CN110046263A

CN110046263A - Multimedia recommendation method, device, server and storage medium

Info

Publication number: CN110046263A
Application number: CN201910332578.XA
Authority: CN
Inventors: 董治; 李深远
Original assignee: Tencent Music Entertainment Technology Shenzhen Co Ltd
Current assignee: Tencent Music Entertainment Technology Shenzhen Co Ltd
Priority date: 2019-04-23
Filing date: 2019-04-23
Publication date: 2019-07-23
Anticipated expiration: 2039-04-23
Also published as: CN110046263B

Abstract

The embodiment of the invention discloses a kind of multimedia recommendation method, device, server and storage mediums, wherein method includes: the candidate cover set for obtaining multimedia album and the multimedia album, the multimedia album includes at least two multi-media segments, and candidate's cover set includes the corresponding cover of each multi-media segment in the multimedia album；User's portrait is obtained, and determines the target cover of the multimedia album from the candidate cover set according to user portrait；The multimedia album and the target cover are exported, so that showing the target cover in user interface when recommending the multimedia album；Multimedia album can preferably be recommended, improve the clicking rate and playback volume of multimedia album.

Description

Multimedia recommendation method, device, server and storage medium

Technical field

The present invention relates to field of computer technology, and in particular to multi-media processing technical field more particularly to a kind of more matchmakers Body recommended method, a kind of Multimedia Recommendation device, a kind of server and a kind of computer storage medium.

Background technique

Multimedia album refers to: the multi-media segment in multiple and different multimedias being carried out to splicing in sequence and is formed by One multimedia comprising this multiple multi-media segment, so-called multimedia is the synthesis of media, generally comprises text, sound And the medias form such as image.When using terminal plays multimedia, terminal can be pushed away in user interface to user user Recommend multimedia album.If terminal detects the clicking operation of the multimedia album in user to user interface, can sequentially broadcast Put each multi-media segment in multimedia album.

In realizing process of the present invention, inventor has found that existing multimedia recommendation method is usually: any to choose one Default cover of the cover as multimedia album, shows the cover in the user interface to recommend the multimedia album；It is such Multimedia recommendation method can not preferably attract user click multimedia album, so as to cause multimedia album clicking rate and broadcast It is high-volume lower.

Summary of the invention

The embodiment of the invention provides a kind of multimedia recommendation method, device, server and storage mediums, can preferably push away Multimedia album is recommended, the clicking rate and playback volume of multimedia album are improved.

On the one hand, the embodiment of the invention provides a kind of multimedia recommendation methods, comprising:

The candidate cover set of multimedia album and the multimedia album is obtained, the multimedia album includes at least Two multi-media segments, candidate's cover set include the corresponding cover of each multi-media segment in the multimedia album；

User's portrait is obtained, and the multimedia album is determined from the candidate cover set according to user portrait Target cover, user portrait draws a portrait the multimedia hobby of indicated user for reflecting the user；

Export the multimedia album and the target cover so that when recommending the multimedia album, with Target cover described in the interface display of family.

On the other hand, a kind of multimedia recommendation method provided in an embodiment of the present invention, comprising:

The candidate cover set of multimedia album and the multimedia album is obtained, the multimedia album includes at least The multi-media segment of two different multimedias, candidate's cover set includes each multi-media segment pair in the multimedia album The cover answered；

User's portrait is obtained, and is drawn a portrait according to the user and determines object element, the object element is described for reflecting The multimedia hobby of the indicated user of user's portrait；

The destination multimedia segment to match with the object element is chosen from the multimedia album, and in the time Select the target cover for selecting the corresponding cover of the destination multimedia segment as the multimedia album in cover set；

In another aspect, the embodiment of the invention provides a kind of Multimedia Recommendation devices, comprising:

Acquiring unit, for obtaining the candidate cover set of multimedia album and the multimedia album, more matchmakers Body album includes at least two multi-media segments, and candidate's cover set includes each multi-media segment in the multimedia album Corresponding cover；

The acquiring unit, for obtaining user's portrait, and according to user portrait from the candidate cover set Determine the target cover of the multimedia album, user's portrait is for reflecting more matchmakers of the indicated user of user's portrait Body hobby；

Output unit, for exporting the multimedia album and the target cover, so that recommending more matchmakers When body album, the target cover is shown in user interface.

