WO2018205642A1 - 视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质 - Google Patents
视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质 Download PDFInfo
- Publication number
- WO2018205642A1 WO2018205642A1 PCT/CN2017/120178 CN2017120178W WO2018205642A1 WO 2018205642 A1 WO2018205642 A1 WO 2018205642A1 CN 2017120178 W CN2017120178 W CN 2017120178W WO 2018205642 A1 WO2018205642 A1 WO 2018205642A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- video
- recommended
- time
- revenue
- user
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/735—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/71—Indexing; Data structures therefor; Storage structures
Definitions
- the present application relates to the field of video playback, and in particular to a video revenue calculation modeling apparatus, method, and video recommendation apparatus, method, and server and storage medium.
- Short video is a kind of Internet content transmission method, which is generally spread on new Internet media.
- the video content is less than 5 minutes.
- Short videos are usually played on a variety of new media platforms, suitable for viewing in mobile and short-lived, high-frequency push video content, usually ranging from a few seconds to a few minutes.
- the content combines skills sharing, humorous, fashion trends, social hotspots, street interviews, public welfare education, advertising creativity, business customization and other topics.
- short video applications which mainly recommend short videos with short durations, and then the videos displayed by such applications for users are usually disorganized, and no corresponding recommendations are recommended according to user interests.
- Short video user experience is poor, short video revenue is not high.
- an object of the embodiments of the present application is to provide a video revenue calculation modeling apparatus, method, and video recommendation apparatus, method, server, and storage medium to improve the above problems.
- an embodiment of the present application provides a video revenue calculation modeling apparatus, where the video revenue calculation modeling apparatus includes:
- a recommended video set extracting unit configured to extract and summarize the video that meets the preset condition according to the first user browsing record from the pre-stored video library into a recommended video set
- a first information sending unit configured to send the recommended video set to a client associated with the first user browsing record
- the information receiving unit is configured to receive, after a preset second time, the video browsing feedback information sent by the client, where the video feedback information includes a time when the user browses each video in the recommended video set, and is recommended from the current time. At least one of a historical display status of each video in the video set, a time of playing each video in the recommended video set, and a total time of each video in the recommended video set;
- a model establishing unit configured to establish a video revenue coefficient calculation model according to the video feedback information.
- the embodiment of the present application further provides a video revenue calculation modeling method, where the video revenue calculation modeling method includes:
- the video that meets the preset condition is extracted and summarized into the recommended video set according to the first user browsing record;
- the video feedback information includes each user browsing the video in the recommended video set, the time from the current time, and each video in the recommended video set. At least one of a historical display status, a time to play each video in the recommended video set, and a total time of each video in the recommended video set;
- a video revenue coefficient calculation model is established according to the video feedback information.
- the embodiment of the present application further provides a video recommendation device, where the video recommendation device includes:
- a data requesting unit configured to request, from a client, a preset second user browsing record in a first time, where the second user browsing record includes that each video of the user browsing the recommended video set is at a current time Second time
- a video revenue expectation value calculation unit configured to calculate a video of each video stored in the video database according to the video revenue coefficient calculation model provided above, the second time when the user browses the video in the recommended video set, and the time decay factor of the current time Expectation of return;
- a second information sending unit configured to send the video stored in the video database to the client according to a video revenue expectation value of each video.
- the embodiment of the present application further provides a video recommendation method, where the video recommendation method includes:
- the video stored in the video database is sent to the client based on the video revenue expectation of each video.
- the embodiment of the present application further provides a server, including:
- One or more processors are One or more processors;
- One or more applications wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more programs configured to:
- the video that meets the preset condition is extracted and summarized into the recommended video set according to the first user browsing record;
- the video feedback information includes each user browsing the video in the recommended video set, the time from the current time, and each video in the recommended video set. At least one of a historical display status, a time to play each video in the recommended video set, and a total time of each video in the recommended video set;
- a video revenue coefficient calculation model is established according to the video feedback information.
- the embodiment of the present application further provides a server, including:
- One or more processors are One or more processors;
- One or more applications wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more programs configured to:
- the video stored in the video database is sent to the client based on the video revenue expectation of each video.
- the embodiment of the present application further provides a computer readable storage medium carrying one or more computer instruction programs, where the computer instruction program is executed by one or more processors, the one or more The processor executes the above video revenue calculation modeling method or the above video recommendation method.
- the video revenue calculation modeling apparatus and method provided by the embodiments of the present application firstly extract and summarize the video that meets the preset condition according to the first user browsing record from the pre-stored video library. And recommending the video set; sending the recommended video set to the client associated with the first user browsing record; and then receiving the video browsing feedback information sent by the client after the preset second time, and finally The video feedback information establishes a video revenue coefficient calculation model.
- the established video revenue coefficient calculation model has high reliability and reference degree.
- the video recommendation apparatus and method provided by the embodiment of the present application firstly requests a second user browsing record in a preset time by a client; and then calculates according to the video revenue coefficient provided above.
- the model the user browses each video in the recommended video set from the second time of the current time and the time decay factor to calculate the video revenue expectation value of each video stored in the video database; finally, according to the video revenue expectation value of each video, the video database is The stored video is sent to the client. Since the established video revenue coefficient calculation model fits each user's own video browsing interests and actual conditions, the established video revenue coefficient calculation model has high reliability and reference degree, so the calculation model based on the established video revenue coefficient is recommended for the user.
- the video is also very suitable for the user's video browsing field and browsing interest, greatly improving the accuracy of video recommendation, and also improving the user experience of browsing video.
- FIG. 1 is a schematic diagram of interaction between a server and a client according to an embodiment of the present application
- FIG. 2 is a structural block diagram of a server according to an embodiment of the present application.
- FIG. 3 is a schematic diagram of functional modules of a video revenue calculation modeling apparatus according to an embodiment of the present application.
- FIG. 4 is a schematic diagram of a display state of a recommended video set on a display interface of a client according to an embodiment of the present disclosure
- FIG. 5 is a flowchart of a video revenue calculation modeling method according to an embodiment of the present application.
- FIG. 6 is a schematic diagram of functional modules of a video recommendation apparatus according to an embodiment of the present disclosure.
- FIG. 7 is a flowchart of a video recommendation method according to an embodiment of the present application.
- FIG. 8 is a block diagram showing an internal structure of a server according to an embodiment of the present application.
- FIG. 9 is a block diagram showing the internal structure of another server according to an embodiment of the present application.
- Icon 100-client; 200-server; 300-network; 400-video revenue calculation modeling device; 500-video recommendation device; 101-processor; 102-memory; 103-storage controller; 104-peripheral interface 301-video set extracting unit; 302-first information transmitting unit; 303-information receiving unit; 304-model establishing unit; 601-data requesting unit; 602-video revenue expectation value calculating unit; 603-sorting unit; Two information transmitting unit; 810-first processor; 820-first non-volatile memory; 821-first operating system; 822-first database; 830-first internal memory; 840-first network interface; a first display screen; 910-second processor; 920-second non-volatile memory; 921-second operating system; 922-second database; 930-second internal memory; 940-second network interface; 950 - second display.
- the server and the storage medium can be applied to the application environment as shown in FIG. 1.
- the client 100 and the server 200 are located in the network 300. Through the network 300, the client 100 performs data interaction with the server 200.
- at least one application (Application, APP) is installed in the client 100, and corresponds to the server 200 to provide services for the user.
- the server 200 can be, but is not limited to, a web server, a database server, a cloud server, and the like.
- the client 100 can be, but is not limited to, a smart phone, a personal computer (PC), a tablet computer, a personal digital assistant (PDA), a mobile Internet device (MID), and the like.
- the operating system of the client 100 may be, but not limited to, an Android system, an IOS (iPhone operating system) system, a Windows phone system, a Windows system, and the like.
- FIG. 2 shows a block diagram of a structure of a server 200 that can be applied to an embodiment of the present application.
- the server includes a video revenue calculation modeling device 400, a video recommendation device 500, a processor 101, a memory 102, a memory controller 103, and a peripheral interface 104.
- the memory 102, the memory controller 103, the processor 101, and the peripheral interface 104 are electrically connected directly or indirectly to each other to implement data transmission or interaction.
- the components can be electrically connected to one another via one or more communication buses or signal lines.
- the video revenue calculation modeling device 400 and the video recommendation device 500 include at least one operating system (OS) that can be stored in the memory 102 in the form of software or firmware or is solidified at the client 100.
