WO2017101328A1 - 智能电视展示播放内容的方法、装置及系统 - Google Patents

智能电视展示播放内容的方法、装置及系统 Download PDF

Info

Publication number
WO2017101328A1
WO2017101328A1 PCT/CN2016/088553 CN2016088553W WO2017101328A1 WO 2017101328 A1 WO2017101328 A1 WO 2017101328A1 CN 2016088553 W CN2016088553 W CN 2016088553W WO 2017101328 A1 WO2017101328 A1 WO 2017101328A1
Authority
WO
WIPO (PCT)
Prior art keywords
movie
user
viewing data
training
recommended
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2016/088553
Other languages
English (en)
French (fr)
Inventor
崔建勇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Le Holdings Beijing Co Ltd
Leshi Zhixin Electronic Technology Tianjin Co Ltd
Original Assignee
Le Holdings Beijing Co Ltd
Leshi Zhixin Electronic Technology Tianjin Co Ltd
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 Le Holdings Beijing Co Ltd, Leshi Zhixin Electronic Technology Tianjin Co Ltd filed Critical Le Holdings Beijing Co Ltd
Priority to US15/247,116 priority Critical patent/US20170169330A1/en
Publication of WO2017101328A1 publication Critical patent/WO2017101328A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

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]
    • H04N21/20—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
    • H04N21/4668—Learning process for intelligent management, e.g. learning user preferences for recommending movies for recommending content, e.g. movies

