
  • a first video set XYAB of the target play track and a second video set XYA of the user video play track are acquired.
  • the intersection of the first video set XYAB and the second video set XYA is determined as the third video set XYA.
  • the union of the first video set XYAB and the second video set XYA is determined as the fourth video set XYAB.
  • the quotient of the number of elements (3) of the third video set XYA and the number of elements (4) of the fourth video set XYAB is calculated, and the similarity of the play track to the video play track is 0.75.
  • the step of determining, in the database, the target playing track with the similarity of the user video playing track in the database that is greater than the preset threshold in the step S11 may further include the following three sub-steps: Step A, determining in the database The playing track of the currently playing video; step B, calculating the number of the playing track; step C, determining that the similarity with the user video playing track is greater than the third in the playing track when the number of the playing track is greater than the second threshold The target playback track of the threshold.
  • the server determines and plays the video in the playing trajectory.
  • the similarity of the trajectory is greater than the target playback trajectory of the third threshold, so as to delete some playing trajectories with less similarity to the video playing trajectory.
  • the server determines in the database that the playing track with the currently playing video A is the first playing track, and the second.
  • the play track, the third play track, the fourth play track, and the fifth play track can calculate that the number of play tracks having the currently played video A is five, because the number of the play tracks (5) is greater than the second threshold ( 3), so the server will determine the target play track with the similarity of the user video play track to the third threshold (0.75) among the above five play tracks.
  • the server calculates that the similarity between the first playback track and the user video playback track is 0.8, the similarity between the second playback track and the user video playback track is 0.78, and the similarity between the third playback track and the user video playback track is 0.6.
  • the similarity between the fourth play track and the user video play track is 0.5, and the similarity between the fifth play track and the user video play track is 0.4, so the server determines the similarity with the user video play track in the above five play tracks.
  • the target play track whose degree is greater than the third threshold (0.75) is the first play track and the second play track, thereby removing the third play track, the fourth play track and the fifth with less similarity to the user video play track. Play the track.
  • Step S12 Determine a to-be-played video after the video is currently played in the target playing track.
  • the server determines the target play track of the currently played video of the user in the database, it is necessary to determine the to-be-played video after the user currently plays the video in the target play track.
  • the video behind the video currently played by the user is called a video to be played.
  • the server needs to separately determine the to-be-played video after the video A in the three target playback tracks, and the videos to be played after the video A in the three target playback tracks are B, B, and C, respectively.
  • Step S13 Calculate the viewing probability of the user when the video to be played is recommended.
  • the server determines the to-be-played video, it is required to calculate the viewing probability of the user when the to-be-played video is recommended.
  • the server pre-counts the situation in which the video is viewed by all users, and when the server recommends the video to be played, the number of times the video to be played is recommended is recorded. If the server detects that the terminal is watching the to-be-played video, the server displays the to-be-played video, indicating that the recommendation is successful, and records the number of times the recommended video is successfully played. Therefore, the server can calculate the quotient of the number of times the user is watching and the number of times the video to be played is recommended when the video to be played is recommended according to the number of times recorded before, so as to obtain the viewing probability of the user when the video to be played is recommended.
  • Step S14 Determine a to-be-played video whose viewing probability is greater than the first threshold as a recommended video.
  • the server After the server calculates the viewing probability of the user when the video to be played is recommended, the server determines the to-be-played video whose viewing probability is greater than the first threshold as the recommended video.
  • the server needs to separately determine the to-be-played video after the video A in the three target playback tracks, and the videos to be played after the video A in the three target playback tracks are B, B, and C, respectively. Since there are two videos to be played as B, the last videos to be played are B and C.
  • the server calculates that the probability of the user watching when the video B to be played is recommended is 0.6, and calculates that the probability of the user watching when the video C to be played is recommended is 0.3, and the preset first threshold is 0.5, so the server will probability.
  • the to-be-played video B that is greater than the first threshold of 0.5 is determined as the recommended video.
  • Step S15 Send the recommended video to the terminal and display it in the video recommendation list.
  • the server when the server receives the user video play track sent by the terminal and the currently played video, the user who uses the terminal is currently watching the video, and in order to be able to push the video of interest to the user, the server needs to Firstly, the target playing track with the video is determined in the database, and some playing tracks with lower similarity are removed by the preset threshold, so as to know the videos to be played that other users may watch after watching the video. The server then determines the to-be-played video whose user's viewing probability is greater than the first threshold when it is recommended as the recommended video, and finally sends the recommended video that the user is most likely to watch to the end. end.
  • the present invention determines the video that the user is most likely to watch by analyzing the viewing behavior of other users and according to the probability. Therefore, the solution provided by the embodiment of the present invention can push the video of interest to the user to improve the viewing experience of the user.
  • FIG. 2 is a flowchart of another video recommendation method according to an embodiment of the present invention. 2 is an alternative embodiment based on FIG. 1. In the embodiment shown in FIG. 2, the same portions as the embodiment shown in FIG. 1 can be referred to and explained in the embodiment shown in FIG. 1. The method shown in Figure 2 includes the following steps.
  • Step S21 Receive a user video play track and a currently played video, and determine, in the database, a target play track that is similar to the user video play track with a similarity greater than a preset threshold, where the target play track includes the currently played video.
  • Step S22 Determine a to-be-played video after the video is currently played in the target playback track.
  • Step S23 Calculate a user viewing probability when the to-be-played video is recommended during the current playing video playing process.
  • the server After the server determines that the video to be played, the server needs to calculate the user viewing probability when the video to be played is recommended during the current playing video playing.
  • the time when the video to be played is recommended is limited to the time when the user is watching the currently playing video, that is, the probability that the user clicks to watch the video to be played while watching the currently playing video, and the target playing track is calculated.
  • the video to be played also appears after the current video is played, and the two are matched with each other. At this time, the calculated user viewing probability is more accurate.