- the processor 101 is configured to execute an executable module stored in the memory 102, for example, the video revenue calculation modeling device 400, a software function module or a computer program included in the video recommendation device 500.
- the memory 102 may be, but not limited to, a random access memory (RAM), a read only memory (Read Only Memory, ROM), a Programmable Read-Only Memory (PROM), and an erasable memory. Erasable Programmable Read-Only Memory (EPROM), Electric Erasable Programmable Read-Only Memory (EEPROM), and the like.
- RAM random access memory
- ROM read only memory
- PROM Programmable Read-Only Memory
- EPROM Erasable Programmable Read-Only Memory
- EEPROM Electric Erasable Programmable Read-Only Memory
- the memory 102 is configured to store a program, and the processor 101 executes the program after receiving the execution instruction, and the method executed by the server defined by the flow process disclosed in any of the foregoing embodiments of the present application may be applied to In processor 101, or implemented by processor 101.
- Processor 101 may be an integrated circuit chip with signal processing capabilities.
- the processor 101 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP Processor, etc.), or a digital signal processor (DSP), an application specific integrated circuit. (ASIC), off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component.
- CPU central processing unit
- NP Processor network processor
- DSP digital signal processor
- ASIC application specific integrated circuit.
- FPGA off-the-shelf programmable gate array
- the general purpose processor may be a microprocessor or the processor 101 may be any conventional processor 101 or the like.
- Peripheral interface 104 couples various input/output devices to processor 101 and memory 102.
- peripheral interface 104, processor 101, and memory controller 103 can be implemented in a single chip. In other instances, they can be implemented by separate chips.
- the structure shown in FIG. 2 is merely illustrative, and the server 200 may further include more or less components than those shown in FIG. 2, or have a different configuration than that shown in FIG. 2.
- the components shown in Figure 2 can be implemented in hardware, software, or a combination thereof.
- the video revenue calculation modeling apparatus 400 includes a recommended video set extraction unit 301, a first information sending unit 302, an information receiving unit 303, and Model building unit 304.
- the recommended video set extracting unit 301 is configured to extract and summarize the video that meets the preset condition according to the first user browsing record from the pre-stored video library into the recommended video set.
- the recommended video set extraction unit 301 may include:
- a data requesting subunit configured to request, from the client 100, the first user browsing record in a preset first time.
- the first user browsing record includes the video tag information.
- the video tag is obtained in many ways, for example, the keyword is extracted by the title of the video, and the keyword is mapped to the corresponding video tag information or Manually tag video with video tag information.
- the video tag information can be "NBA Basketball Game”, “American Election”, “Beauty Fashion Show”, etc., and is merely an example.
- the first user browsing record is a browsing record cached by the user before the video browsed by the client 100.
- the first user browsing history is selected from the current time, which is more suitable for the video browsing preference of the user at the current time.
- the preset first time is preferably 3 days or a week or a half month.
- the video set induction subunit is configured to find a video associated with the video tag information in a pre-stored video library, and classify all the found videos into a recommended video set.
- Video A is about "military” and video is about “basketball”.
- video A is about "military” and video is about "basketball”.
- all videos containing "military” and “basketball” are found. , as a recommended video set recommended to users.
- the video set induction subunit is configured to find a video associated with the video tag information in a pre-stored video library, and classify all the found videos into candidate video sets. .
- the recommended video set extraction unit 301 further includes: a selection subunit, configured to select a candidate video from the candidate video set, and each candidate video has not generated a display record at the client 100, and the selected candidate video is summarized as Recommended video set.
- the number of videos in the recommended video set may be 10, 50, 100, etc., depending on the actual situation, and is not limited herein.
- the first information sending unit 302 is configured to send the recommended video set to the client 100 associated with the first user browsing record.
- the client 100 associated with the first user browsing record refers to the client 100 that generates and caches the first user browsing record.
- the information receiving unit 303 is configured to receive the video browsing feedback information sent by the client 100 after a preset second time.
- the video feedback information includes a time when the user browses each video in the recommended video set, a historical display state of each video in the recommended video set, and a time of playing each video in the recommended video set. And the total time of each video in the recommended video set.
- the video feedback information may further include: a time when the user browses each video in the recommended video set, a time from the current time, a historical display state of each video in the recommended video set, and each of the recommended video sets. Any one or more of the time of the video and the total time of each video in the recommended video set is merely illustrative.
- the model establishing unit 304 is configured to establish a video revenue coefficient calculation model according to the video feedback information.
- the model establishing unit 304 includes: a video feedback effect calculation model establishing subunit,
- feedback u, i, j is a video feedback effect value
- isview u j is a historical display state value of each video of the recommended video set; for example, -1 indicates that the video is displayed on the screen of the client 100 by the user. However, there is no click to play, 0 means that the video is not displayed on the client 100, 1 means that the video is displayed on the client 100 and the user has clicked to play.
- Viewdays u, i User browses the time of each video in the recommended video set from the current time, in days. For example, if the user browses the video on the same day, it takes 1 and the user browses the video yesterday and takes 2, and so on.
- Playtime u,j represents the time when the user plays each video of the recommended video set
- videotime j represents the total time of each video of the recommended video set
- precent u,j represents the completion rate of the user watching each video.
- the total duration of a military video is 5 minutes
- the user's viewing completion rate of the military video is 60%.
- Lapse is the time decay factor for each video
- Ti represents the set of labels for the video that the user has viewed
- Tj represents the set of labels for the video contained in the recommended video set.
- lapse is a constant and lapse>0, indicating time decay factor
- lapse is a manual configuration parameter. The larger the value of the configuration parameter, the more attention is paid to the behavior of the user browsing the video recently.
- 0 is taken, the user's browsing time is displayed. The behavior is treated the same. In actual business, it is generally taken as 1, and the weight of each day in the future is in turn And so on. Exception, consider The reason for assigning a value of 0 is that videos without the same tag are considered irrelevant.
- the foregoing operations are performed on a batch user, and a large amount of video feedback information can be obtained, and multiple video revenue coefficients are obtained according to a large amount of video feedback information.
- the average value of all signals whose video effect value is not 0 is used to measure the profit of short video. Adding 1 to this is a correction value to ensure that the video yield coefficients of the same label are greater than or equal to 0, but not the same label. There is no correlation between the videos, and the video yield coefficient is also zero.
- the established video revenue coefficient calculation model is obtained according to the user's real-time video feedback information of the recommended video set, which is scientific and reasonable, and conforms to the actual situation of the user currently browsing the video.
- the embodiment of the present application further provides a video revenue calculation modeling method, which needs to be described in the video revenue calculation modeling method provided by the embodiment of the present application, the basic principle, the technical effects generated, and the first The same is true for an embodiment.
- the video revenue calculation modeling method includes:
- Step S501 Extract, from the pre-stored video library, the video that meets the preset condition according to the first user browsing record into a recommended video set.
- step S501 is performed by the recommended video set extracting unit 301 described above.
- step S501 includes but is not limited to the following two modes:
- the first type requesting, by the client 100, the first user browsing record in a preset first time, wherein the first user browsing record includes video tag information; in a pre-stored video library Find the video associated with the video tag information and group all the found videos into a recommended video set.
- a second type requesting, by the client 100, the first user browsing record in a preset first time, wherein the first user browsing record includes video tag information; in a pre-stored video library Finding a video associated with the video tag information, and summarizing all the found videos into a candidate video set; selecting candidate videos from the candidate video set, and each candidate video has not generated a display record at the client 100 , the selected candidate videos are summarized into recommended video sets.
- Step S502 Send the recommended video set to the client 100 associated with the first user browsing record.
- step S502 is performed by the first information transmitting unit 302 described above.
- Step S503 Receive video browsing feedback information sent by the client 100 after a preset second time.
- step S503 is performed by the above-described information receiving unit 303.
- the video feedback information includes a time when the user browses each video in the recommended video set, a historical display state of each video in the recommended video set, and a time of playing each video in the recommended video set. And the total time of each video in the recommended video set.
- Step S504 Establish a video revenue coefficient calculation model according to the video feedback information.
- step S504 is performed by the model establishing unit 304 described above.
- the flow of calculating a model based on establishing a video revenue coefficient is:
- a video revenue coefficient calculation model is established, where earn i,j represents the video revenue coefficient of each video.