Definitions

  • the present invention relates to the field of television service technologies, and in particular, to a method, device and system for displaying content by a smart television.
  • Smart TV is an intelligent multimedia terminal that conforms to the trend of “high definition”, “networking” and “intelligence” of TV sets. It has the content of obtaining programs from various channels such as the Internet, video equipment and computers.
  • the integrated operator interface uses the features that consumers want most clearly on the big screen.
  • the smart TV can realize various application services such as network search, network television, video on demand (VOD), digital music, network news, network video telephone, and the like.
  • VOD video on demand
  • smart TV has a lot of program content for users to watch, and because of its characteristics of the Internet, program content can be searched and viewed throughout the network, so the fast and accurate positioning of program content is particularly important.
  • the inventors found that the existing smart TV has few film recommendations, and the current recommendation function is a mandatory recommendation by the manufacturer, which is contrary to the user-centered concept. What users want to see is that the movie they want to see is in the most conspicuous position, so that they can watch quickly and no longer have to worry about finding the source.
  • the embodiment of the invention provides a method, a device and a system for displaying a content on a smart TV, so as to solve the technical problem that the user searches for a large amount of video and affects the viewing experience of the user caused by the lack of the smart recommendation movie function on the television end in the prior art. .
  • An embodiment of the present invention provides a method for displaying content on a smart TV.
  • the method is applicable to a server, and includes:
  • the predicted recommended movie is fed back to the television to recommend to the user.
  • Another aspect of the present invention provides a method for displaying a content on a smart television.
  • the method is applicable to the television terminal, and the method mainly includes:
  • the viewing data is transmitted to the server.
  • Another aspect of the embodiments of the present invention provides a device for displaying content on a smart television, the device being suitable for being deployed on a server, including:
  • a receiving module configured to receive the viewing data viewed by the user transmitted by the television end
  • a prediction module configured to predict, according to the viewing data, a recommended movie that is highly correlated with the viewing data
  • a recommendation module configured to feed back the predicted recommended movie to the television terminal for recommendation to the user.
  • Another aspect of an embodiment of the present invention provides a smart television display device for playing content, the device Suitable for deployment on the TV side, including:
  • a data acquisition module configured to acquire viewing data of the user according to a movie that is viewed by the user each time
  • a transmitting module configured to transmit the viewing data to a server.
  • Another embodiment of the present invention further provides a system for displaying content of a smart television, including a server and a television end.
  • the server includes the above-mentioned device suitable for being deployed on a server to display content for playback
  • the television terminal includes the above suitable A device that does not display content on a smart TV on the TV side.
  • the method, device and system for displaying the content of the smart TV provided by the present invention can predict the type of the movie that the user likes to watch according to the viewing data of the user, and recommend the same type of movie to the screen display on the television end.
  • the user solves the technical problem that the user is inconvenient to search for the video in the prior art, and the operator forcibly recommends the video that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, facilitating the user to watch the movie, and enhancing the user view.
  • the technical effect of the shadow experience is a technical problem that the user is inconvenient to search for the video in the prior art, and the operator forcibly recommends the video that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, facilitating the user to watch the movie, and enhancing the user view.
  • FIG. 1 is a flowchart of a method for displaying a content on a smart television according to an embodiment of the present invention
  • FIG. 2 is a service diagram of a smart television display method for playing content according to an embodiment of the present invention Flowchart on the device;
  • FIG. 3 is a schematic diagram of a forward propagation algorithm in a method for displaying content of a smart television according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of an error back propagation algorithm in a method for displaying content of a smart television according to an embodiment of the present invention
  • FIG. 5 is a schematic diagram of a BP prediction network in a method for displaying content of a smart television according to an embodiment of the present invention
  • FIG. 6 is a schematic flowchart diagram of a method for displaying content played by a smart television according to an embodiment of the present invention
  • FIG. 7 is a schematic structural diagram of an apparatus for displaying content of a smart television on a television end according to an embodiment of the present invention.
  • FIG. 8 is a schematic structural diagram of an apparatus for displaying content of a smart television on a server according to an embodiment of the present invention.
  • FIG. 9 is a schematic structural diagram of a system for displaying content of a smart television according to an embodiment of the present invention.
  • FIG. 10 is a schematic structural diagram of an apparatus for displaying content of a smart television according to an embodiment of the present invention.
  • FIG. 11 is a schematic structural diagram of an apparatus for displaying content of a smart television according to an embodiment of the present invention.
  • this embodiment provides a method for displaying content on a smart TV.
  • the method is suitable for deployment on a television end, and includes:
  • the above viewing data includes each movie name and category label.
  • the viewing data of the movie that the user has watched can be obtained, including the movie name and the category label.
  • the TV terminal may periodically send the user's viewing data to the server, or the server may query the user's viewing data.
  • the viewing data of the above user is sent to the server, so that the server predicts the type of the movie that the user likes to watch according to the viewing data of the user, and feeds the corresponding movie type as a recommended movie to the television. Therefore, the foregoing solution further includes: receiving, by the server, a recommended movie that is highly correlated with a movie viewed by the user; and displaying the recommended movie to the user.
  • the television terminal only needs to send the viewing data of the user to the server, so that the server can perform prediction of the recommended movie, and since the prediction process is performed on the server, the burden on the television end is reduced. At the same time, it can meet the viewing needs of users.
  • the embodiment further provides a smart television display method for displaying content on a server, which mainly includes:
  • Movies that are highly correlated with the type of movie in the viewing data are usually the ones that the user likes, so this movie is suitable for recommending to the user as a recommended movie.
  • the accuracy of the predicted recommended movie in the present invention is different from the name and category of the movie that the user watches each time depending on the viewing data.
  • the recommendation is also based on the recommendation algorithm. By repeatedly inputting the movie name and category label viewed by the user as input samples into the recommendation algorithm for repeated training, the prediction degree of the recommendation algorithm will become more and more accurate, and The training result after multiple trainings is recommended to the user as the final predicted film with high correlation with the viewing data.
  • Neural network algorithm is a kind of algorithm widely used in machine learning algorithms.
  • BP algorithm Back Propagation
  • back propagation algorithm By calculating the error of the reverse transmission and accurately calculating the degree, it is called a back propagation algorithm. Both of these algorithms can be used as the recommendation algorithm in this embodiment.
  • the algorithm description is divided into two phases: the training phase and the prediction phase.
  • the input sample is a movie name and each movie name is marked with a plurality of tags, such as 10, and these 10 tags are trained as their characteristics.
  • the input result, that is, the target is the category label (category) of the movie.
  • category label category label
  • a certain type of movie is also marked with 10 labels as a feature for identifying it.
  • the sample is input to the input layer and then multiplied by the operation of the weight matrix with the corresponding weight.
  • the mathematical expression is as follows (1):
  • Aj is the output of the jth layer
  • W for the weight matrix
  • X for the input sample (data) matrix
  • the calculation method is as follows: the weight matrix is multiplied by the input sample matrix.
  • Xi, Wii are all elements of the corresponding matrix.
  • the weight matrix W directly affects the validity of the above formula 1, and thus the initial value for this W is obtained by the special dictionary training algorithm K-SVD.
  • K-SVD is a mathematical algorithm. A matrix can be decomposed to get the part we care about most, the feature.
  • the output of the neuron depends on the input and weight (weight) of the signal, and the activation function selects the nonlinear sigmoid for calculation.
  • the output of the activation function is as follows (2):
  • Oj is the output of the jth layer after the activation function.
  • the activation function uses sigmoid, which tends to 1 for larger input values (the structure after the sample is multiplied by the weight); when the input is 0, the active output is 0.5; when the negative value of the input value tends to be small, the activation The output is 0.
  • the curve of the activation function is smoothed before use, so that the gradient operation can be performed, and the accuracy of the operation is also improved.
  • the purpose of the training process is to get the desired output for a given input. Since the choice of the initial weight must not be optimal, there will be a residual between the target and the actual output, as shown in the following equation (3).
  • Ej is the difference between the actual output and the real value. We generally use this error as the target for optimization, and t j is the target output of the jth layer, that is, the real output value corresponding to the sample. As shown in Figure 3.
  • ⁇ w(k+1) and ⁇ w(k) represent the update of the weight values after the k+1th and kth iterations, respectively
  • ⁇ and ⁇ are two constant parameters, one for balanced update and one for The magnitude of the gradient is updated during the gradient descent method.
  • ⁇ usually chooses a number between 0 and 1, Is the gradient of the error function to the weight value.
  • the output depends on the output value of the activation function, and the activation function is the equation (7) obtained on the basis of the weight:
  • the derivative of the output relative to the weight value can be derived, as follows:
  • the core is carried out by the stochastic gradient descent method.
  • the process of backpropagation is the key point of the BP network. From the output layer to the input layer, the gradient is used layer by layer to obtain the current optimal value. Until the weights of the network are all updated, the input weight matrix and the output weight matrix are updated, and an iteration is completed, and the algorithm is iterated until convergence. As shown in Figure 4.
  • the input sample is the movie name and type tag that the user clicks
  • the output training result is the type tag of the predicted user's favorite movie.
  • Xi is an input sample
  • Yi is an intermediate node
  • Zi is the last output.
  • Xi is equivalent to the name of the movie that the user clicks, and outputs the type of movie that the predicted user is interested in.
  • the user's click on the movie record will be recorded and uploaded to the server for prediction to get the user's preference type. Then, when the system makes the recommendation, the preference will be recorded, and the server will recommend the type of the user's preference. What is reflected in the user's TV is that the same type of video will be recommended on the search page.
  • the foregoing 202 implementation method is as follows:
  • a movie whose matching degree meets the preset recommendation standard value is used as a recommended movie having a high degree of relevance to the movie data.