  • Step S24 Calculating a product of the similarity and the user's viewing probability to obtain a probability of recommending the video to be played.
  • the server calculates the user viewing probability when the video to be played by other users during the video playing process is recommended, the product of the similarity between the target playing track and the user video playing track and the user viewing probability needs to be calculated to obtain a recommended recommendation.
  • the probability of playing a video is the probability of playing a video.
  • the similarity between the target playback track and the user video playback track is higher, the more similar the similarity between the two tracks is, the more likely the user is to watch the video to be played after the currently playing video in the target playing track with high similarity, so
  • the recommendation value of the to-be-played video is larger; the smaller the similarity between the target playback track and the user's video playback track, the smaller the similarity between the two tracks, and the less likely the user is to view the target playback track with low similarity.
  • the viewing probability of the user is also related to the similarity between the target playing track and the user video playing track, so the product of the similarity and the user viewing probability needs to be calculated, thereby more accurately calculating the probability of recommending the video to be played.
  • the target playback trajectory is XYAB and the video playback trajectory is XYA
  • the similarity between the target playback trajectory and the user video playback trajectory is 0.75
  • the server has calculated the video.
  • the viewing probability when the video B to be played is recommended during playback is 0.3, and the product of the similarity 0.75 and the viewing probability of 0.3 is calculated, and the probability of recommending the video B to be played is 0.225.
  • the similarity between each target playback track and the user video playback track needs to be calculated separately, and the viewing is recommended according to each similarity and the video to be played.
  • the server determines in the database that there are two target playback tracks having the currently played video and the similarity with the video playing track is greater than a preset threshold.
  • the two target playing tracks are XYAB and EAB, respectively, and the server determines the two target playing.
  • the videos to be played after the video A in the track are B and B, respectively. Assuming that the server calculates the viewing probability of 0.3 when the video B to be played is recommended during video playback, the similarity between the two target playback trajectories and the video playback trajectory is calculated separately.
  • the similarity between the target play track XYAB and the user U video play track XYA is 0.75
  • the similarity between the target play track EAB and the user U video play track XYA is 0.2
  • the probability of calculating the recommended video B to be played is 0.75.
  • Step S25 Determine a to-be-played video whose viewing probability is greater than the first threshold as a recommended video.
  • Step S26 Send the recommended video to the terminal and display it in the video recommendation list.
  • FIG. 3 is a schematic diagram of a video recommendation apparatus according to an embodiment of the present invention.
  • the apparatus includes a receiving module 11, a first determining module 12, a second determining module 13, a calculating module 14, a third determining module 15, and a transmitting module 16, wherein:
  • the receiving module 11 is configured to receive a user video playing track and a currently playing video.
  • the first determining module 12 is configured to determine, in the database, a target playing track whose similarity with the user video playing track is greater than a preset threshold, where the target playing track includes the currently playing video.
  • the second determining module 13 is configured to determine a to-be-played video after the currently playing video in the target playing track.
  • the calculating module 14 is configured to calculate a viewing probability of the user when the video to be played is recommended.
  • the third determining module 15 is configured to determine a to-be-played video whose viewing probability is greater than the first threshold as the recommended video.
  • the sending module 16 is configured to send the recommended video to the terminal and display it in the video recommendation list.
  • the calculation module 14 may include the following sub-module: a first acquisition sub-module, configured to acquire the number of times the user views when the video to be played is recommended.
  • the second obtaining sub-module is configured to obtain the number of times the video to be played is recommended.
  • the first calculation sub-module is configured to calculate a quotient of the number of times the user views the recommended video to be played and the number of times the video to be played is recommended, and obtain the viewing probability of the user when the video to be played is recommended.
  • calculation module 14 may further include the following sub-module: a second calculation sub-module, configured to calculate a user viewing probability when the to-be-played video is recommended during the playing of the currently played video; It is used to calculate the product of the similarity and the user's viewing probability, and the probability of recommending the video to be played is obtained.
  • a second calculation sub-module configured to calculate a user viewing probability when the to-be-played video is recommended during the playing of the currently played video. It is used to calculate the product of the similarity and the user's viewing probability, and the probability of recommending the video to be played is obtained.
  • the first determining module 12 may further include the following sub-module: a third acquiring sub-module, configured to acquire a first video set of the target playing track and a second video set of the user video playing track; a module, configured to determine an intersection of the first video set and the second video set as a third video set, and a second determining submodule, configured to determine a union of the first video set and the second video set as a fourth video set And a fourth calculating sub-module, configured to calculate a quotient of the number of elements of the third video set and the number of elements of the fourth video set, to obtain a similarity between the target playing track and the user video playing track.
  • a third acquiring sub-module configured to acquire a first video set of the target playing track and a second video set of the user video playing track
  • a module configured to determine an intersection of the first video set and the second video set as a third video set
  • a second determining submodule configured to determine a union of the first video set and the second
  • the first determining module 12 may further include the following submodule: a third determining submodule for determining a play track having a currently played video in a database; and a fifth calculating submodule for calculating a play track.
  • a quantity determining a sub-module configured to determine whether the number of playing tracks is greater than a second threshold; if the number of playing tracks is greater than a second threshold, determining, in the playing track, that the similarity with the user video playing track is greater than a third threshold Playing a track; if the number of playing tracks is less than or equal to the second threshold, determining that the playing track is the target playing track.
  • the embodiment of the present invention further provides a server, including the video recommendation apparatus provided in the embodiment shown in FIG.
  • the embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium can store a program, and when the program is executed, each implementation manner of the video recommendation method provided by the embodiment shown in FIG. 1 and FIG. 2 can be implemented. Some or all of the steps.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Graphics (AREA)
  • Computing Systems (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