- an embodiment of the present application further provides a video recommendation apparatus 500.
- the video recommendation apparatus 500 includes a data request unit 601, a video benefit expectation value calculation unit 602, a sorting unit 603, and a second information sending unit 604.
- the data requesting unit 601 is configured to request a client 100 to record a second user browsing record in a preset time.
- the second user browsing record also includes video tag information and a second time when the user browses each video in the recommended video set from the current time.
- the video tag is obtained by a plurality of methods, for example, extracting a keyword by using a title of the video, and mapping the keyword to the corresponding video tag information or manually tagging the video with the video tag information by the operator.
- the second user browsing history is selected from the current time, which is more suitable for the video browsing preference of the user at the current time.
- the second time preset is also preferably 3 days or a week or a half month.
- the video revenue expectation value calculation unit 602 is configured to calculate the video revenue coefficient calculation model according to the first embodiment or the second embodiment, and the user browses the second time of the current time and the time decay factor of each video in the recommended video set to calculate the storage in the video database. The video revenue expectation for each video.
- the video revenue expectation value calculation unit 602 is configured to use the calculation formula Calculating the expected value of the video revenue, wherein viewdays a, i indicates that the user browses the second time of each video in the recommended video set from the current time, lapse represents the time decay factor, expect a, j video revenue expectation value of each video, earn i , j represents the video revenue coefficient of each video, and the calculation method of the ear i, j is as described in the first embodiment or the second embodiment, and will not be repeated here.
- the sorting unit 603 is configured to sort the videos stored in the video database in descending order according to the video revenue expectation value of each video.
- the ordering of the videos stored in the video database may not be arranged in descending order according to the expected value of the video revenue of each video, or may be other arrangements, which will not be exemplified herein.
- the information sending unit is configured to send the video stored in the video database to the client 100 in an arranged order.
- the video sequence displayed on the client 100 is arranged in the order of the expected value of the video revenue, and the user first sees the video at the client 100 as the video that best meets the user's viewing needs.
- the method for sending the video to the information sending unit may be not only the foregoing manner, but also, for example, determining whether the video revenue expectation value is greater than a preset threshold, and only the video revenue expectation value is greater than a preset threshold. The video is sent to the client 100.
- the embodiment of the present application further provides a video recommendation method.
- the video recommendation method provided by the embodiment of the present application has the same basic principles and technical effects as the foregoing embodiment, and is briefly described.
- the video recommendation method includes:
- Step S701 Request a second user browsing record in a preset time period from a client 100.
- the second user browsing record includes a second time when the user browses each video in the recommended video set from the current time. It can be understood that step S701 is performed by the above-described data requesting unit 601.
- Step S702 Calculate a video revenue coefficient calculation model according to Embodiment 1 or Embodiment 2, calculate a second time of each video of the user in the recommended video set, and a time decay factor to calculate each video stored in the video database. Video revenue expectation.
- step S702 is performed by the video benefit expectation value calculation unit 602 described above.
- step S702 is:
- Step S703 Describing the videos stored in the video library in descending order according to the video revenue expectation value of each video.
- step S703 is performed by the sorting unit 603 described above.
- Step S704 Send the video stored in the video database to the client 100 in the arranged order.
- step S704 is performed by the second information transmitting unit 604 described above.
- the present application provides a server as an embodiment, which is specifically as follows:
- the server comprising: one or more processors; a memory; one or more applications, wherein the one or more applications are stored in the memory and configured to be processed by the one or more Executing, the one or more programs are configured to: extract, from the pre-stored video library, the video that meets the preset condition according to the first user browsing record into a recommended video set; and the recommended video set Sending to the client associated with the first user browsing record; receiving the video browsing feedback information sent by the client after the preset second time, the video feedback information including the user browsing each of the recommended video sets
- the video is at least one of a time from the current time, a historical display state of each video in the recommended video set, a time of playing each video in the recommended video set, and a total time of each video in the recommended video set;
- the video feedback information establishes a video yield coefficient calculation model.
- the server includes a first processor 810, a first memory 820, a first internal memory 830, a first network interface 840, and a first display screen 850 connected by a system bus.
- the first processor 810 is configured to implement a computing function and a function of controlling the operation of the terminal device, and the first processor 810 is configured to perform the video revenue calculation modeling method provided by the above embodiments.
- the first processor 810 is configured to extract, from the pre-stored video library, the video that meets the preset condition according to the first user browsing record into a recommended video set; and send the recommended video set to the first
- the user browses the associated client of the record; and receives the video browsing feedback information sent by the client after the preset second time, where the video feedback information includes the time when the user browses each video in the recommended video set respectively from the current time.
- the first memory 820 is a non-volatile storage medium, and the first storage system 821, the first database 822, and a computer program for implementing the video revenue calculation modeling method provided by the above embodiments, and the execution of the computer program Candidate intermediate data as well as result data.
- the first network interface 840 is for communicating with a server, and the first network interface 840 includes a radio frequency transceiver.
- the step of extracting and summarizing the video that meets the pre-set condition into the recommended video set according to the first user browsing record in the pre-stored video library of the server includes: to the client Requesting the first user browsing record in a preset first time, wherein the first user browsing record includes video tag information; and the pre-stored video library is found to be associated with the video tag information Video and group all the videos found into a recommended video set.
- the step of extracting and summarizing the video that meets the pre-set condition into the recommended video set according to the first user browsing record in the pre-stored video library of the server includes: to the client Requesting the first user browsing record in a preset first time, wherein the first user browsing record includes video tag information; and the pre-stored video library is found to be associated with the video tag information Video, and all the found videos are summarized into candidate video sets; candidate videos are selected from the candidate video sets, and each candidate video has not generated display records on the client, and the selected candidate videos are summarized into recommended video sets. .
- the video feedback information of the server includes a time when the user browses each video in the recommended video set, a historical display state of each video in the recommended video set, and each of the recommended video sets.
- the time of the video and the total time of each video in the recommended video set, the step of establishing a video revenue coefficient calculation model according to the video feedback information includes:
- the step of establishing, by the server, the video revenue coefficient calculation model according to the video feedback information further includes:
- a video revenue coefficient calculation model is established, where earn i,j represents the video revenue coefficient of each video, u represents a single user, and U represents a collection of all users.
- the present application provides another server as an embodiment, which is specifically as follows:
- the server comprising: one or more processors; a memory; one or more applications, wherein the one or more applications are stored in the memory and configured to be processed by the one or more Executing, the one or more programs are configured to: request a second user browsing record in a preset time from a client, where the second user browsing record includes a user browsing recommended video set Each video is at a second time from the current time; according to the video revenue coefficient calculation model provided above, the user browses each video in the recommended video set, the second time from the current time, and the time decay factor to calculate each stored in the video database.
- the video revenue expectation of the video; the video stored in the video database is sent to the client based on the video revenue expectation of each video.
- the server includes a second processor 910, a second memory 920, a second internal memory 930, a second network interface 940, and a second display screen 950 connected by a system bus.
- the second processor 910 is configured to implement a computing function and a function of controlling the operation of the terminal device, and the second processor 910 is configured to perform the video recommendation method provided by the above embodiments.
- the second processor 910 is configured to request, from a client, a preset second user browsing record in a first time, where the second user browsing record includes a video browsing the current time of each video in the recommended video set.
- the second time calculating the model according to the video yield coefficient provided above, the second time of each video of the user browsing the recommended video set from the current time, and the time decay factor to calculate the video revenue expectation value of each video stored in the video database;
- the video stored in the video database is sent to the client based on the video revenue expectation of each video.
- the second memory 920 is a non-volatile storage medium, and the second storage system 921, the second database 922, and a computer program for implementing the video recommendation method provided by the above embodiments, and a candidate intermediate generated by the execution of the computer program Data and result data.
- the second network interface 940 is for communicating with a server, and the second network interface 940 includes a radio frequency transceiver.
- the server calculates the model according to the video revenue coefficient provided above, the second time of each video of the user browsing the recommended video set, and the time decay factor of the current time, and calculates each stored in the video database.
- the steps for video video revenue expectations for a video include:
- the step of the server sending the video stored in the video database to the client according to the video revenue expectation value of each video comprises: storing the video revenue in the video database according to the expected value of the video revenue of each video.
- the videos are arranged in descending order; the videos stored in the video database are sent to the client in the arranged order.