  • the method for displaying the content of the smart TV according to the embodiment of the present invention can predict the type of the movie that the user likes to watch according to the viewing data of the user, and recommend the same type of movie to the user through the screen display on the television end.
  • the invention solves the technical problem that the user is inconvenient to search for a video in the prior art, and the operator forcibly recommends a movie that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, facilitating the user to watch the movie, and enhancing the user's viewing experience.
  • the embodiment specifically provides a smart television display playing content, including:
  • the television end acquires the viewing data of the user according to the video that the user views each time, and transmits the viewing data to the server.
  • the server receives the viewing data viewed by the user transmitted by the television end;
  • the user has an account, and can obtain the viewing data of the user according to the account number.
  • the viewing data of the clicked video is transmitted to the network periodically when the user is idle, and the server collects the data through the network.
  • the type tag is taken out, and the lower tag is a tag that is pre-marked to the movie.
  • three labels "Fantasy”, “Fantasy” and “Comedy” are marked beforehand, then these three will be used as sample features for prediction.
  • the server trains the viewing data according to a recommendation algorithm.
  • the movie name and the category label of the movie are input as an input sample to the recommendation algorithm to obtain an output result; wherein, a process of obtaining an output result is used as a training process;
  • the corresponding output result is used as a training result, wherein the training result includes a category label.
  • the recommendation algorithm learning and training the obtained viewing data, inputting the name and type label of the movie, and associating the movie name and type as one-to-one correspondence training data.
  • the above obtained image data is input into the recommendation algorithm as a sample. There is an output when inputting a sample.
  • the process of inputting and outputting each time is a training, and the output result/training result includes: the type of the movie and the score.
  • Each sample input will have an output, and multiple sample inputs will have multiple outputs. Multiple training can improve the accuracy of the recommended algorithm.
  • each output is a calculated value, and this value is made to be different from the preset target value to obtain an error.
  • the error is optimized by the error back propagation process described above, so that the error is reduced to Set a smaller predictive standard threshold. At this point, it is considered that the precise network is obtained, that is, the input sample movie and a prediction relationship to which it belongs. Since the training process is carried out in the server, there is no pressure on the user's TV.
  • the type tag learned according to the user's viewing data is matched with the classification of the target movie library. Since there may be multiple types of the obtained type tags, and not necessarily each of them is matched, the score is matched according to the matching result. As a correlation with a movie that scores above the default recommended standard value High video and the top n most relevant videos as recommended videos.
  • the movie of the target movie library needs to be classified first. That is, the type label of the movie in the target movie library is marked on the server side in advance, and a relatively classified network is obtained according to the type label classification.
  • the server feeds back the predicted recommended movie to the television terminal for recommendation to the user.
  • the television end receives the recommended movie that is highly correlated with the movie viewed by the user, and displays the recommended movie to the user.
  • the accuracy of the prediction can be improved by training the recommendation algorithm.
  • the method, device and system for displaying the content of the smart TV according to the embodiment of the present invention can predict the type of the movie that the user likes to watch according to the viewing data of the user, and pass the same type of movie on the television end.
  • the screen display is further recommended to the user, which solves the technical problem that the user is inconvenient to search for the video in the prior art, and the operator forces to recommend the movie that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, and facilitating the user view. Shadow, enhance the technical effect of the user's viewing experience.
  • the embodiment provides a device for displaying content on a smart television.
  • the apparatus includes:
  • the data acquisition module 41 is configured to acquire the viewing data of the user according to the movie that the user views each time;
  • the transmitting module 42 is configured to transmit the viewing data to the server.
  • the device further includes:
  • a receiving module configured to receive, by the server, a recommended movie that is highly correlated with a movie viewed by the user;
  • a presentation module for presenting the recommended movie to the user.
  • the device provided in Figure 7 of the present embodiment is suitable for installation on a television or a television or set top box.
  • the embodiment provides a device for displaying content on a smart television, as shown in FIG.
  • the receiving module 51 is configured to receive the viewing data viewed by the user transmitted by the television end;
  • a prediction module 52 configured to predict, according to the viewing data, a recommended movie that is highly correlated with the viewing data
  • the recommendation module 53 is configured to feed back the predicted recommended movie to the television terminal for recommendation to the user.
  • the prediction module 52 includes:
  • a training unit configured to train the viewing data according to a recommendation algorithm
  • a result obtaining unit configured to obtain a category label of the recommended movie with high correlation according to the training result
  • a matching unit configured to match the obtained category label with a preset category label in the movie library, and use the movie whose matching degree meets the preset recommendation standard value as the recommended movie with high correlation with the movie data.
  • the movie data includes a name of the movie viewed by the user each time and a category label of the movie;
  • the training unit is specifically configured to input the movie name and the category label of the movie as an input sample into the recommendation algorithm for each movie title and the category label of the movie.
  • An output result wherein, the process of obtaining an output result is used as a training process; and when the output result obtained after the plurality of training processes meets a preset prediction standard value, the corresponding output result is used as a training result, wherein
  • the training result includes a category tag.
  • the device for displaying the content of the smart TV can predict the type of the movie that the user likes to watch according to the viewing data of the user, and recommend the same type of movie through the screen display on the television end.
  • the user solves the technical problem that the user searches for the video inconveniently in the prior art, and the operator forcibly recommends the movie that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, facilitating the user to watch the movie, and enhancing the user.
  • the technical effect of the viewing experience is a technical problem that the user searches for the video inconveniently in the prior art, and the operator forcibly recommends the movie that the non-user wants to watch, thereby improving the accuracy of the movie search and recommendation, facilitating the user to watch the movie, and enhancing the user.
  • the device provided in FIG. 8 of the embodiment is suitable for being installed on a server or a server.
  • FIG. 7 and FIG. 8 provided by the embodiment of the present invention may perform the foregoing method embodiments, and the specific implementation principles and technical effects thereof may be referred to the foregoing method embodiments, and details are not described herein again.
  • the present embodiment continues to provide a system for displaying content of a smart television, including a server 60 and a television terminal 70.
  • the server 60 includes the apparatus in the embodiment shown in FIG. 8, and the television terminal 7
  • the device in the embodiment shown in FIG. 7 above is included, and the specific structure and function are not described herein.
  • FIG. 10 is a block diagram showing the structure of an apparatus for displaying content on a smart television according to another embodiment of the present invention.
  • the smart TV display device 1100 for playing content may be a host server having a computing capability, a personal computer PC, or a portable computer or terminal that can be carried.
  • the specific embodiments of the present invention do not limit the specific implementation of the computing node.
  • the smart television display device 1100 for playing content includes a processor 1110 and a pass A Communication Interface 1120, a Memory Array 1130, and a Bus 1140.
  • the processor 1110, the communication interface 1120, and the memory 1130 complete communication with each other through the bus 1140.
  • the communication interface 1120 is configured to communicate with a network element, where the network element includes, for example, a virtual machine management center, shared storage, and the like.
  • the processor 1110 is configured to execute a program.
  • the processor 1110 may be a central processing unit CPU, or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.
  • ASIC Application Specific Integrated Circuit
  • the memory 1130 is used to store files.
  • the memory 1130 may include a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.
  • Memory 1130 can also be a memory array.
  • the memory 1130 may also be partitioned, and the blocks may be combined into a virtual volume according to certain rules.
  • the above program may be program code including computer operating instructions. This program can be used to execute:
  • the predicted recommended movie is fed back to the television to recommend to the user.
  • the predicting a movie that is highly correlated with the viewing data according to the viewing data includes:
  • a movie whose matching degree meets the preset recommendation standard value is used as a recommended movie having a high degree of relevance to the movie data.
  • the movie data includes a movie name viewed by the user each time and a category label of the movie; the training data is performed on the movie data according to a recommendation algorithm.
  • the corresponding output result is used as a training result, wherein the training result includes a category label.
  • FIG. 11 is a block diagram showing the structure of an apparatus for displaying content on a smart television according to another embodiment of the present invention.
  • the smart TV display device 2200 for playing content may be a host server having a computing capability, a personal computer PC, or a portable computer or terminal that can be carried.
  • the specific embodiments of the present invention do not limit the specific implementation of the computing node.
  • the smart television display device 2200 includes a processor 2210, a communications interface 2220, a memory array 2230, and a bus 2240.
  • the processor 2210, the communication interface 2220, and the memory 2230 complete communication with each other through the bus 2240.
  • the communication interface 2220 is for communicating with a network element, wherein the network element includes, for example, a virtual machine management center, shared storage, and the like.
  • the processor 2210 is for executing a program.
  • the processor 2210 may be a central processing unit CPU, or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.
  • ASIC Application Specific Integrated Circuit
  • the memory 2230 is for storing files.
  • the memory 2230 may include a high speed RAM memory and may also include a non-volatile memory such as at least one disk memory.
  • Memory 2230 can also be a memory array.
  • the memory 2230 may also be partitioned, and the blocks may be combined into a virtual volume according to certain rules.
  • the above program may be program code including computer operating instructions. This program can be used to execute:
  • the viewing data is transmitted to the server.
  • the method further includes:
  • embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) including computer usable program code.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. Instructions are provided for implementation in a block or blocks of a flow or a flow and/or block diagram of a flowchart The steps of the feature.
  • the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, ie may be located A place, or it can be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment. Those of ordinary skill in the art can understand and implement without deliberate labor.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Graphics (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)