一种视频推荐方法及装置，该方法包括：接收用户视频播放轨迹和当前播放视频；在数据库中确定与所述用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，所述目标播放轨迹包括所述当前播放视频；确定所述目标播放轨迹中所述当前播放视频后的待播放视频；计算所述待播放视频被推荐时用户的观看概率；将所述观看概率大于第一阈值的所述待播放视频确定为推荐视频；将所述推荐视频发送至终端。由于用户的观看行为具有相似性，所以该方法通过分析其他用户的观看行为，并根据概率来确定出用户最有可能观看的视频。因此，可以给用户推送感兴趣的视频，以提高用户的观影体验。

Description

一种视频推荐方法及装置
本申请要求于2016年3月25日提交中国专利局、申请号为201610178391.5、发明名称为“一种视频推荐方法及装置”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及通信技术领域，尤其涉及视频推荐方法及装置。
背景技术
目前，随着网络技术的不断更新，在视频网站上观看视频成为很多用户的首选。为了更好的将优质视频推送给用户观看，视频网站通常会将热门视频放在推荐列表中，但是推荐列表中的热门视频仅仅是某段时间内观看的人数多而已，并不一定是用户真正感兴趣的视频，很多用户当下并不会观看推荐列表中的热门视频。也就是说推荐列表中的热门视频被用户观看的概率较低，推送技术无法达到较好的推送效果。
因此，如何推荐给用户感兴趣的视频，成为目前亟需解决的技术问题。
发明内容
本发明实施例提供了一种视频推荐方法及装置，以推荐给用户感兴趣的视频，从而提高用户的观影体验。
根据本发明实施例的第一方面，提供一种视频推荐方法，包括：
接收用户视频播放轨迹和当前播放视频；
在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，目标播放轨迹包括当前播放视频；
确定目标播放轨迹中当前播放视频后的待播放视频；
计算待播放视频被推荐时用户的观看概率；
将观看概率大于第一阈值的待播放视频确定为推荐视频；
将推荐视频发送至终端。
可选的，计算待播放视频被推荐时用户的观看概率的步骤包括：
获取待播放视频被推荐时用户观看的次数；
获取待播放视频被推荐的次数；
计算待播放视频被推荐时用户观看的次数与待播放视频被推荐的次数之商，得到待播放视频被推荐时用户的观看概率。
可选的，计算待播放视频被推荐时用户的观看概率的步骤包括：
计算在当前播放视频的播放过程中，待播放视频被推荐时的用户观看概率；
计算相似度和用户观看概率的乘积，得出推荐待播放视频的概率。
可选的，与用户视频播放轨迹的相似度的计算方法包括：
获取目标播放轨迹的第一视频集合和用户视频播放轨迹的第二视频集合；
将第一视频集合与第二视频集合的交集确定为第三视频集合；
将第一视频集合与第二视频集合的并集确定为第四视频集合；
计算第三视频集合的元素数量与第四视频集合的元素数量之商，得到目标播放轨迹与视频播放轨迹的相似度。
可选的，在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹的步骤包括：
在数据库中确定具有当前播放视频的播放轨迹；
计算播放轨迹的数量；
判断播放轨迹的数量是否大于第二阈值；
若播放轨迹的数量大于第二阈值，则在播放轨迹中确定与用户视频播放轨迹的相似度大于第三阈值的为目标播放轨迹；
若播放轨迹的数量小于或等于第二阈值，则确定播放轨迹为目标播放轨迹。
根据本发明实施例的第二方面，提供一种视频推荐装置，包括：
接收模块：接收用户视频播放轨迹和当前播放视频；
第一确定模块，用于在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，目标播放轨迹包括当前播放视频；
第二确定模块，用于确定目标播放轨迹中当前播放视频后的待播放视频；
计算模块，用于计算待播放视频被推荐时用户的观看概率；
第三确定模块，用于将观看概率大于第一阈值的待播放视频确定为推荐视频；
发送模块，用于将推荐视频发送至终端。
可选的，计算模块包括：
第一获取子模块，用于获取待播放视频被推荐时用户观看的次数；
第二获取子模块，用于获取待播放视频被推荐的次数；
第一计算子模块，用于计算待播放视频被推荐时用户观看的次数与待播放视频被推荐的次数之商，得到待播放视频被推荐时用户的观看概率。
可选的，计算模块包括：
第二计算子模块，用于计算在当前播放视频的播放过程中待播放视频被推荐时的用户观看概率；
第三计算子模块，用于计算相似度和用户观看概率的乘积，得出推荐待播放视频的概率。
可选的，第一确定模块包括：
第三获取子模块，用于获取目标播放轨迹的第一视频集合和用户视频播放轨迹的第二视频集合；
第一确定子模块，用于将第一视频集合与第二视频集合的交集确定为第三视频集合；
第二确定子模块，用于将第一视频集合与第二视频集合的并集确定为第四视频集合；
第四计算子模块，用于计算第三视频集合的元素数量与第四视频集合的元素数量之商，得到目标播放轨迹与用户视频播放轨迹的相似度。
可选的，第一确定模块包括：
第三确定子模块，用于在数据库中确定具有当前播放视频的播放轨迹；
第五计算子模块，用于计算播放轨迹的数量；
判断子模块，用于判断播放轨迹的数量是否大于第二阈值；
若播放轨迹的数量大于第二阈值，则在播放轨迹中确定与用户视频播放轨迹的相似度大于第三阈值的为目标播放轨迹；
若播放轨迹的数量小于或等于第二阈值，则确定播放轨迹为目标播放轨迹。