- the application further provides a computer readable storage medium carrying one or more computer instruction programs thereon, the one or more processors executing one or more processors executing a read-only separation
- the method for downloading the speed of the mode includes: requesting the network download and waiting for the connection to succeed; reading the network file stream after the connection is successful; creating the first write operation task, and executing by the pre-started first thread queue; creating the second write operation The task is executed by a pre-started second thread queue.
- the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed.
- the foregoing storage medium includes: a mobile storage device, a random access memory (RAM), a read-only memory (ROM), a magnetic disk, or an optical disk.
- RAM random access memory
- ROM read-only memory
- magnetic disk or an optical disk.
- optical disk A medium that can store program code.
- the above-described integrated unit of the present application may be stored in a computer readable storage medium if it is implemented in the form of a software function module and sold or used as a stand-alone product.
- the technical solution of the embodiments of the present application may be embodied in the form of a software product in essence or in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for making
- a computer device which may be a personal computer, server, or network device, etc.
- the foregoing storage medium includes various media that can store program codes, such as a mobile storage device, a RAM, a ROM, a magnetic disk, or an optical disk.
- the video revenue calculation modeling apparatus and method provided by the present application firstly extracts and summarizes a video that meets a preset condition according to a first user browsing record from a pre-stored video library into a recommended video set; And sending the recommended video set to the client associated with the first user browsing record; then receiving the video browsing feedback information sent by the client after the preset second time, and finally according to the video feedback
- the information establishes a video yield coefficient calculation model.
- the video recommendation apparatus and method provided by the present application firstly requests a client for a preset second user browsing record in a first time; and then calculates a model according to the video revenue coefficient provided above, and the user browses the recommended video.
- Each video in the set calculates the video revenue expectation value of each video stored in the video database from the second time of the current time and the time attenuation factor; finally, the video stored in the video database is sent to the video according to the video revenue expectation value of each video.
- Said client Since the established video revenue coefficient calculation model fits each user's own video browsing interests and actual conditions, the established video revenue coefficient calculation model has high reliability and reference degree, so the calculation model based on the established video revenue coefficient is recommended for the user.
- the video is also very suitable for the user's video browsing field and browsing interest, greatly improving the accuracy of video recommendation, and also improving the user experience of browsing video.
- each block of the flowchart or block diagram can represent a module, a program segment, or a portion of code that includes one or more of the Executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the blocks may also occur in a different order than those illustrated in the drawings.
- each block of the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented in a dedicated hardware-based system that performs the specified function or function. Or it can be implemented by a combination of dedicated hardware and computer instructions.
- each functional module in each embodiment of the present application may be integrated to form a separate part, or each module may exist separately, or two or more modules may be integrated to form a separate part.
- the functions, if implemented in the form of software functional modules and sold or used as separate products, may be stored in a computer readable storage medium.
- the technical solution of the present application which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including
- the instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present application.
- the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like. .
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
本申请实施例提供了一种视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质,涉及视频播放领域。该视频收益计算建模装置与方法首先通过从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;再将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;然后在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,最后依据所述视频反馈信息建立视频收益系数计算模型。通过根据实时根据用户的浏览反馈信息建立的视频收益系数计算模型,贴合每个用户自身的视频浏览兴趣爱好与实际情况,建立的视频收益系数计算模型可靠性与参考度高。
Description
交互参考
本申请要求以下优先权:2017年05月12日提出的申请号:201710334750.6,名称:“视频收益计算建模装置与方法及视频推荐装置与方法”的中国专利,本申请参考引用了如上所述申请的全部内容。
本申请涉及视频播放领域,具体而言,涉及一种视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质。
短视频是一种互联网内容传播方式,一般是在互联网新媒体上传播的时长在5分钟以内的视频传播内容;随着移动终端普及和网络的提速,短平快的大流量传播内容逐渐获得各大平台、粉丝和资本的青睐。短视频通常在各种新媒体平台上播放,适合在移动状态和短时休闲状态下观看,高频推送的视频内容,时间通常在几秒到几分钟不等。内容融合了技能分享、幽默搞怪、时尚潮流、社会热点、街头采访、公益教育、广告创意、商业定制等主题。
现有技术中,有一类应用为短视频应用,这类应用主要给用户推荐时长较短的短视频,然后这类应用程序为用户展示的视频通常杂乱无章,没有根据用户的兴趣爱好来推荐相应的短视频,用户的体验感差,短视频的收益也不高。
发明内容
有鉴于此,本申请实施例的目的在于提供一种视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质,以改善上述的问题。