Abstract

一种智能电视展示播放内容的方法、装置及系统，涉及电视业务技术领域，解决了现有技术中电视端没有智能推荐影片功能导致的用户搜索影片操作量大、影响用户的观影体验的技术问题。其中，方法包括：接收到电视端传送的用户观看的观影数据；根据所述观影数据预测与所述观影数据相关度高的推荐影片；将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。

Description

智能电视展示播放内容的方法、装置及系统
本申请要求在2015年12月15日提交中国专利局、申请号为201510953575X、发明名称为“智能电视展示播放内容的方法、装置及系统”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
本发明涉及电视业务技术领域，尤其涉及一种智能电视展示播放内容的方法、装置及系统。
背景技术
智能电视是顺应电视机“高清化”、“网络化”、“智能化”的趋势而出现的一种智能多媒体终端，具备从因特网、视频设备、计算机等多种渠道获得节目内容，通过简单易用的整合式操作界面将消费者最需要的内容在大屏幕上清晰地展现的功能。与传统电视的应用平台相比，智能电视可实现网络搜索、网络电视、视频点播(VOD)、数字音乐、网络新闻、网络视频电话等各种应用服务。目前智能电视有很多节目内容可供用户观看，又由于其互联网的特性，节目内容可以在全网进行搜索观看，因此对节目内容的快速精准定位显得尤其重要。
发明人在实现本发明的过程中发现，现有智能电视的影片推荐很少，目前的推荐功能是厂商自主进行的强制推荐，这跟用户为中心的理念背驰。用户所希望看到的是，自己想要看到的影片在最显眼的位置，从而可以迅速的观看，不必再为寻找片源烦恼。
现有技术中用户若想看某类型的电影只能明确知道目的影片的名称，通过搜索进行观看，通常有两种方式：(1)从类型中进入，然后层层深入寻找。(2)通过搜索功能找到。
但是，很多情况下，用户并非清楚的知道影片名称，很有可能仅是对 某类或某种影片感兴趣，想要观看。若通过搜索的方式查找影片将会带来很多的操作量。因此现有技术中的这种通过搜索获得想观看的影片的方式会给用户带来较多繁琐的操作，并且消耗用户的时间，影响用户的观影体验。
发明内容
本发明实施例提供一种智能电视展示播放内容的方法、装置及系统，以解决现有技术中电视端没有智能推荐影片功能导致的用户搜索影片操作量大、影响用户的观影体验的技术问题。
本发明实施例一方面提供一种智能电视展示播放内容的方法，该方法在服务器适用，主要包括：
接收到电视端传送的用户观看的观影数据；
根据所述观影数据预测与所述观影数据相关度高的推荐影片；
将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
本发明实施例另一方面提供一种智能电视展示播放内容的方法，该方法在电视端适用，主要包括：
根据用户每次观看的影片获取所述用户的观影数据；
将所述观影数据传送到服务器。
本发明实施例另一方面提供一种智能电视展示播放内容的装置，该装置适合部署在服务器上，包括：
接收模块，用于接收到电视端传送的用户观看的观影数据；
预测模块，用于根据所述观影数据预测与所述观影数据相关度高的推荐影片；
推荐模块，用于将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
本发明实施例另一方面提供一种智能电视展示播放内容的装置，该装 置适合部署在电视端上，包括：
数据获取模块，用于根据用户每次观看的影片获取所述用户的观影数据；
传送模块，用于将所述观影数据传送到服务器。
本发明实施例另一方面还提供一种智能电视展示播放内容的系统，包括服务器和电视端；该服务器包括上述适合部署在服务器上的智能电视展示播放内容的装置，所述电视端包括上述适合不是在电视端的智能电视展示播放内容的装置。
本发明提供的上述智能电视展示播放内容的方法、装置及系统，可根据用户的观影数据预测用户喜欢看的影片的类型，并将同样类型的影片通过在电视端上的屏幕显示进而推荐给用户，解决了现有技术中用户搜索影片操作不方便、以及运营商强制推荐非用户想观看的影片的技术问题，进而实现了提高电影查找和推荐的精确度、方便用户观影，增强用户观影体验的技术效果。
附图说明
附图用来提供对本发明的进一步理解，并且构成说明书的一部分，与本发明的实施例一起用于解释本发明，并不构成对本发明的限制。在附图中：
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图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性的前提下，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智能电视展示播放内容的方法的电视端上的流程图；
图2为本发明实施例提供的一种智能电视展示播放内容的方法的服务 器上的流程图；
图3为本发明实施例提供的一种智能电视展示播放内容的方法中前向传播算法的示意图；
图4为本发明实施例提供的一种智能电视展示播放内容的方法中误差反向传播算法的示意图；
图5为本发明实施例提供的一种智能电视展示播放内容的方法中BP预测网络示意图；
图6为本发明实施例提供的一种智能电视展示播放内容的方法的流程示意图；
图7为本发明实施例提供的电视端上的一种智能电视展示播放内容的装置的结构示意图；
图8为本发明实施例提供的服务器上的一种智能电视展示播放内容的装置的结构示意图；
图9为本发明实施例提供的一种智能电视展示播放内容的系统的结构示意图；
图10为本发明实施例提供的一种智能电视展示播放内容的装置的结构示意图；
图11为本发明实施例提供的一种智能电视展示播放内容的装置的结构示意图。
具体实施方式
以下结合附图对本发明的优选实施例进行说明，应当理解，此处所描述的优选实施例仅用于说明和解释本发明，并不用于限定本发明。
为使本发明实施例的目的、技术方案和优点更加清楚，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、完整地描述，显然，所描述的实施例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作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。
如图1所示，本实施例提供一种智能电视展示播放内容的方法，该方法适合部署在电视端，主要包括：
101，根据用户每次观看的影片获取所述用户的观影数据；
上述观影数据包括每个影片名称和类别标签。
电视端上通常会记录用户每次点击哪部影片进行观看，根据该记录可以得到将用户观看过的的影片的观影数据，包括影片名称和类别标签。
102，将所述观影数据传送到服务器。
可以是电视端定期主动将用户的观影数据发送到服务器，也可以是服务器下来查询用户的观影数据。
将上述用户的观影数据发送到服务器上，是为了便于服务器根据用户的观影数据预测该用户喜欢观看的影片类型，并将相应的影片类型作为推荐影片反馈到电视端上。因此，上述方案还包括：接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影片；将所述推荐影片展示给所述用户。
本实施例提供的上述方法中，电视端只需将用户的观影数据发送到服务器端，便于服务器端进行推荐影片的预测，并且由于该预测过程是在服务器上进行，因此降低了电视端的负担，同时又能满足用户的观影需求。
对应于上述电视端上的适用方法，如图2所示，本实施例还提供了一种适合部署在服务器上的一种智能电视展示播放内容的方法，主要包括：
201，接收到电视端传送的用户观看的观影数据；
202，根据所述观影数据预测与所述观影数据相关度高的推荐影片；
与观影数据中的影片类型相关度高的影片通常就是用户所喜欢的影片，因此这种影片适合作为推荐影片推荐给用户。本发明中的预测推荐影片的准确度除了取决于观影数据的用户每次观看的影片的名称和类别标 签外，还取决于推荐算法，通过将用户每次观看的影片名称和类别标签作为输入样本输入到推荐算法中进行多次的反复训练，该推荐算法的预测度便会越来越准确，并将多次训练后的训练结果作为最后预测的与所述观影数据相关度高的推荐影片推荐给用户。
下面具体介绍一下本实施例中提供的推荐算法。
神经网络算法是机器学习算法中广泛使用的一类算法，BP算法(Back Propagation)，也被称为反向传播算法。通过计算反向传输的误差，精确计算程度，所以被成为反向传播算法。这两种算法都可以用作本实施例中的推荐算法。整体来讲，算法描述，分为两个阶段：分别是训练阶段和预测阶段。
1.训练阶段
输入样本以及样本所对应的目标，通过前向计算，得到一个初步目标值，与精确的目标值进行对比，接下来反向传播误差，减少输出值与预设的目标值的差别，直到误差减少到我们预设的预测标准值。