根据本发明实施例的第三方面，提供一种服务器，包括：第二方面所述的视频推荐装置。
根据本发明实施例的第四方面，本发明实施例提供一种计算机存储介质，其中，该计算机存储介质可存储有程序，该程序执行时可实现本发明第一方面提供的视频推荐方法的各实现方式中的部分或全部步骤。
与现有技术相比，本实施例提供的技术方案具有以下优点和特点：
在本发明提供的方案中，在接收到终端发送的用户视频播放轨迹和当前播放视频时，说明使用终端的用户当前正在观看视频，为了能够为该用户推送感兴趣的视频，需要先在数据库中确定具有该视频的目标播放轨迹，同时通过预设阈值去除一些相似度较低的播放轨迹，以便于得知其他用户观看该视频以后可能会看的待播放视频。再将被推荐时用户的观看概率大于第一阈值的待播放视频确定为推荐视频，最后将用户最有可能观看的推荐视频发送给终端。由于用户的观看行为具有相似性，所以本发明通过分析其他用户的观看行为，并根据用户的观看概率来确定出用户最有可能观看的视频。因此，本发明提供的方案可以给用户推送感兴趣的视频，以提高用户的观影体验。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本发明实施例提供的一种视频推荐方法的流程图。
图2为本发明实施例提供的另一种视频推荐方法的流程图。
图3为本发明实施例提供的一种视频推荐装置的示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整的描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
图1为本发明实施例提供的一种视频推荐方法的流程图。本发明提供的视频推荐方法可以给用户推送感兴趣的视频，以提高用户的观影体验。该方法包括以下步骤。
步骤S11、接收用户视频播放轨迹和当前播放的视频，在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，目标播放轨迹包括当前播放视频。
本实施例提供的方法可以应用于服务器内，服务器可以为电脑等设备，终端可以为智能电视、智能手机、平板电脑和家用电脑等设备。
其中，用户视频播放轨迹为用户在一定时间内观看的视频顺序。例如，用户在最近5天内观看了A电影、B电影、C电影和D电影，那么该用户的视频播放轨迹为A电影、B电影、C电影和D电影。
数据库设置在服务器端，数据库用于存储所有用户的播放轨迹。例如，用户X的播放轨迹为A电影、B电影、C电影和D电影，用户Y的播放轨迹为A电影、B电影、E电影和F电影，用户Z的播放轨迹为A电影、B电影、H电影和I电影。
在用户使用终端观看视频时，终端会将该用户的视频播放轨迹和当前播放的视频发送给服务器，以便于服务器经过分析后推送合适的视频。在接收到终端发送的视频播放轨迹和当前播放的视频时，服务器会先在数据库中确定出具有用户当前播放的视频且与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，以便于通过这些目标播放轨迹了解到其他用户在看完用户当前播放的视频以后，会看哪些视频，同时通过预设阈值，能够去除一些相似度较低的播放轨迹。
例如，假设用户U在最近2天内使用终端观看了视频X和视频Y，在用户U使用终端观看视频A时，终端会先确定出用户U的视频播放轨迹为视频X、视频Y和视频A，以及当前播放的视频A，然后终端会将该用户U的视频播放轨迹XYA和当前播放的视频A发送给服务器，以便于服务器经过分析后推送合适的视频。 在服务器接收到终端发送的视频播放轨迹XYA和当前播放的视频A时，服务器会先在数据库中确定出具有用户当前播放的视频A且与视频播放轨迹XYA的相似度大于预设阈值的目标播放轨迹，以便于通过这些目标播放轨迹了解到其他用户在看完用户当前播放的视频以后，会看哪些视频，并去除一些相似度较低的播放轨迹。
步骤S11中提到的关于目标播放轨迹与用户视频播放轨迹的相似度的计算方法具有很多方式，下面提供一种方式。
首先，获取目标播放轨迹的第一视频集合和用户视频播放轨迹的第二视频集合。
然后，将第一视频集合与第二视频集合的交集确定为第三视频集合。
其次，将第一视频集合与第二视频集合的并集确定为第四视频集合。
最后，计算第三视频集合的元素数量与第四视频集合的元素数量之商，得到目标播放轨迹与用户视频播放轨迹的相似度。
例如，假设目标播放轨迹为XYAB，目标播放轨迹的第一视频集合即为XYAB，用户视频播放轨迹为XYA，用户视频播放轨迹的第二视频集合即为XYA。首先，获取目标播放轨迹的第一视频集合XYAB和用户视频播放轨迹的第二视频集合XYA。然后，将第一视频集合XYAB与第二视频集合XYA的交集确定为第三视频集合XYA。其次，将第一视频集合XYAB与第二视频集合XYA的并集确定为第四视频集合XYAB。最后，计算第三视频集合XYA的元素数量(3个)与第四视频集合XYAB的元素数量(4个)之商，得到播放轨迹与视频播放轨迹的相似度0.75。目标播放轨迹与用户视频播放轨迹的相似度越高，说明这两个轨迹的相似性越大；目标播放轨迹与用户视频播放轨迹的相似度越低，说明这两个轨迹的相似性越小。
在本发明的另一些实施例中，步骤S11中在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹的步骤还可以包括以下三个子步骤：步骤A、在数据库中确定具有当前播放视频的播放轨迹；步骤B、计算上述播放轨迹的数量；步骤C、在上述播放轨迹的数量大于第二阈值时，在这些播放轨迹中确定与用户视频播放轨迹的相似度大于第三阈值的目标播放轨迹。其中，如果在数据库中确定出具有用户当前播放视频的播放轨迹的数量大于第二阈值，说明具有视频的播放轨迹的样本有点过多，所以服务器会在播放轨迹中确定与视频播放 轨迹的相似度大于第三阈值的目标播放轨迹，以便于删除一些与视频播放轨迹的相似度较小的播放轨迹。