第一方面,本申请实施例提供了一种视频收益计算建模装置,所述视频收益计算建模装置包括:
推荐视频集提取单元,用于从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;
第一信息发送单元,用于将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;
信息接收单元,用于在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;
模型建立单元,用于依据所述视频反馈信息建立视频收益系数计算模型。
第二方面,本申请实施例还提供了一种视频收益计算建模方法,所述视频收益计算建模方法包括:
从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;
将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;
在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;
依据所述视频反馈信息建立视频收益系数计算模型。
第三方面,本申请实施例还提供了一种视频推荐装置,所述视频推荐装置包括:
数据请求单元,用于向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;
视频收益期望值计算单元,用于依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;
第二信息发送单元,用于依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
第四方面,本申请实施例还提供了一种视频推荐方法,所述视频推荐方法包括:
向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;
依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;
依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
第五方面,本申请实施例还提供了一种服务器,其包括:
一个或多个处理器;
存储器;
一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配 置用于:
从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;
将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;
在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;
依据所述视频反馈信息建立视频收益系数计算模型。
第六方面,本申请实施例还提供了一种服务器,其包括:
一个或多个处理器;
存储器;
一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配置用于:
向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;
依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;
依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
第七方面,本申请实施例还提供了一种计算机可读存储介质,其上承载一个或多个计算机指令程序,所述计算机指令程序被一个或多个处理器执行时,所述一个或多个处理器执行上述视频收益计算建模方法或上述视 频推荐方法。
与现有技术相比,本申请实施例提供的视频收益计算建模装置与方法,首先通过从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;再将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;然后在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,最后依据所述视频反馈信息建立视频收益系数计算模型。通过根据实时根据用户的浏览反馈信息建立的视频收益系数计算模型,贴合每个用户自身的视频浏览兴趣爱好与实际情况,建立的视频收益系数计算模型可靠性与参考度高。
与现有技术相比,本申请实施例提供的视频推荐装置与方法,首先通过向一客户端请求预设定的第一时间内的第二用户浏览记录;然后依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;最后依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。由于建立的视频收益系数计算模型贴合每个用户自身的视频浏览兴趣爱好与实际情况,建立的视频收益系数计算模型可靠性与参考度高,因此依据建立的视频收益系数计算模型为用户推荐的视频,也非常贴合用户的视频浏览领域与浏览兴趣,大幅度提高了视频推荐的精准度,同时也提高了用户浏览视频的体验感。
为使本申请的上述目的、特征和优点能更明显易懂,下文特举较佳实施例,并配合所附附图,作详细说明如下。
为使本申请实施例的目的、技术方案和优点更加清楚,下面将结合本申请实施例中的附图,对本申请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申请一部分实施例,而不是全部的实施例。通常在此处附图中描述和示出的本申请实施例的组件可以以各种不同的配置来布置和设计。因此,以下对在附图中提供的本申请的实施例的详细描述并非旨在限制要求保护的本申请的范围,而是仅仅表示本申请的选定实施例。基于本申请中的实施例,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,都属于本申请保护的范围。
图1为本申请实施例提供的服务器与客户端之间的交互示意图;
图2为本申请实施例提供的服务器的结构框图;
图3为本申请实施例提供的视频收益计算建模装置功能模块示意图;
图4为本申请实施例提供的推荐视频集在客户端的显示界面的显示状态示意图;
图5为本申请实施例提供的视频收益计算建模方法的流程图;
图6为本申请实施例提供的视频推荐装置的功能模块示意图;
图7为本申请实施例提供的视频推荐方法的流程图;
图8为本申请实施例的一服务器的内部结构框图;
图9为本申请实施例的另一服务器的内部结构框图。
图标:100-客户端;200-服务器;300-网络;400-视频收益计算建模装置;500-视频推荐装置;101-处理器;102-存储器;103-存储控制器;104-外设接口;301-视频集提取单元;302-第一信息发送单元;303-信息接收单元;304-模型建立单元;601-数据请求单元;602-视频收益期望值计算单元;603-排序单元;604-第二信息发送单元;810-第一处理器;820-第一非易失性存储器;821-第一操作系统;822-第一数据库;830-第一内存储器;840-第一网络接口;850-第一显示屏;910-第二处理器;920-第二非易失性存储器;921-第二操作系统;922-第二数据库;930-第二内存储器;940-第二网络接口;950-第二显示屏。
下面将结合本申请实施例中附图,对本申请实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。通常在此处附图中描述和示出的本申请实施例的组件可以以各种不同的配置来布置和设计。因此,以下对在附图中提供的本申请的实施例的详细描述并非旨在限制要求保护的本申请的范围,而是仅仅表示本申请的选定实施例。基于本申请的实施例,本领域技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例,都属于本申请保护的范围。
本申请实施例提出的视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质,本申请较佳实施例所提供的视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质,可应用于如图1所示的应用环境中。如图1所示,客户端100、服务器200位于网络300中,通过该网络300,客户端100与服务器200进行数据交互。于本申请实施例中,客户端100中安装有至少一个应用程序(Application,APP),与服务器200相对应,为用户提供服务。该服务器200可以是,但不限于,网络服务器、数据库服务器、云端服务器等等。该客户端100可以是,但不限于智能手机、个人电脑(personal computer,PC)、平板电脑、个人数字助理(personal digital assistant,PDA)、移动上网设备(mobile Internet device,MID)等。所述客户端100的操作系统可以是,但不限于,安卓(Android)系统、IOS(iPhone operating system)系统、Windows phone系统、Windows系统等。
图2示出了一种可应用于本申请实施例中的服务器200的结构框图。所述服务端包括视频收益计算建模装置400、视频推荐装置500、处理器101、存储器102、存储控制器103以及外设接口104。
所述存储器102、存储控制器103、处理器101以及外设接口104,各元件相互之间直接或间接地电性连接,以实现数据的传输或交互。例如,这些元件相互之间可通过一条或多条通讯总线或信号线实现电性连接。所述视频收益计算建模装置400、视频推荐装置500包括至少一个可以软件或固件(firmware)的形式存储于所述存储器102中或固化在所述客户端100的操作系统(operating system,OS)中的软件功能模块。所述处理器101用于执行存储器102中存储的可执行模块,例如,所述视频收益计算建模装置400、视频推荐装置500包括的软件功能模块或计算机程序。
其中,存储器102可以是,但不限于,随机存取存储器(Random Access Memory,RAM),只读存储器102Read Only Memory,ROM),可编程只读存储器(Programmable Read-Only Memory,PROM),可擦除只读存储器(Erasable Programmable Read-Only Memory,EPROM),电可擦除只读存储器(Electric Erasable Programmable Read-Only Memory,EEPROM)等。其中,存储器102用于存储程序,所述处理器101在接收到执行指令后,执行所述程序,前述本申请实施例任一实施例揭示的流过程定义的服务端所执行的方法可以应用于处理器101中,或者由处理器101实现。
处理器101可能是一种集成电路芯片,具有信号的处理能力。上述的处理器101可以是通用处理器,包括中央处理器(Central Processing Unit,简称CPU)、网络处理器(Network Processor,简称NP)等;还可以是数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器101也可以是任何常规的处理器101等。
外设接口104将各种输入/输出装置耦合至处理器101以及存储器 102。在一些实施例中,外设接口104、处理器101以及存储控制器103可以在单个芯片中实现。在其他一些实例中,他们可以分别由独立的芯片实现。
可以理解,图2所示的结构仅为示意,服务器200还可包括比图2中所示更多或者更少的组件,或者具有与图2所示不同的配置。图2中所示的各组件可以采用硬件、软件或其组合实现。
第一实施例
请参阅图3,本申请实施例提供了一种视频收益计算建模装置400,所述视频收益计算建模装置400包括推荐视频集提取单元301、第一信息发送单元302、信息接收单元303以及模型建立单元304。
推荐视频集提取单元301用于从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集。
具体地,推荐视频集提取单元301可以包括:
数据请求子单元,用于向所述客户端100请求预设定的第一时间内的所述第一用户浏览记录。
其中,所述第一用户浏览记录包括有视频标签信息,本申请实施例中,视频标签的获取途径很多,例如:通过视频的标题提取关键字,并将关键字映射为相应的视频标签信息或者通过运营人员人工对视频打上视频标签信息。视频标签信息可以为“NBA篮球赛”、“美国大选”、“选美时尚秀”等等,在此仅仅是举例说明。本申请实施例中,第一用户浏览记录为用户先前在客户端100浏览过的视频后缓存的浏览记录。本申请实施例中,考虑到不同时间段用户关注的时代热点、时尚潮流类的视频内容不同,因此挑选距离当前时间最近第一用户浏览记录,更贴合用户当前时间的视频浏览偏好,因此,预设定的第一时间优选为3天或一周或半个月。
视频集归纳子单元,用于在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