(1)数据准备：输入样本是影片名称和每个影片名称被标记多个标签，如10个，这10个标签作为其特征进行训练。输入结果，也就是目标就是影片的类别标签(种类)，比如某类影片也被标记10个标签，作为识别其的特征。
(2)网络构造：一般情况下，分类越精细需要的网络结构也就会越复杂。这里使用中间层为1层的网络
(a)前向传输
样本输入至输入层，然后经过与权重矩阵的操作与对应的权重相乘，数学表达式如下(一)：
Figure PCTCN2016088553-appb-000001
其中，Aj为第j层的输出，
Figure PCTCN2016088553-appb-000002
分别是是示意性的代表X和W矩阵中的元素，W代表权重矩阵，X代表输入样本(数据)矩阵；计算方法如 上：权重矩阵与输入样本矩阵相乘。Xi，Wii均是对应矩阵的元素。
特别地，本实施例中考虑到权值矩阵W是直接影响到上述公式一的有效性，因此对于这个W的初始值是通过特别的字典训练算法K-SVD得到的。K-SVD，是一种数学算法。可以对一个矩阵进行分解，得到其中我们最关心的部分，即特征。
因此可以看出，神经元的输出依赖于信号的输入和权值(权重)，激活函数选择非线性的sigmoid便于计算。激活函数的输出如下式(二)：
Figure PCTCN2016088553-appb-000003
Oj是第j层经过激活函数之后的输出。
激活函数使用sigmoid，对于较大的输入值(样本与权值相乘之后的结构)都趋向于1；在输入为0时，激活输出为0.5；在输入值的负值趋于很小时，激活输出为0。特别地，在本实施例中，激活函数的曲线在使用前经过了平滑处理，这样即可以进行梯度运算，也提高了运算的准确性。
训练过程的目的就是对于给定的输入得到理想的输出。因为初始权值的选择一定不是最优的，所以在目标与实际输出之间会有残差，下式(三)。
Figure PCTCN2016088553-appb-000004
Ej是实际输出与真实值之间的差，我们一般使用这个误差作为目标进行优化，tj为第j层的目标输出，也即样本所对应的真实输出值。如图3所示。
对误差求平方之后，所有误差均为正值，将所有样本的误差值累加得到如下数学模型(四)：
Figure PCTCN2016088553-appb-000005
(b)误差反向传播
在前向传播中得到了误差是如何依赖于输入样本，输出，以及权重值的关系。在找到此关系之后，任务变为如何根据误差对权值进行更新，得到最优化的权值，得到网络。在计算更新权值时我们使用随机梯度下降法，下式(五)：
Figure PCTCN2016088553-appb-000006
其中Δw(k+1)和Δw(k)分别代表的是第k+1次和第k次迭代后权重值的更新，α与η是两个常量参数，一个用来平衡更新，一个用来在梯度下降法时更新梯度的大小。α通常选择0到1之间的数，
Figure PCTCN2016088553-appb-000007
是误差函数对权重值的梯度。
在误差回传过程中，我们需要找到输出相对于权重值的偏导数
Figure PCTCN2016088553-appb-000008
既然是反向传播，因此需要先求得误差相对于输出的偏导数，参照下式(六)：
Figure PCTCN2016088553-appb-000009
然后，反向链路上，输出依赖于激活函数的输出值，激活函数是在权重基础之上得到的式(七)：
Figure PCTCN2016088553-appb-000010
根据以上六、七两式，利用求导的链式法则，可以推导得到输出相对于权重值的导数，如下式：
Figure PCTCN2016088553-appb-000011
综上，可以得到权重的更新，式(八)：
Δwji＝2η(Oj-tj)Oj(1-Oj)xj   (八)
根据以上的算法推导过程，核心是利用随机梯度下降法进行的，反向传播的过程是BP网络的关键之处，从输出层到输入层逐层使用梯度下降，求得当前的最优值，直到网络的权值全部更新，即输入权值矩阵及输出权值矩阵均得到更新，完成一次迭代，算法迭代直到收敛。如图4所示。
2、预测阶段
当网络被训练好之后，我们就可以应用到实际使用中。输入的样本就是用户点击的影片名称和类型标签，输出的训练结果就是预测的用户所喜好的影片的类型标签。如图5所示的BP预测网络示意图，Xi是输入样本，Yi是中间节点，Zi是最后的输出。其中Xi相当于用户点击的影片名称，输出预测的用户所感兴趣的影片类型。
203，将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
用户的点击影片记录会被记录，并上传到服务器，进行预测得到用户喜好类型，那么系统在进行推荐时，则偏好会被记录，从服务器会将用户偏好的类型进行推荐。在用户的电视机中体现出来的就是，在搜索页面，其同类型的影片会被推荐出来。
可选的实施方式中，上述202实现方法如下：
根据推荐算法对所述观影数据进行训练；
根据训练结果获得相关度高的推荐影片的类别标签；
将获得的所述类别标签与影片库中预设的类别标签进行匹配；
将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
本发明实施例提供的上述智能电视展示播放内容的方法，可根据用户的观影数据预测用户喜欢看的影片的类型，并将同样类型的影片通过在电视端上的屏幕显示进而推荐给用户，解决了现有技术中用户搜索影片操作不方便、以及运营商强制推荐非用户想观看的影片的技术问题，进而实现了提高电影查找和推荐的精确度、方便用户观影，增强用户观影体验的技术效果。
详细的上述202的具体实现过程可参照图6所示的实施例。
如图6所示，根据上述图1和2的方法，本实施例具体提供一种智能电视展示播放内容，包括：
301，电视端根据用户每次观看的影片获取所述用户的观影数据，并将所述观影数据传送到服务器。
302，服务器接收到电视端传送的用户观看的观影数据；
目前用户都有账号，可根据账号，获取用户的观影数据，比如定期会在用户网络闲时，将点击影片的观影数据，传送到网络，通过网络送回服务器，服务器会将数据进行收集，并且将其类型标签取出，了下标签即预先对影片标好的tag。以便作为分类依据，比如《捉妖记》，事前将其标上三个标签“古装”，“奇幻”，“喜剧”，那么这三个将作为其样本特征，用来进行预测。
303，服务器根据推荐算法对所述观影数据进行训练；
具体而言，针对每次用户观看的影片名称和该影片的类别标签，将所述 影片名称和所述影片的类别标签作为一个输入样本输入到所述推荐算法中得到一个输出结果；其中，将得到一个输出结果的过程作为一个训练过程；
在通过多次训练过程后得到的输出结果符合预设的预测标准值时，将所述对应的输出结果作为训练结果，其中，所述训练结果包括类别标签。
例如：使用推荐算法，对得到的观影数据进行学习和训练，输入影片的名称及类型标签，将影片名字及类型对应起来作为一一对应的训练数据。将以上得到观影数据作为样本输入到推荐算法中，输入一个样本会有一个输出，每次输入并得到输出的过程就是一个训练，输出结果/训练结果包括：影片类型及得分。每次样本输入都会有一个输出，多次样本输入就会有多个输出，多次的训练可以提高推荐算法的准确性。
准确性的提高是因为每次输出是一次计算得到的数值，将这个数值与预设的目标值做差，得到一个误差，通过上述的误差反向传播过程优化这个误差，使之减小到预先设定的一个较小的预测标准阀值。此时就认为得到了精确网络，即输入的样本影片和其所属了下分类的一种预测关系。由于训练过程是在服务器中进行，所以不会对用户电视机有使用压力。
304，提取训练结果中包括的类型标签，将该类型标签作为预测的与用户观看的影片相关度高的影片的类型标签。
305，将获得的所述类别标签与影片库中预设的类别标签进行匹配；并将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
将根据用户的观影数据学习得到的类型标签与目标影片库的分类进行匹配，由于获得的类型标签可能有多个，而并不一定每个都匹配上，因此根据匹配的结果情况进行打分，将得分超过预设推荐标准值的影片作为与相关度 高的影片，并将相关程度最高的前n部影片作为推荐影片。
需要说明的是：在上述训练之前，本实施例的方案中，需要先将目标影片库的影片进行分类。即事先在服务器端对目标影片库中的影片标注类型标签，并按照类型标签分类得到一个较为初始的分类网络。
306，服务器将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
307，电视端接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影片；并将所述推荐影片展示给所述用户。
本实施例提供的方法中，通过对推荐算法进行训练可提高预测的准确性，输入的样本数据量越多，训练得到的算法越准确，预测的类别越详细推荐准确程度越高。