例如，假设在服务器中预先将第二阈值设定为3个，将第三阈值设定为0.75，假设服务器在数据库中确定出具有当前播放视频A的播放轨迹分别为第一播放轨迹、第二播放轨迹、第三播放轨迹、第四播放轨迹和第五播放轨迹，可以计算出具有当前播放视频A的播放轨迹的数量为5个，由于上述播放轨迹的数量(5个)大于第二阈值(3个)，所以服务器会在上述5个播放轨迹中确定与用户视频播放轨迹的相似度大于第三阈值(0.75)的目标播放轨迹。假设服务器经过计算得出第一播放轨迹与用户视频播放轨迹的相似度为0.8，第二播放轨迹与用户视频播放轨迹的相似度为0.78，第三播放轨迹与用户视频播放轨迹的相似度为0.6，第四播放轨迹与用户视频播放轨迹的相似度为0.5，第五播放轨迹与用户视频播放轨迹的相似度为0.4，所以服务器会在上述5个播放轨迹中确定出与用户视频播放轨迹的相似度大于第三阈值(0.75)的目标播放轨迹为第一播放轨迹和第二播放轨迹，从而便去除掉与用户视频播放轨迹的相似度较小的第三播放轨迹、第四播放轨迹和第五播放轨迹。
步骤S12、确定目标播放轨迹中当前播放视频后的待播放视频。
其中，服务器在数据库中确定具有用户当前播放视频的目标播放轨迹以后，需要再确定目标播放轨迹中用户当前播放视频后的待播放视频。其中，在目标播放轨迹中，用户当前播放的视频后面的那个视频被称为待播放视频。
例如，假设服务器已经在数据库中确定出具有用户当前播放的视频A的目标播放轨迹有3个，这3个目标播放轨迹分别为XYAB、EAB和XAC。此时，服务器需要分别确定出这3个目标播放轨迹中视频A后的待播放视频，这3个目标播放轨迹中视频A后的待播放视频分别为B、B和C。
步骤S13、计算待播放视频被推荐时用户的观看概率。
其中，在服务器确定待播放视频以后，需要计算出待播放视频被推荐时用户的观看概率。
计算待播放视频被推荐时用户的观看概率的方式有很多种，下面提供一种详细的方式：首先，获取待播放视频被推荐时用户观看的次数。然后，获取待播放视频被推荐的次数。最后，计算待播放视频被推荐时用户观看的次数与待播放视 频被推荐的次数之商，得到待播放视频被推荐时用户的观看概率。
在上述计算待播放视频被推荐时用户观看的概率的方法中，服务器会预先统计所有用户的观看视频的情况，服务器在推荐待播放视频时，会记录下推荐该待播放视频的次数。如果服务器检测到终端在被推荐该待播放视频的情况下，观看了该待播放视频，说明本次推荐成功了，也会记录下推荐该待播放视频成功的次数。所以服务器可以根据之前记录的次数，来计算待播放视频被推荐时用户观看的次数与待播放视频被推荐的次数之商，以得到待播放视频被推荐时用户的观看概率。
例如，假设服务器向50个终端推荐了待播放视频B的次数为50次，服务器检测到在这50个终端中，在被推荐该待播放视频的情况下且观看了该待播放视频的次数为20次，所以待播放视频被推荐时用户的观看概率为20/50＝0.4。
步骤S14、将观看概率大于第一阈值的待播放视频确定为推荐视频。
其中，在服务器计算待播放视频被推荐时用户的观看概率后，服务器会将观看概率大于第一阈值的待播放视频确定为推荐视频。
例如，假设服务器已经在数据库中确定出具有用户当前播放的视频A的目标播放轨迹有3个，这3个目标播放轨迹分别为XYAB、EAB和XAC。此时，服务器需要分别确定出这3个目标播放轨迹中视频A后的待播放视频，这3个目标播放轨迹中视频A后的待播放视频分别为B、B和C。由于存在两个待播放视频均为B，所以最后得到的待播放视频为B和C。假设服务器计算出待播放视频B被推荐时用户观看的概率为0.6，并计算出待播放视频C被推荐时用户观看的概率为0.3，预先设定的第一阈值为0.5，所以服务器会将概率大于第一阈值0.5的待播放视频B确定为推荐视频。
步骤S15、将推荐视频发送至终端，并显示在视频推荐列表中。
在图1所示的实施例中，服务器接收到终端发送的用户视频播放轨迹和当前播放的视频时，说明使用终端的用户当前正在观看视频，为了能够为该用户推送感兴趣的视频，服务器需要先在数据库中确定具有该视频的目标播放轨迹，同时通过预设阈值去除一些相似度较低的播放轨迹，以便于得知其他用户观看该视频以后可能会看的待播放视频。服务器再将被推荐时用户的观看概率大于第一阈值的待播放视频确定为推荐视频，最后将用户最有可能观看的推荐视频发送给终 端。由于用户的观看行为具有相似性，所以本发明通过分析其他用户的观看行为，并根据概率来确定出用户最有可能观看的视频。因此，本发明实施例提供的方案可以给用户推送感兴趣的视频，以提高用户的观影体验。
图2为本发明实施例提供的另一种视频推荐方法的流程图。图2为基于图1的一个可选的实施例，在图2所示的实施例中，与图1所示的实施例相同的部分可以参见图1所示的实施例中介绍和解释。图2所示的方法包括以下步骤。
步骤S21、接收用户视频播放轨迹和当前播放的视频，在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，目标播放轨迹包括当前播放视频。
步骤S22、确定目标播放轨迹中当前播放视频后的待播放视频。
步骤S23、计算在当前播放视频播放过程中，待播放视频被推荐时的用户观看概率。
其中，在服务器确定出待播放视频以后，服务器需要计算在当前播放视频播放过程中待播放视频被推荐时的用户观看概率。与步骤S13相比，将待播放视频被推荐的时间限定在用户在观看当前播放视频的过程中，即计算的是用户在观看当前播放视频时，点击观看待播放视频的概率，而目标播放轨迹中待播放视频也是在当前播放视频之后出现，两者是相互匹配的，此时计算出来的用户观看概率更加精准。