例如:用户最近一周看了两个视频,视频A是关于“军事”的,视频是关于“篮球”的,则在预存储的视频库中把所有包含“军事”、“篮球”的视频找出来,作为推荐给用户的推荐视频集。
本申请实施例中,较佳地,所述视频集归纳子单元用于在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集。
所述推荐视频集提取单元301还包括:选择子单元,用于从候选视频集中选择候选视频,且每个候选视频均未在所述客户端100产生过显示记录,将选中的候选视频归纳为推荐视频集。
为了让推荐视频集中的视频更符合用户的视频浏览偏好以及更能提起用户的浏览兴趣,因此需要对先前在用户的客户端100显示但未播放过的视频进行剔除(显示但未播放,说明用户对此视频的兴趣度不高),因此将候选视频集中的在用户的客户端100显示但未播放过的视频进行剔除后,作为推荐视频集,在客户端100被用户点击并播放的概率更大,也更贴合用户的浏览偏好。本申请实施例中,推荐视频集中的视频的个数可以为10个、50个、100个等等,具体依据实际情况而定,在此不做限制。
第一信息发送单元302用于将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端100。
需要说明的是,本申请实施例中,与所述第一用户浏览记录关联的客户端100是指产生并缓存有第一用户浏览记录的客户端100。
信息接收单元303用于在预设定的第二时间后接收所述客户端100发送的视频浏览反馈信息。
本申请实施例中,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中。本申请实施例中,当然地,视频反馈信息还可以包括用户浏览 推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的任一一个或多个,在此仅仅是举例说明。
模型建立单元304用于依据所述视频反馈信息建立视频收益系数计算模型。
具体地,所述模型建立单元304包括:视频反馈效果计算模型建立子单元,
其中,feedback
u,i,j为视频反馈效果值,isview
u,j为推荐视频集的每个视频的历史展示状态值;例如,-1表示该视频在用户在客户端100的屏幕上展示了但没有点击进行播放,0表示该视频在客户端100上没有展示,1表示该视频在客户端100展示了且用户有点击进行播放。viewdays
u,i用户浏览推荐视频集中的每个视频距离当前时刻的时间,单位为天,例如:用户当天浏览过视频则取1,用户昨天浏览过视频则取2,以此类推。
其中,
playtime
u,j表示用户播放推荐视频集的每个视频的时间,videotime
j表示推荐视频集的每个视频的总时间,precent
u,j表示用户观看每个视频的完成率。例如,一个军事视频的总时长为5分钟,用户点击播放观看了3分钟,则该军事视频的用户观看 完成率为60%。
lapse为每个视频的时间衰减因子,Ti表示用户浏览过的视频的标签集合,Tj表示推荐视频集中包含的视频的标签集合。其中,lapse为一常数且lapse>0,表示时间衰减因子,lapse为一个人工配置参数,配置参数的值越大,表示越重视用户最近浏览视频的行为,取0时表示用户的所有时间的浏览行为都同等对待。实际业务中一般取1,往后的时间每天的权重依次是
以此类推。例外,考虑
的时候赋值为0的原因是没有相同标签的视频视为不相关。
视频收益计算建模子单元,用于依据算式
本申请实施例中,对批量用户采取上述操作,可以得到大量的视频反馈信息,根据大量的视频反馈信息得到多个视频收益系数。上述的算式中,
所有视频效果值不为0的信号平均值用于衡量短视频的收益,在此基础上加上1则是一个修正值,确保相同标签的视频收益系数都大于或等于0,而没有相同标签的视频之间则认为没有 相关性,视频收益系数也为0。
其中,相同标签的视频收益系数都大于或等于0的证明过程如下:
过程如下,由feedback
u,i,j的定义可知
-1≤feedback
u,i,j
所以
即可得出
该建立好的视频收益系数计算模型,根据用户的实时对推荐视频集的视频反馈信息得出,科学合理,符合用户当前浏览视频的实际情况。
第二实施例
请参阅图5,本申请实施例还提供了一种视频收益计算建模方法,需要说明的是,本申请实施例所提供的视频收益计算建模方法,其基本原理及产生的技术效果和第一实施例相同,为简要描述,本申请实施例部分未提及之处,可参考上述的实施例中相应内容。所述视频收益计算建模方法包括:
步骤S501:从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集。
可以理解地,步骤S501由上述的推荐视频集提取单元301执行。
具体地,步骤S501执行流程包括但不限于以下两种方式:
第一种:向所述客户端100请求预设定的第一时间内的所述第一用户 浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
第二种:向所述客户端100请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集;从候选视频集中选择候选视频,且每个候选视频均未在所述客户端100产生过显示记录,将选中的候选视频归纳为推荐视频集。
步骤S502:将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端100。
可以理解地,步骤S502由上述的第一信息发送单元302执行。
步骤S503:在预设定的第二时间后接收所述客户端100发送的视频浏览反馈信息。
可以理解地,步骤S503由上述的信息接收单元303执行。
本申请实施例中,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间。
步骤S504:依据所述视频反馈信息建立视频收益系数计算模型。
可以理解地,步骤S504由上述的模型建立单元304执行。
具体地,依据建立视频收益系数计算模型的流程为:
首先依据算式
建立视频反馈效果计算模型,其中,feedback
u,i,j为视频反馈效果值,isview
u,j为推荐视频集的每个视频的历史展示状态值,viewdays
u,i用户浏览推荐视频集中的每个视频距离当前时刻的时间,
playtime
u,j表示用户播放推荐视频集的每个视频的时间,videotime
j表示推荐视频集的每个视频的总时间,precent
u,j表示用户观看每个视频的完成率,lapse为每个视频的时间衰减因子,T
i表示用户浏览过的短视频的标签集合,T
j推荐视频集中包含的短视频的标签集合。
然后依据算式
建立视频收益系数计算模型,其中,earn
i,j表示每个视频的视频收益系数。
第三实施例
请参阅图6,本申请实施例还提供了一种视频推荐装置500,所述视频推荐装置500包括数据请求单元601、视频收益期望值计算单元602、排序单元603以及第二信息发送单元604。
数据请求单元601用于向一客户端100请求预设定的第一时间内的第二用户浏览记录。
其中,所述第二用户浏览记录也包括有视频标签信息以及用户浏览推 荐视频集中的每个视频距离当前时刻的第二时间。
本申请实施例中,视频标签的获取途径很多,例如:通过视频的标题提取关键字,并将关键字映射为相应的视频标签信息或者通过运营人员人工对视频打上视频标签信息。本申请实施例中,考虑到不同时间段用户关注的时代热点、时尚潮流类的视频内容不同,因此挑选距离当前时间最近第二用户浏览记录,更贴合用户当前时间的视频浏览偏好,因此,预设定的第二时间也优选为3天或一周或半个月。
视频收益期望值计算单元602用于依据实施例一或实施例二提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值。
具体地,所述视频收益期望值计算单元602用于依据算式
计算出视频收益期望值,其中,viewdays
a,i表示用户浏览推荐视频集中的每个视频距离当前时刻的第二时间、lapse表示时间衰减因子,expect
a,j每个视频的视频收益期望值,earn
i,j表示每个视频的视频收益系数,earn
i,j的计算方式如实施例一或实施例二所描述,在此就不再多做赘述。
排序单元603用于依据每个视频的视频收益期望值将视频数据库中存储的视频进行降序排列。
本申请实施例中,对视频数据库中存储的视频的排序方式不仅仅可以依据每个视频的视频收益期望值降序排列,也可以为其他的一些排列方式,在此就不再举例说明。
所述信息发送单元用于将视频数据库中存储的视频按照排列后的顺序发送至所述客户端100。
在客户端100展示的视频顺序即按照视频收益期望值的先后顺序进行排列,用户在客户端100首先看到视频为最符合用户的观看需求的视频。本申请实施例中,对于信息发送单元发送视频的方式不仅仅可以为上述的方式,也可以为例如,先判断视频收益期望值是否大于预设定阈值,并且仅仅将视频收益期望值大于预设定阈值的视频发送值客户端100。
第四实施例
请参阅图7,本申请实施例还提供了一种视频推荐方法,需要说明的是,本申请实施例所提供的视频推荐方法,其基本原理及产生的技术效果和上述实施例相同,为简要描述,本申请实施例部分未提及之处,可参考上述的实施例中相应内容。所述视频推荐方法包括:
步骤S701:向一客户端100请求预设定的第一时间内的第二用户浏览记录。
其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间。可以理解地,步骤S701由上述的数据请求单元601执行。
步骤S702:依据实施例一或实施例二提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值。
可以理解地,步骤S702由上述的视频收益期望值计算单元602执行。
具体地,步骤S702的具体实施方式为:
依据算式
计算出视频收益 期望值,其中,viewdays
a,i表示用户浏览推荐视频集中的每个视频距离当前时刻的第二时间、lapse表示时间衰减因子,expect
a,j每个视频的视频收益期望值,earn
i,j表示每个视频的视频收益系数。
步骤S703:依据每个视频的视频收益期望值将视频库中存储的视频进行降序排列。
可以理解地,步骤S703由上述的排序单元603执行。
步骤S704:将视频数据库中存储的视频按照排列后的顺序发送至所述客户端100。
可以理解地,步骤S704由上述的第二信息发送单元604执行。
为进一步说明上述实施例视频收益计算建模方法,本申请提供一服务器作为实施例,具体如下:
所述服务器,包括:一个或多个处理器;存储器;一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配置用于:从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;依据所述视频反馈信息建立视频收益系数计算模型。
在本实施例中,如图8所示,所述服务器包括通过系统总线连接的第一处理器810、第一存储器820、第一内存储器830、第一网络接口840和第一显示屏850。第一处理器810用于实现计算功能和控制终 端装置工作的功能,第一处理器810被配置为执行上述实施例提供的视频收益计算建模方法。第一处理器810用于从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;依据所述视频反馈信息建立视频收益系数计算模型。第一存储器820是一种非易失性存储介质,第一存储有操作系统821、第一数据库822和用于实现上述实施例提供的视频收益计算建模方法的计算机程序,以及执行计算机程序产生的候选中间数据以及结果数据。第一网络接口840用于与服务器通信,第一网络接口840包括射频收发器。
在本实施例中,所述服务器的所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
在本实施例中,所述服务器的所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集;从候选视频集中选择候选视频,且每个候选视频均未在所述客户端产生过显示记录,将选中的候选视频归纳为推荐视频集。
在本实施例中,所述服务器的所述视频反馈信息包括用户浏览推荐视频集中的每个视频距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间,所述依据所述视频反馈信息建立视频收益系数计算模型的步骤包括:
依据算式
建立视频反馈效果计算模型,其中,feedback
u,i,j为视频反馈效果值,isview
u,j为推荐视频集的每个视频的历史展示状态值,viewdays
u,i用户浏览推荐视频集中的每个视频距离当前时刻的时间,
playtime
u,j表示用户播放推荐视频集的每个视频的时间,videotime
j表示推荐视频集的每个视频的总时间,precent
u,j表示用户观看每个视频的完成率,lapse为每个视频的时间衰减因子,Ti表示用户浏览过的视频的标签集合,Tj表示推荐视频集中包含的视频的标签集合。