由此，本发明实施例提供的上述智能电视展示播放内容的方法、装置及系统，可根据用户的观影数据预测用户喜欢看的影片的类型，并将同样类型的影片通过在电视端上的屏幕显示进而推荐给用户，解决了现有技术中用户搜索影片操作不方便、以及运营商强制推荐非用户想观看的影片的技术问题，进而实现了提高电影查找和推荐的精确度、方便用户观影，增强用户观影体验的技术效果。
为了便于上述图1和图6实施例中电视端侧的方法实现，本实施例提供一种智能电视展示播放内容的装置，如图7所示，该装置包括：
数据获取模块41，用于根据用户每次观看的影片获取所述用户的观影数据；
传送模块42，用于将所述观影数据传送到服务器。
可选地，该装置还包括：
接收模块，用于接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影片；
展示模块，用于将所述推荐影片展示给所述用户。
本实施例图7提供的装置适合安装在电视端或就是一台电视或机顶盒。
相应地，为了便于上述图2和图6实施例中服务器上的方法实现，本实施例提供一种智能电视展示播放内容的装置，如图8所示，包括：
接收模块51，用于接收到电视端传送的用户观看的观影数据；
预测模块52，用于根据所述观影数据预测与所述观影数据相关度高的推荐影片；
推荐模块53，用于将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
其中，所述预测模块52包括：
训练单元，用于根据推荐算法对所述观影数据进行训练；
结果获取单元，用于根据训练结果获得相关度高的推荐影片的类别标签；
匹配单元，用于将获得的所述类别标签与影片库中预设的类别标签进行匹配，并将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
其中，所述影片数据包括每次用户观看的影片名称和该影片的类别标签；
相应地，所述训练单元，具体用于针对每次用户观看的影片名称和该影片的类别标签，将所述影片名称和所述影片的类别标签作为一个输入样本输入到所述推荐算法中得到一个输出结果；其中，将得到一个输出结果的过程作为一个训练过程；在通过多次训练过程后得到的输出结果符合预设的预测标准值时，将所述对应的输出结果作为训练结果，其中，所述训练结果包括类别标签。
由此，本发明实施例提供的上述智能电视展示播放内容的装置，可根据用户的观影数据预测用户喜欢看的影片的类型，并将同样类型的影片通过在电视端上的屏幕显示进而推荐给用户，解决了现有技术中用户搜索影片操作不方便、以及运营商强制推荐非用户想观看的影片的技术问题，进而实现了提高电影查找和推荐的精确度、方便用户观影，增强用户观影体验的技术效果。
本实施例图8所提供的该装置适合安装在服务器上或就是一台服务器，
本发明实施例提供的图7和8中的装置，可执行上述方法实施例，其具体实现原理和技术效果，可参见上述方法实施例，本实施例此处不再赘述。
如图9所示，本实施继续提供一种智能电视展示播放内容的系统，包括服务器60和电视端70；该服务器60包括上述图8所示的实施例中的装置，该所述电视端7包括上述图7中所示的实施例中的装置，具体结构和作用在此不赘述。
图10示出了本发明的另一个实施例的一种智能电视展示播放内容的装置的结构框图。所述智能电视展示播放内容的装置1100可以是具备计算能力的主机服务器、个人计算机PC、或者可携带的便携式计算机或终端等。本发明具体实施例并不对计算节点的具体实现做限定。
该智能电视展示播放内容的装置1100包括处理器(processor)1110、通 信接口(Communications Interface)1120、存储器(memory array)1130和总线1140。其中，处理器1110、通信接口1120、以及存储器1130通过总线1140完成相互间的通信。
通信接口1120用于与网元通信，其中网元包括例如虚拟机管理中心、共享存储等。
处理器1110用于执行程序。处理器1110可能是一个中央处理器CPU，或者是专用集成电路ASIC(Application Specific Integrated Circuit)，或者是被配置成实施本发明实施例的一个或多个集成电路。
存储器1130用于存放文件。存储器1130可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。存储器1130也可以是存储器阵列。存储器1130还可能被分块，并且所述块可按一定的规则组合成虚拟卷。
在一种可能的实施方式中，上述程序可为包括计算机操作指令的程序代码。该程序具体可用于执行：
接收到电视端传送的用户观看的观影数据；
根据所述观影数据预测与所述观影数据相关度高的推荐影片；
将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
在一种可能的实现方式中，所述根据所述观影数据预测与所述观影数据相关度高的推荐影片，包括：
根据推荐算法对所述观影数据进行训练；
根据训练结果获得相关度高的推荐影片的类别标签；
将获得的所述类别标签与影片库中预设的类别标签进行匹配；
将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
在一种可能的实现方式中，所述影片数据包括每次用户观看的影片名称和该影片的类别标签；所述根据推荐算法对所述影片数据进行训练包 括：
针对每次用户观看的影片名称和该影片的类别标签，将所述影片名称和所述影片的类别标签作为一个输入样本输入到所述推荐算法中得到一个输出结果；其中，将得到一个输出结果的过程作为一个训练过程；
在通过多次训练过程后得到的输出结果符合预设的预测标准值时，将所述对应的输出结果作为训练结果，其中，所述训练结果包括类别标签。
图11示出了本发明的另一个实施例的一种智能电视展示播放内容的装置的结构框图。所述智能电视展示播放内容的装置2200可以是具备计算能力的主机服务器、个人计算机PC、或者可携带的便携式计算机或终端等。本发明具体实施例并不对计算节点的具体实现做限定。
该智能电视展示播放内容的装置2200包括处理器(processor)2210、通信接口(Communications Interface)2220、存储器(memory array)2230和总线2240。其中，处理器2210、通信接口2220、以及存储器2230通过总线2240完成相互间的通信。
通信接口2220用于与网元通信，其中网元包括例如虚拟机管理中心、共享存储等。
处理器2210用于执行程序。处理器2210可能是一个中央处理器CPU，或者是专用集成电路ASIC(Application Specific Integrated Circuit)，或者是被配置成实施本发明实施例的一个或多个集成电路。
存储器2230用于存放文件。存储器2230可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。存储器2230也可以是存储器阵列。存储器2230还可能被分块，并且所述块可按一定的规则组合成虚拟卷。
在一种可能的实施方式中，上述程序可为包括计算机操作指令的程序代码。该程序具体可用于执行：
根据用户每次观看的影片获取所述用户的观影数据；
将所述观影数据传送到服务器。
在一种可能的实现方式中，该方法还包括：
接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影片；
将所述推荐影片展示给所述用户。
本领域内的技术人员应明白，本发明的实施例可提供为方法、系统、或计算机程序产品。因此，本发明可采用完全硬件实施例、完全软件实施例、或结合软件和硬件方面的实施例的形式。而且，本发明可采用在一个或多个其中包含有计算机可用程序代码的计算机可用存储介质(包括但不限于磁盘存储器和光学存储器等)上实施的计算机程序产品的形式。
本发明是参照根据本发明实施例的方法、设备(系统)、和计算机程序产品的流程图和/或方框图来描述的。应理解可由计算机程序指令实现流程图和/或方框图中的每一流程和/或方框、以及流程图和/或方框图中的流程和/或方框的结合。可提供这些计算机程序指令到通用计算机、专用计算机、嵌入式处理机或其他可编程数据处理设备的处理器以产生一个机器，使得通过计算机或其他可编程数据处理设备的处理器执行的指令产生用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的装置。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定 的功能的步骤。
前述对本发明的具体示例性实施方案的描述是为了说明和例证的目的。这些描述并非想将本发明限定为所公开的精确形式，并且很显然，根据上述教导，可以进行很多改变和变化。对示例性实施例进行选择和描述的目的在于解释本发明的特定原理及其实际应用，从而使得本领域的技术人员能够实现并利用本发明的各种不同的示例性实施方案以及各种不同的选择和改变。本发明的范围意在由权利要求书及其等同形式所限定。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。