步骤S24、计算相似度和用户观看概率的乘积，得出推荐待播放视频的概率。
其中，在服务器计算出其他用户在视频播放过程中待播放视频被推荐时的用户观看概率以后，需要计算目标播放轨迹与用户视频播放轨迹的相似度和用户观看概率的乘积，以得出推荐待播放视频的概率。因为目标播放轨迹与用户视频播放轨迹的相似度越高，说明这两个轨迹的相似性越大，用户就越有可能观看相似度高的目标播放轨迹中当前播放视频后的待播放视频，所以该待播放视频的推荐价值就越大；目标播放轨迹与用户视频播放轨迹的相似度越小，说明这两个轨迹的相似性越小，用户就越没有可能观看相似度低的目标播放轨迹中当前播放视频后的待播放视频，所以该待播放视频的推荐价值就越小。因此，决定用户观看待播放视频的因素不仅包括其他用户在视频播放过程中待播放视频被推荐时的用 户观看概率，还与目标播放轨迹与用户视频播放轨迹的相似度有关，所以需要计算相似度和用户观看概率的乘积，从而更加精准的计算出推荐待播放视频的概率。
例如，假设目标播放轨迹为XYAB，视频播放轨迹为XYA，根据图1所示实施例中相似度的计算方法，可知目标播放轨迹与用户视频播放轨迹的相似度0.75，且服务器已经计算出在视频播放过程中待播放视频B被推荐时的观看概率为0.3，计算相似度0.75与观看概率0.3之积，得到推荐待播放视频B的概率为0.225。
如果至少存在两个目标播放轨迹的待播放视频是相同的，那么需要分别计算出每个目标播放轨迹与用户视频播放轨迹的相似度，并根据每个相似度和待播放视频被推荐时的观看概率乘积之和，得到推荐待播放视频的概率。
例如，假设用户U的视频播放轨迹XYA，当前播放的视频A。服务器在数据库中确定具有当前播放的视频且与视频播放轨迹的相似度大于预设阈值的目标播放轨迹有两个，这两个目标播放轨迹分别为XYAB和EAB，服务器确定出这2个目标播放轨迹中视频A后的待播放视频分别为B和B。假设服务器计算出在视频播放过程中待播放视频B被推荐时的观看概率0.3，那么分别计算这两个目标播放轨迹与视频播放轨迹的相似度。其中，经过计算，目标播放轨迹XYAB与用户U视频播放轨迹XYA的相似度为0.75，目标播放轨迹EAB与用户U视频播放轨迹XYA的相似度为0.2，最后计算推荐待播放视频B的概率为0.75×0.3+0.2×0.3＝0.225+0.06＝0.285。
步骤S25、将观看概率大于第一阈值的待播放视频确定为推荐视频。
步骤S26、将推荐视频发送至终端，并显示在视频推荐列表中。
图3为本发明实施例提供的一种视频推荐装置的示意图。参照图3，该装置包括接收模块11、第一确定模块12、第二确定模块13、计算模块14、第三确定模块15和发送模块16，其中：
接收模块11，用于接收用户视频播放轨迹和当前播放视频。
第一确定模块12，用于在数据库中确定与用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，目标播放轨迹包括当前播放视频。
第二确定模块13，用于确定目标播放轨迹中当前播放视频后的待播放视频。
计算模块14，用于计算待播放视频被推荐时用户的观看概率。
第三确定模块15，用于将观看概率大于第一阈值的待播放视频确定为推荐视频。
发送模块16，用于将推荐视频发送给终端，并显示在视频推荐列表中。
另外，可选的，计算模块14可以包括以下子模块：第一获取子模块，用于获取待播放视频被推荐时用户观看的次数。第二获取子模块，用于获取待播放视频被推荐的次数。第一计算子模块，用于计算待播放视频被推荐时用户观看的次数与待播放视频被推荐的次数之商，得到待播放视频被推荐时用户的观看概率。
另外，可选的，计算模块14还可以包括以下子模块：第二计算子模块，用于计算在当前播放视频的播放过程中待播放视频被推荐时的用户观看概率；第三计算子模块，用于计算相似度和用户观看概率的乘积，得出推荐待播放视频的概率。
另外，可选的，第一确定模块12还可以包括以下子模块：第三获取子模块，用于获取目标播放轨迹的第一视频集合和用户视频播放轨迹的第二视频集合；第一确定子模块，用于将第一视频集合与第二视频集合的交集确定为第三视频集合；第二确定子模块，用于将第一视频集合与第二视频集合的并集确定为第四视频集合；第四计算子模块，用于计算第三视频集合的元素数量与第四视频集合的元素数量之商，得到目标播放轨迹与用户视频播放轨迹的相似度。
另外，可选的，第一确定模块12还可以包括以下子模块：第三确定子模块，用于在数据库中确定具有当前播放视频的播放轨迹；第五计算子模块，用于计算播放轨迹的数量；判断子模块，用于判断播放轨迹的数量是否大于第二阈值；若播放轨迹的数量大于第二阈值，则在播放轨迹中确定与用户视频播放轨迹的相似度大于第三阈值的为目标播放轨迹；若播放轨迹的数量小于或等于第二阈值，则确定播放轨迹为目标播放轨迹。
关于上述实施例中的装置，其中各个模块执行操作的具体方式已经在有关该方法的实施例中进行了详细描述，此处将不做详细阐述说明。
需要说明的是，本发明实施例还提供了一种服务器，包括图3所示实施例提供的视频推荐装置。
此外，本发明实施例还提供了一种计算机存储介质，其中，该计算机存储介质可存储有程序，该程序执行时可实现图1和图2所示实施例提供的视频推荐方法的各实现方式中的部分或全部步骤。
以上，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以权利要求的保护范围为准。