在本实施例中,所述服务器的所述依据所述视频反馈信息建立视频收益系数计算模型的步骤还包括:依据算式
建立视频收益系数计算模型,其中,earn
i,j表示每个视频的视频收益系 数,u表示单个用户,U表示所有用户的集合。
为进一步说明上述实施例视频推荐方法,本申请提供另一服务器作为实施例,具体如下:
所述服务器,包括:一个或多个处理器;存储器;一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配置用于:向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
在本实施例中,如图9所示,所述服务器包括通过系统总线连接的第二处理器910、第二存储器920、第二内存储器930、第二网络接口940和第二显示屏950。第二处理器910用于实现计算功能和控制终端装置工作的功能,第二处理器910被配置为执行上述实施例提供的视频推荐方法。第二处理器910用于向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。第二存储器920是一种非易失性存储介质,第二存储有操作系统921、第二数据库922和用于实现上述实施例提供的视频推荐方法的计算机程序,以及执行计算机程序产生的候选中间数据以及结果数据。第二网络接口940用于与服务器通信,第二网络接口940包 括射频收发器。
在本实施例中,所述服务器的所述依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值的步骤包括:
依据算式
计算出视频收益期望值,其中,viewdays
a,i表示用户浏览推荐视频集中的每个视频距离当前时刻的第二时间、lapse表示时间衰减因子,expect
a,j每个视频的视频收益期望值,earn
i,j表示每个视频的视频收益系数。
在本实施例中,所述服务器的所述依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端的步骤包括:依据每个视频的视频收益期望值将视频数据库中存储的视频进行降序排列;将视频数据库中存储的视频按照排列后的顺序发送至所述客户端。
本申请还提供一种计算机可读存储介质,其上承载一个或多个计算机指令程序,计算机指令程序被一个或多个处理器执行时,一个或多个处理器执行实现一种基于读写分离模式的下载提速方法,包括:请求网络下载并等待连接成功;连接成功后读取网络文件流;创建第一写入操作任务,并由预先启动的第一线程队列执行;创建第二写入操作任务,并由预先启动的第二线程队列执行。
本领域普通技术人员可以理解:实现上述方法实施例的全部或部分步骤可以通过程序指令相关的硬件来完成,前述的程序可以存储于一计算机可读取存储介质中,该程序在执行时,执行包括上述任意方法实施 例的步骤;而前述的存储介质包括:移动存储设备、随机存取存储器(RAM,Random Access Memory)、只读存储器(ROM,Read-Only Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
或者,本申请上述集成的单元如果以软件功能模块的形式实现并作为独立的产品销售或使用时,也可以存储在一个计算机可读取存储介质中。基于这样的理解,本申请实施例的技术方案本质上或者说对相关技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机、服务器、或者网络设备等)执行申请各个实施例所述方法的全部或部分。而前述的存储介质包括:移动存储设备、RAM、ROM、磁碟或者光盘等各种可以存储程序代码的介质。
综上所述,本申请提供的视频收益计算建模装置与方法,首先通过从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;再将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;然后在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,最后依据所述视频反馈信息建立视频收益系数计算模型。通过根据实时根据用户的浏览反馈信息建立的视频收益系数计算模型,贴合每个用户自身的视频浏览兴趣爱好与实际情况,建立的视频收益系数计算模型可靠性与参考度高。
再者,本申请提供的视频推荐装置与方法,首先通过向一客户端请求预设定的第一时间内的第二用户浏览记录;然后依据上述提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;最后依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。由于建立的视频收益系数计算模型贴合每个用户自身的视频浏览兴趣爱好与实际情况,建立的视频收益系数计算模型可靠性与参 考度高,因此依据建立的视频收益系数计算模型为用户推荐的视频,也非常贴合用户的视频浏览领域与浏览兴趣,大幅度提高了视频推荐的精准度,同时也提高了用户浏览视频的体验感。
在本申请所提供的几个实施例中,应该理解到,所揭露的装置和方法,也可以通过其它的方式实现。以上所描述的装置实施例仅仅是示意性的,例如,附图中的流程图和框图显示了根据本申请的多个实施例的装置、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上,流程图或框图中的每个方框可以代表一个模块、程序段或代码的一部分,所述模块、程序段或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意,在有些作为替换的实现方式中,方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如,两个连续的方框实际上可以基本并行地执行,它们有时也可以按相反的顺序执行,这依所涉及的功能而定。也要注意的是,框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合,可以用执行规定的功能或动作的专用的基于硬件的系统来实现,或者可以用专用硬件与计算机指令的组合来实现。
另外,在本申请各个实施例中的各功能模块可以集成在一起形成一个独立的部分,也可以是各个模块单独存在,也可以两个或两个以上模块集成形成一个独立的部分。
所述功能如果以软件功能模块的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM, Read-Only Memory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上所述仅为本申请的优选实施例而已,并不用于限制本申请,对于本领域的技术人员来说,本申请可以有各种更改和变化。凡在本申请的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本申请的保护范围之内。应注意到:相似的标号和字母在下面的附图中表示类似项,因此,一旦某一项在一个附图中被定义,则在随后的附图中不需要对其进行进一步定义和解释。
Claims (25)
- 一种视频收益计算建模装置,所述视频收益计算建模装置包括:推荐视频集提取单元,用于从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;第一信息发送单元,用于将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;信息接收单元,用于在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;模型建立单元,用于依据所述视频反馈信息建立视频收益系数计算模型。
- 根据权利要求1所述的视频收益计算建模装置,所述推荐视频集提取单元包括:数据请求子单元,用于向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;视频集归纳子单元,用于在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
- 根据权利要求1所述的视频收益计算建模装置,所述视频集归纳子单元用于在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集;所述推荐视频集提取单元还包括:选择子单元,用于从候选视频集中选择候选视频,且每个候选视频均未在所述客户端产生过显示记录,将选中的候选视频归纳为推荐视频集。
- 根据权利要求1所述的视频收益计算建模装置,所述视频反馈信息包括用户浏览推荐视频集中的每个视频距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间,所述模型建立单元包括:视频反馈效果计算模型建立子单元,
- 一种视频收益计算建模方法,所述视频收益计算建模方法包括:从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;依据所述视频反馈信息建立视频收益系数计算模型。
- 根据权利要求6所述的视频收益计算建模方法,所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
- 根据权利要求6所述的视频收益计算建模方法,所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集;从候选视频集中选择候选视频,且每个候选视频均未在所述客户端产生过显示记录,将选中的候选视频归纳为推荐视频集。
- 根据权利要求6所述的视频收益计算建模方法,所述视频反馈信息包括用户浏览推荐视频集中的每个视频距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间,所述依据所述视频反馈信息建立视频收益系数计算模型的步骤包括:
- 一种视频推荐装置,所述视频推荐装置包括:数据请求单元,用于向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;视频收益期望值计算单元,用于依据权利要求1~5任一项提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;第二信息发送单元,用于依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
- 根据权利要求11所述的视频推荐装置,所述视频推荐装置还包 括:排序单元,用于依据每个视频的视频收益期望值将视频数据库中存储的视频进行降序排列;所述信息发送单元用于将视频数据库中存储的视频按照排列后的顺序发送至所述客户端。
- 一种视频推荐方法,所述视频推荐方法包括:向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;依据权利要求6~10任一项提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
- 根据权利要求14所述的视频推荐方法,所述依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端的步骤包括:依据每个视频的视频收益期望值将视频数据库中存储的视频进行降序排列;将视频数据库中存储的视频按照排列后的顺序发送至所述客户端。
- 一种服务器,其包括:一个或多个处理器;存储器;一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配置用于:从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集;将所述推荐视频集发送至与所述第一用户浏览记录关联的客户端;在预设定的第二时间后接收所述客户端发送的视频浏览反馈信息,所述视频反馈信息包括用户浏览推荐视频集中的每个视频分别距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间中的至少之一;依据所述视频反馈信息建立视频收益系数计算模型。
- 根据权利要求17所述的所述服务器,所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为推荐视频集。
- 根据权利要求17所述的所述服务器,所述从预存储的视频库中,依据第一用户浏览记录将符合预设定条件的视频提取并归纳为推荐视频集的步骤包括:向所述客户端请求预设定的第一时间内的所述第一用户浏览记录,其中,所述第一用户浏览记录包括有视频标签信息;在预存储的视频库中查找出与所述视频标签信息关联的视频,并将查找到的所有视频归纳为候选视频集;从候选视频集中选择候选视频,且每个候选视频均未在所述客户端产生过显示记录,将选中的候选视频归纳为推荐视频集。
- 根据权利要求17所述的所述服务器,所述视频反馈信息包括用户浏览推荐视频集中的每个视频距离当前时刻的时间、推荐视频集中的每个视频的历史展示状态、播放推荐视频集中的每个视频的时间以及推荐视频集中的每个视频的总时间,所述依据所述视频反馈信息建立视频收益系数计算模型的步骤包括:
- 一种服务器,其包括:一个或多个处理器;存储器;一个或多个应用程序,其中所述一个或多个应用程序被存储在所述存储器中并被配置为由所述一个或多个处理器执行,所述一个或多个程序配置用于:向一客户端请求预设定的第一时间内的第二用户浏览记录,其中,所述第二用户浏览记录包括有用户浏览推荐视频集中的每个视频距离当前时刻的第二时间;依据权利要求6~10任一项提供的视频收益系数计算模型、用户浏览推荐视频集中的每个视频距离当前时刻的第二时间以及时间衰减因子计算出视频数据库中存储的每个视频的视频收益期望值;依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端。
- 根据权利要求22所述的所述服务器,所述依据每个视频的视频收益期望值将视频数据库中存储的视频发送至所述客户端的步骤包括:依据每个视频的视频收益期望值将视频数据库中存储的视频进行降序排列;将视频数据库中存储的视频按照排列后的顺序发送至所述客户端。
- 一种计算机可读存储介质,其上承载一个或多个计算机指令程序,所述计算机指令程序被一个或多个处理器执行时,所述一个或多个处理器执行权利要求6至10或14至16任一项所述的方法。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710334750.6A CN107341172B (zh) | 2017-05-12 | 2017-05-12 | 视频收益计算建模装置与方法及视频推荐装置与方法 |
CN201710334750.6 | 2017-05-12 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2018205642A1 true WO2018205642A1 (zh) | 2018-11-15 |
Family
ID=60221227
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2017/120178 WO2018205642A1 (zh) | 2017-05-12 | 2017-12-29 | 视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN107341172B (zh) |
WO (1) | WO2018205642A1 (zh) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107341172B (zh) * | 2017-05-12 | 2020-06-19 | 阿里巴巴(中国)有限公司 | 视频收益计算建模装置与方法及视频推荐装置与方法 |
TWI779889B (zh) * | 2021-10-20 | 2022-10-01 | 中華電信股份有限公司 | 一種串流動態銷售裝置、系統及方法 |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104486649A (zh) * | 2014-12-18 | 2015-04-01 | 北京百度网讯科技有限公司 | 视频内容评级方法及装置 |
CN104994408A (zh) * | 2015-06-25 | 2015-10-21 | 青岛海信电器股份有限公司 | 一种智能电视节目推荐方法、装置及智能电视 |
CN105183856A (zh) * | 2015-09-08 | 2015-12-23 | 百度在线网络技术(北京)有限公司 | 一种评价信息内容质量的方法及装置 |
CN105320766A (zh) * | 2015-10-28 | 2016-02-10 | 百度在线网络技术(北京)有限公司 | 信息推送方法和装置 |
CN105872629A (zh) * | 2016-03-18 | 2016-08-17 | 合网络技术(北京)有限公司 | 内容推荐方法、装置及系统 |
CN106294601A (zh) * | 2016-07-28 | 2017-01-04 | 腾讯科技(深圳)有限公司 | 数据处理方法和装置 |
CN106294830A (zh) * | 2016-08-17 | 2017-01-04 | 合智能科技(深圳)有限公司 | 多媒体资源的推荐方法及装置 |
US20170055014A1 (en) * | 2015-08-21 | 2017-02-23 | Vilynx, Inc. | Processing video usage information for the delivery of advertising |
CN106469210A (zh) * | 2016-09-02 | 2017-03-01 | 腾讯科技(深圳)有限公司 | 媒体类别标签的展示方法和装置 |
CN106534902A (zh) * | 2016-12-14 | 2017-03-22 | 北京数码视讯软件技术发展有限公司 | 一种行为分析方法及系统 |
CN106570720A (zh) * | 2016-09-20 | 2017-04-19 | 中央电视台 | 多媒体数据定价处理方法及装置 |
CN107341172A (zh) * | 2017-05-12 | 2017-11-10 | 广州优视网络科技有限公司 | 视频收益计算建模装置与方法及视频推荐装置与方法 |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20160098797A (ko) * | 2015-02-11 | 2016-08-19 | 삼성전자주식회사 | 영상처리장치, 영상처리장치의 제어방법 및 시스템 |
-
2017
- 2017-05-12 CN CN201710334750.6A patent/CN107341172B/zh active Active
- 2017-12-29 WO PCT/CN2017/120178 patent/WO2018205642A1/zh active Application Filing
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104486649A (zh) * | 2014-12-18 | 2015-04-01 | 北京百度网讯科技有限公司 | 视频内容评级方法及装置 |
CN104994408A (zh) * | 2015-06-25 | 2015-10-21 | 青岛海信电器股份有限公司 | 一种智能电视节目推荐方法、装置及智能电视 |
US20170055014A1 (en) * | 2015-08-21 | 2017-02-23 | Vilynx, Inc. | Processing video usage information for the delivery of advertising |
CN105183856A (zh) * | 2015-09-08 | 2015-12-23 | 百度在线网络技术(北京)有限公司 | 一种评价信息内容质量的方法及装置 |
CN105320766A (zh) * | 2015-10-28 | 2016-02-10 | 百度在线网络技术(北京)有限公司 | 信息推送方法和装置 |
CN105872629A (zh) * | 2016-03-18 | 2016-08-17 | 合网络技术(北京)有限公司 | 内容推荐方法、装置及系统 |
CN106294601A (zh) * | 2016-07-28 | 2017-01-04 | 腾讯科技(深圳)有限公司 | 数据处理方法和装置 |
CN106294830A (zh) * | 2016-08-17 | 2017-01-04 | 合智能科技(深圳)有限公司 | 多媒体资源的推荐方法及装置 |
CN106469210A (zh) * | 2016-09-02 | 2017-03-01 | 腾讯科技(深圳)有限公司 | 媒体类别标签的展示方法和装置 |
CN106570720A (zh) * | 2016-09-20 | 2017-04-19 | 中央电视台 | 多媒体数据定价处理方法及装置 |
CN106534902A (zh) * | 2016-12-14 | 2017-03-22 | 北京数码视讯软件技术发展有限公司 | 一种行为分析方法及系统 |
CN107341172A (zh) * | 2017-05-12 | 2017-11-10 | 广州优视网络科技有限公司 | 视频收益计算建模装置与方法及视频推荐装置与方法 |
Also Published As
Publication number | Publication date |
---|---|
CN107341172A (zh) | 2017-11-10 |
CN107341172B (zh) | 2020-06-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11290413B2 (en) | Trend detection for content targeting using an information distribution system | |
US9595053B1 (en) | Product recommendation using sentiment and semantic analysis | |
US10305851B1 (en) | Network-based content discovery using messages of a messaging platform | |
US9936330B2 (en) | Methods for exchanging data amongst mobile applications using superlinks | |
JP6334696B2 (ja) | ハッシュタグおよびコンテンツ提示 | |
US9990436B2 (en) | Personal trends module | |
US20200356572A1 (en) | Search ranking method and apparatus, electronic device and storage medium | |
CN107870984A (zh) | 识别搜索词的意图的方法和装置 | |
JP2016531518A (ja) | 動的電話番号割り当て | |
CN108540508B (zh) | 用于推送信息的方法、装置和设备 | |
CN109862100B (zh) | 用于推送信息的方法和装置 | |
US10922722B2 (en) | System and method for contextual video advertisement serving in guaranteed display advertising | |
US9367861B2 (en) | Timer-based ad placement in content retrieval applications | |
US20160180374A1 (en) | Viewable impressions system | |
WO2022242428A1 (zh) | 一种信息处理方法、装置、设备、存储介质及程序产品 | |
US10062101B2 (en) | Maximizing the propensity of content consumption according to content complexity and content consumption time | |
WO2020151548A1 (zh) | 关注页面的排序方法及装置 | |
CN111699487A (zh) | 用于快速且安全的内容提供的系统 | |
CN110866040A (zh) | 用户画像生成方法、装置和系统 | |
CN110347900B (zh) | 一种关键词的重要度计算方法、装置、服务器及介质 | |
WO2018205642A1 (zh) | 视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质 | |
CN109981712B (zh) | 用于推送信息的方法和装置 | |
CN114461919A (zh) | 信息推荐的模型训练方法及装置 | |
US9665890B1 (en) | Determining lookback windows | |
CN112131502A (zh) | 数据处理方法、数据处理装置、电子设备和介质 |
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: 17908962 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: 17908962 Country of ref document: EP Kind code of ref document: A1 |