Claims (13)

  1. 一种智能电视展示播放内容的方法，其特征在于，包括：
    接收到电视端传送的用户观看的观影数据；
    根据所述观影数据预测与所述观影数据相关度高的推荐影片；
    将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述观影数据预测与所述观影数据相关度高的推荐影片，包括：
    根据推荐算法对所述观影数据进行训练；
    根据训练结果获得相关度高的推荐影片的类别标签；
    将获得的所述类别标签与影片库中预设的类别标签进行匹配；
    将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
  3. 根据权利要求2所述的方法，其特征在于，所述影片数据包括每次用户观看的影片名称和该影片的类别标签；所述根据推荐算法对所述影片数据进行训练包括：
    针对每次用户观看的影片名称和该影片的类别标签，将所述影片名称和所述影片的类别标签作为一个输入样本输入到所述推荐算法中得到一个输出结果；其中，将得到一个输出结果的过程作为一个训练过程；
    在通过多次训练过程后得到的输出结果符合预设的预测标准值时，将所述对应的输出结果作为训练结果，其中，所述训练结果包括类别标签。
  4. 一种智能电视展示播放内容的方法，其特征在于，包括：
    根据用户每次观看的影片获取所述用户的观影数据；
    将所述观影数据传送到服务器。
  5. 根据权利要求4所述的方法，其特征在于，该方法还包括：
    接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影 片；
    将所述推荐影片展示给所述用户。
  6. 一种智能电视展示播放内容的装置，其特征在于，包括：
    接收模块，用于接收到电视端传送的用户观看的观影数据；
    预测模块，用于根据所述观影数据预测与所述观影数据相关度高的推荐影片；
    推荐模块，用于将预测到的所述推荐影片反馈到所述电视端推荐给所述用户。
  7. 根据权利要求6所述的装置，其特征在于，所述预测模块包括：
    训练单元，用于根据推荐算法对所述观影数据进行训练；
    结果获取单元，用于根据训练结果获得相关度高的推荐影片的类别标签；
    匹配单元，用于将获得的所述类别标签与影片库中预设的类别标签进行匹配，并将匹配度符合预设推荐标准值的影片作为与所述影片数据相关度高的推荐影片。
  8. 根据权利要求7所述的装置，其特征在于，所述影片数据包括每次用户观看的影片名称和该影片的类别标签；
    所述训练单元，具体用于针对每次用户观看的影片名称和该影片的类别标签，将所述影片名称和所述影片的类别标签作为一个输入样本输入到所述推荐算法中得到一个输出结果；其中，将得到一个输出结果的过程作为一个训练过程；在通过多次训练过程后得到的输出结果符合预设的预测标准值时，将所述对应的输出结果作为训练结果，其中，所述训练结果包括类别标签。
  9. 一种智能电视展示播放内容的装置，其特征在于，包括：
    数据获取模块，用于根据用户每次观看的影片获取所述用户的观影数据；
    传送模块，用于将所述观影数据传送到服务器。
  10. 根据权利要求9所述的装置，其特征在于，该装置还包括：
    接收模块，用于接收到所述服务器反馈的与所述用户观看的影片相关度高的推荐影片；
    展示模块，用于将所述推荐影片展示给所述用户。
  11. 一种智能电视展示播放内容的系统，其特征在于，包括服务器和电视端；所述服务器如权利要求6-8中任意一项所述的装置，所述电视端包括如权利要求9或10中任意一项所述的装置。
  12. [根据细则91更正 26.07.2016]　
    一种电子装置，包含处理器，以及存储能够被所述处理器执行的程序的存储器，其特征在于，处理器被配置为执行权利要求1-5任一项所述的视频播放方法。
  13. [根据细则91更正 26.07.2016]　
    一种计算机存储介质，其特征在于，所述计算机存储介质可存储有程序，所述程序执行时可实现包括权利要求1-5任一项所述的视频播放方法。
PCT/CN2016/088553 2015-12-15 2016-07-05 智能电视展示播放内容的方法、装置及系统 Ceased WO2017101328A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/247,116 US20170169330A1 (en) 2015-12-15 2016-08-25 Method and Electronic Device for Displaying Play Content in Smart Television