Claims (10)

  1. 一种视频推荐方法，其特征在于，包括：
    接收用户视频播放轨迹和当前播放视频；
    在数据库中确定与所述用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，所述目标播放轨迹包括所述当前播放视频；
    确定所述目标播放轨迹中所述当前播放视频后的待播放视频；
    计算所述待播放视频被推荐时用户的观看概率；
    将所述观看概率大于第一阈值的所述待播放视频确定为推荐视频；
    将所述推荐视频发送至终端。
  2. 根据权利要求1所述的视频推荐方法，其特征在于，所述计算所述待播放视频被推荐时用户的观看概率的步骤包括：
    获取所述待播放视频被推荐时用户观看的次数；
    获取所述待播放视频被推荐的次数；
    计算所述待播放视频被推荐时用户观看的次数与所述待播放视频被推荐的次数之商，得到所述待播放视频被推荐时用户的观看概率。
  3. 根据权利要求1所述的视频推荐方法，其特征在于，所述计算所述待播放视频被推荐时用户的观看概率的步骤包括：
    计算在所述当前播放视频的播放过程中，所述待播放视频被推荐时的用户观看概率；
    计算所述相似度和所述用户观看概率的乘积，得出推荐所述待播放视频的概率。
  4. 根据权利要求1所述的视频推荐方法，其特征在于，所述与所述用户视频播放轨迹的相似度的计算方法包括：
    获取所述目标播放轨迹的第一视频集合和所述用户视频播放轨迹的第二视频集合；
    将所述第一视频集合与所述第二视频集合的交集确定为第三视频集合；
    将所述第一视频集合与所述第二视频集合的并集确定为第四视频集合；
    计算所述第三视频集合的元素数量与所述第四视频集合的元素数量之商，得到所述目标播放轨迹与所述视频播放轨迹的相似度。
  5. 根据权利要求1所述的视频推荐方法，其特征在于，所述在数据库中确定与所述用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹的步骤包括：
    在数据库中确定具有所述当前播放视频的播放轨迹；
    计算所述播放轨迹的数量；
    判断所述播放轨迹的数量是否大于第二阈值；
    若所述播放轨迹的数量大于所述第二阈值，则在所述播放轨迹中确定与所述用户视频播放轨迹的相似度大于第三阈值的为目标播放轨迹；
    若所述播放轨迹的数量小于或等于所述第二阈值，则确定所述播放轨迹为目标播放轨迹。
  6. 一种视频推荐装置，其特征在于，包括：
    接收模块：接收用户视频播放轨迹和当前播放视频；
    第一确定模块，用于在数据库中确定与所述用户视频播放轨迹的相似度大于预设阈值的目标播放轨迹，其中，所述目标播放轨迹包括所述当前播放视频；
    第二确定模块，用于确定所述目标播放轨迹中所述当前播放视频后的待播放视频；
    计算模块，用于计算所述待播放视频被推荐时用户的观看概率；
    第三确定模块，用于将所述观看概率大于第一阈值的所述待播放视频确 定为推荐视频；
    发送模块，用于将所述推荐视频发送至终端。
  7. 根据权利要求6所述的视频推荐装置，其特征在于，所述计算模块包括：
    第一获取子模块，用于获取所述待播放视频被推荐时用户观看的次数；
    第二获取子模块，用于获取所述待播放视频被推荐的次数；
    第一计算子模块，用于计算所述待播放视频被推荐时用户观看的次数与所述待播放视频被推荐的次数之商，得到所述待播放视频被推荐时用户的观看概率。
  8. 根据权利要求6所述的视频推荐装置，其特征在于，所述计算模块包括：
    第二计算子模块，用于计算在所述当前播放视频的播放过程中所述待播放视频被推荐时的用户观看概率；
    第三计算子模块，用于计算所述相似度和所述用户观看概率的乘积，得出推荐所述待播放视频的概率。
  9. 根据权利要求6所述的视频推荐装置，其特征在于，所述第一确定模块包括：
    第三获取子模块，用于获取所述目标播放轨迹的第一视频集合和所述用户视频播放轨迹的第二视频集合；
    第一确定子模块，用于将所述第一视频集合与所述第二视频集合的交集确定为第三视频集合；
    第二确定子模块，用于将所述第一视频集合与所述第二视频集合的并集确定为第四视频集合；
    第四计算子模块，用于计算所述第三视频集合的元素数量与所述第四视 频集合的元素数量之商，得到所述目标播放轨迹与所述用户视频播放轨迹的相似度。
  10. 根据权利要求6所述的视频推荐装置，其特征在于，第一确定模块包括：
    第三确定子模块，用于在数据库中确定具有所述当前播放视频的播放轨迹；
    第五计算子模块，用于计算所述播放轨迹的数量；
    判断子模块，用于判断所述播放轨迹的数量是否大于第二阈值；
    若所述播放轨迹的数量大于所述第二阈值，则在所述播放轨迹中确定与所述用户视频播放轨迹的相似度大于第三阈值的为目标播放轨迹；
    若所述播放轨迹的数量小于或等于所述第二阈值，则确定所述播放轨迹为目标播放轨迹。
PCT/CN2016/088488 2016-03-25 2016-07-04 一种视频推荐方法及装置 WO2017161744A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610178391.5A CN105847984A (zh) 2016-03-25 2016-03-25 一种视频推荐方法及装置
CN201610178391.5 2016-03-25

Publications (1)

Publication Number Publication Date
WO2017161744A1 true WO2017161744A1 (zh) 2017-09-28