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510953575.XA CN105915957A (zh) 2015-12-15 2015-12-15 智能电视展示播放内容的方法、装置及系统
CN201510953575.X 2015-12-15

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/247,116 Continuation US20170169330A1 (en) 2015-12-15 2016-08-25 Method and Electronic Device for Displaying Play Content in Smart Television

Publications (1)

Publication Number Publication Date
WO2017101328A1 true WO2017101328A1 (zh) 2017-06-22

Family

ID=56744279

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/088553 Ceased WO2017101328A1 (zh) 2015-12-15 2016-07-05 智能电视展示播放内容的方法、装置及系统

Country Status (2)

Country Link
CN (1) CN105915957A (zh)
WO (1) WO2017101328A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449754A (zh) * 2020-03-26 2021-09-28 百度在线网络技术（北京）有限公司 标签的匹配模型训练和展示方法、装置、设备及介质
CN114880515A (zh) * 2022-05-05 2022-08-09 广州市影擎电子科技有限公司 一种基于数字化虚拟服务的智能影院的控制方法及系统
CN116910045A (zh) * 2023-07-20 2023-10-20 深圳市酷开网络科技股份有限公司 影片聚合补全处理方法、装置

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107229943A (zh) * 2017-04-26 2017-10-03 厦门吉比特网络技术股份有限公司 一种网络游戏玩家流失的预测方法
CN112231580B (zh) * 2020-11-10 2024-04-02 腾讯科技（深圳）有限公司 基于人工智能的信息推荐方法、装置、电子设备及存储介质
CN113537215A (zh) * 2021-07-19 2021-10-22 山东福来克思智能科技有限公司 一种视频标签标注的方法及设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103686236A (zh) * 2013-11-19 2014-03-26 乐视致新电子科技（天津）有限公司 推荐视频资源的方法及系统
CN103686403A (zh) * 2013-12-04 2014-03-26 康佳集团股份有限公司 一种电视机桌面的显示方法和装置
CN104394471A (zh) * 2014-11-19 2015-03-04 四川长虹电器股份有限公司 一种智能推荐用户喜爱节目的方法
CN104539981A (zh) * 2014-11-26 2015-04-22 四川长虹电器股份有限公司 一种热门电视频道实时推荐系统及方法
WO2015070807A1 (zh) * 2013-11-15 2015-05-21 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

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6922680B2 (en) * 2002-03-19 2005-07-26 Koninklijke Philips Electronics N.V. Method and apparatus for recommending an item of interest using a radial basis function to fuse a plurality of recommendation scores
CN103714130B (zh) * 2013-12-12 2017-08-22 深圳先进技术研究院 视频推荐系统及方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015070807A1 (zh) * 2013-11-15 2015-05-21 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
CN103686236A (zh) * 2013-11-19 2014-03-26 乐视致新电子科技（天津）有限公司 推荐视频资源的方法及系统
CN103686403A (zh) * 2013-12-04 2014-03-26 康佳集团股份有限公司 一种电视机桌面的显示方法和装置
CN104394471A (zh) * 2014-11-19 2015-03-04 四川长虹电器股份有限公司 一种智能推荐用户喜爱节目的方法
CN104539981A (zh) * 2014-11-26 2015-04-22 四川长虹电器股份有限公司 一种热门电视频道实时推荐系统及方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113449754A (zh) * 2020-03-26 2021-09-28 百度在线网络技术（北京）有限公司 标签的匹配模型训练和展示方法、装置、设备及介质
CN113449754B (zh) * 2020-03-26 2023-09-22 百度在线网络技术（北京）有限公司 标签的匹配模型训练和展示方法、装置、设备及介质
CN114880515A (zh) * 2022-05-05 2022-08-09 广州市影擎电子科技有限公司 一种基于数字化虚拟服务的智能影院的控制方法及系统
CN116910045A (zh) * 2023-07-20 2023-10-20 深圳市酷开网络科技股份有限公司 影片聚合补全处理方法、装置

Also Published As

Publication number Publication date
CN105915957A (zh) 2016-08-31

Similar Documents

Publication Publication Date Title
US20210027160A1 (en) End-to-end deep collaborative filtering
US11288444B2 (en) Optimization techniques for artificial intelligence
CA3111094C (en) Noise contrastive estimation for collaborative filtering
US8589434B2 (en) Recommendations based on topic clusters
CN113515690B (zh) 内容召回模型的训练方法、内容召回方法、装置及设备
WO2017101328A1 (zh) 智能电视展示播放内容的方法、装置及系统
Huang et al. Learning to learn a cold-start sequential recommender
CN110347781B (zh) 文章倒排方法、文章推荐方法、装置、设备及存储介质
US20170169330A1 (en) Method and Electronic Device for Displaying Play Content in Smart Television
US20250316062A1 (en) Self-supervised audio-visual learning for correlating music and video
CN108563648B (zh) 数据显示方法和装置、存储介质及电子装置
CN114722268B (zh) 媒体资源的处理方法和装置、存储介质及电子设备
CN113987262A (zh) 一种视频推荐信息确定方法、装置、电子设备及存储介质
CN113641900A (zh) 信息推荐方法及装置
CN118797098A (zh) 视频排序推荐方法、装置、电子设备、存储介质及产品
CN117909542A (zh) 一种视频推荐方法、装置、设备及存储介质
CN115129963B (zh) 搜索处理方法和装置
WO2023174075A1 (zh) 内容检测模型的训练方法、内容检测方法及装置
CN113934872A (zh) 一种搜索结果的排序方法、装置、设备以及存储介质
CN114860988A (zh) 一种视频媒资推荐排序方法、装置及电子设备
CN117056561B (zh) 视频推荐方法、装置、设备、存储介质及程序产品
Chen et al. Content-based video relevance prediction with multi-view multi-level deep interest network
WO2025241598A1 (zh) 一种数据处理方法以及相关装置
CN118861985A (zh) 偏差确定方法、双塔模型训练方法、推荐方法及相关产品
CN121637078A (zh) 基于内容的数据匹配方法和模型训练方法、装置、设备

Legal Events

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

Ref document number: 16874391

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16874391

Country of ref document: EP

Kind code of ref document: A1