Family

ID=56583550

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/088488 WO2017161744A1 (zh) 2016-03-25 2016-07-04 一种视频推荐方法及装置

Country Status (2)

Country Link
CN (1) CN105847984A (zh)
WO (1) WO2017161744A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108495143A (zh) * 2018-03-30 2018-09-04 百度在线网络技术（北京）有限公司 视频推荐的方法和装置
CN113626698A (zh) * 2021-08-06 2021-11-09 北京奇艺世纪科技有限公司 视频推荐方法、装置、电子设备及可读存储介质

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107454442B (zh) * 2017-09-07 2021-02-05 阿里巴巴（中国）有限公司 一种推荐视频的方法和装置
CN107786895B (zh) * 2017-10-18 2019-09-17 北京奇艺世纪科技有限公司 一种播放页视频推荐的质量评估方法及装置
CN108153863B (zh) * 2017-12-25 2021-12-17 北京奇艺世纪科技有限公司 一种视频信息的表示方法及装置
CN108235045B (zh) * 2018-01-04 2019-09-10 武汉斗鱼网络科技有限公司 一种直播间推荐方法、电子设备及可读存储介质
CN108419100B (zh) * 2018-01-29 2020-10-02 山东云缦智能科技有限公司 一种用户电影播放行为相似度的获取方法及系统
CN108536814B (zh) * 2018-04-04 2022-06-21 武汉斗鱼网络科技有限公司 直播间推荐方法、计算机可读存储介质及电子设备

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102164308A (zh) * 2010-02-17 2011-08-24 索尼公司 信息处理装置、信息处理方法和程序
CN103648031A (zh) * 2013-11-15 2014-03-19 乐视致新电子科技（天津）有限公司 一种智能电视的节目推荐方法及装置
JP2014200007A (ja) * 2013-03-29 2014-10-23 ニフティ株式会社 推薦プログラム、推薦情報受信プログラム、推薦方法、推薦情報受信方法及び推薦装置
CN104123325A (zh) * 2013-04-28 2014-10-29 北京百度网讯科技有限公司 多媒体文件的推荐方法和推荐服务器
CN104935964A (zh) * 2015-06-02 2015-09-23 四川九天揽月文化传媒有限公司 一种智能电视节目分组筛选推送方法
CN105100165A (zh) * 2014-05-20 2015-11-25 深圳市腾讯计算机系统有限公司 网络服务推荐方法和装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102630052B (zh) * 2012-04-16 2014-10-15 上海交通大学 面向实时流的电视节目推荐系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102164308A (zh) * 2010-02-17 2011-08-24 索尼公司 信息处理装置、信息处理方法和程序
JP2014200007A (ja) * 2013-03-29 2014-10-23 ニフティ株式会社 推薦プログラム、推薦情報受信プログラム、推薦方法、推薦情報受信方法及び推薦装置
CN104123325A (zh) * 2013-04-28 2014-10-29 北京百度网讯科技有限公司 多媒体文件的推荐方法和推荐服务器
CN103648031A (zh) * 2013-11-15 2014-03-19 乐视致新电子科技（天津）有限公司 一种智能电视的节目推荐方法及装置
CN105100165A (zh) * 2014-05-20 2015-11-25 深圳市腾讯计算机系统有限公司 网络服务推荐方法和装置
CN104935964A (zh) * 2015-06-02 2015-09-23 四川九天揽月文化传媒有限公司 一种智能电视节目分组筛选推送方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108495143A (zh) * 2018-03-30 2018-09-04 百度在线网络技术（北京）有限公司 视频推荐的方法和装置
CN113626698A (zh) * 2021-08-06 2021-11-09 北京奇艺世纪科技有限公司 视频推荐方法、装置、电子设备及可读存储介质

Also Published As

Publication number Publication date
CN105847984A (zh) 2016-08-10

Similar Documents

Publication Publication Date Title
WO2017161744A1 (zh) 一种视频推荐方法及装置
US9602886B2 (en) Methods and systems for displaying contextually relevant information from a plurality of users in real-time regarding a media asset
US11750895B2 (en) Crowd-sourced program boundaries
US20160366463A1 (en) Information pushing method, terminal and server
US10880025B1 (en) Identification of concurrently broadcast time-based media
US10469918B1 (en) Expanded previously on segments
KR20170033360A (ko) 자막 및 더빙에 대한 선호도의 자동 검출
CN105049882A (zh) 一种视频推荐方法及装置
US20120042041A1 (en) Information processing apparatus, information processing system, information processing method, and program
WO2017107464A1 (zh) 一种视频跟播方法及装置
US10897658B1 (en) Techniques for annotating media content
CN104965874A (zh) 信息处理方法及装置
EP3346396A1 (en) Multimedia resource quality assessment method and apparatus
CN111182316B (zh) 媒体资源的流切换方法和装置、存储介质及电子装置
Salas et al. Subjective quality evaluations using crowdsourcing
US20220312079A1 (en) Systems and methods to provide adaptive play settings
JP5451545B2 (ja) ノイズ除去条件決定装置、ノイズ除去条件決定方法、及びプログラム
EP2902924A1 (en) Method for automatically selecting a real-time video stream among a plurality of available real-time video streams, and associated system
CN110546932A (zh) 使用媒体查看数据提高设备映射图准确度的系统和方法
US11902619B2 (en) Systems and methods for providing media content
CN105992065B (zh) 随选视讯社交互动方法和系统
KR101380963B1 (ko) 관련 정보 제공 시스템 및 제공 방법
CN117336558A (zh) 数据发送方法、装置、电子设备和存储介质
CN110929151A (zh) 用户推荐方法、电子设备和存储介质
CN105516279A (zh) 一种用户账户创建方法及装置

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16895094

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16895094

Country of ref document: EP

Kind code of ref document: A1