Acquiring unit, for obtaining the candidate cover set of multimedia album and the multimedia album, more matchmakers Body album includes the multi-media segment of at least two different multimedias, and candidate's cover set includes in the multimedia album The corresponding cover of each multi-media segment；

Acquiring unit for obtaining user's portrait, and draws a portrait according to the user and determines object element, the object element For reflecting the multimedia hobby of the indicated user of user's portrait；

Selection unit, for choosing the destination multimedia piece to match with the object element from the multimedia album Section, and select the corresponding cover of the destination multimedia segment as the multimedia album in the candidate cover set Target cover；

In another aspect, the embodiment of the present invention provides a kind of server, including processor, transceiver and memory, the place It manages device, transceiver and memory to be connected with each other, wherein the memory is for storing computer program, the computer program Including the first program instruction, the processor is configured for calling first program instruction, executes following steps:

Alternatively, the computer program includes the second program instruction, the processor is configured for calling described second Program instruction executes following steps:

In another aspect, the embodiment of the present invention provides a kind of computer storage medium, which is stored with One computer program instructions, first computer program instructions are suitable for being loaded by processor and executing following steps:

Alternatively, the computer storage medium is stored with second computer program instruction, the second computer program instruction is suitable In being loaded by processor and execute following steps:

The embodiment of the present invention when recommending multimedia album, then draw a portrait according to user from time by available user's portrait Select the target cover that multimedia album to be recommended is determined in cover set；Since user draws a portrait the multimedia for reflecting user Like, therefore the hobby of user can be met to a certain extent according to target cover determined by user's portrait, so that User's click play multimedia album can be attracted when user interface shows the target cover, and then multimedia can be improved The clicking rate and playback volume of album.

Detailed description of the invention

Technical solution in order to illustrate the embodiments of the present invention more clearly, below will be to needed in embodiment description Attached drawing is briefly described, it should be apparent that, drawings in the following description are some embodiments of the invention, general for this field For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.

Fig. 1 a is a kind of structural schematic diagram of interactive system provided in an embodiment of the present invention；

Fig. 1 b is the schematic diagram of a kind of multimedia album provided in an embodiment of the present invention and candidate cover set；

Fig. 1 c is a kind of schematic diagram of user interface provided in an embodiment of the present invention；

Fig. 2 is a kind of flow diagram of multimedia recommendation method provided in an embodiment of the present invention；

Fig. 3 be another embodiment of the present invention provides a kind of multimedia recommendation method flow diagram；

Fig. 4 is a kind of method schematic diagram for being inserted into transition animation provided in an embodiment of the present invention；

Fig. 5 a is a kind of method schematic diagram of determining target cover provided in an embodiment of the present invention；

Fig. 5 b is another method schematic diagram for determining target cover provided in an embodiment of the present invention；

Fig. 6 be another embodiment of the present invention provides a kind of multimedia recommendation method flow diagram；

Fig. 7 is the structural schematic diagram that the present invention implements a kind of Multimedia Recommendation device provided；

Fig. 8 is a kind of structural schematic diagram of another Multimedia Recommendation device for implementing to provide of the present invention；

Fig. 9 is a kind of structural schematic diagram of server provided in an embodiment of the present invention.

Specific embodiment

Cover refers to the image that can represent multimedia entire content；Multimedia cover can be used for not broadcasting in multimedia Show the multimedia entire content when putting, to attract user's click play multimedia.Therefore, recommending multimedia to user When album, the cover of available multimedia album, and the cover of multimedia album is showed into user in user interface, thus Attract user's click play multimedia album.Based on this, the embodiment of the present invention proposes a kind of more matchmakers for multimedia album Body suggested design can determine the target cover for meeting user preferences using the Multimedia Recommendation scheme, so that it is more to improve this The clicking rate and playback volume of media album.The Multimedia Recommendation scheme can be executed by server or terminal, server herein It can be the service equipment for carrying out multimedia-data procession, can be data processing server, web server etc.； Server can be an independent service equipment, be also possible to the cluster device collectively formed by multiple service equipments；Herein Terminal can be the portable devices and desktop computer etc. such as smart phone, laptop computer, tablet computer.

By taking server executes the Multimedia Recommendation scheme as an example: user is during using terminal plays multimedia, eventually End can send a multimedia request to server, as shown in Figure 1a.Server is asked receiving the multimedia for carrying out self terminal After asking, the candidate cover set of available multimedia album and multimedia album.Multimedia album includes at least two The multi-media segment of different multimedia, each multi-media segment sequential in the multimedia album, to constitute the multimedia Album；Including the corresponding cover of multi-media segment each in multimedia album, the cover of the multi-media segment in candidate cover set For representing the entire content of the multi-media segment.It as shown in Figure 1 b, include three multi-media segments in multimedia album, respectively It is: multi-media segment a, multi-media segment b and multi-media segment c；Correspondingly, including these three more matchmakers in candidate cover set The cover of body segment is respectively: cover 1, cover 2 and cover 3.

Server can also obtain user's portrait of terminal, and user's portrait refers to the picture of the user using the terminal Picture, user portrait is for reflecting the multimedia hobby of the indicated user of user portrait；Such as: may include in user's portrait History multimedia information analyzes the history multimedia information and knows that user once played multimedia A and multimedia A Broadcasting frequency it is larger, then can predict that user likes multimedia A, i.e., the user in the case of this draws a portrait reflected multimedia Hobby is " multimedia A ".After obtaining user's portrait, it can be drawn a portrait according to user and determine to accord with from candidate cover set The target cover of user preferences is closed, and the target cover and multimedia album are sent to terminal.Terminal is receiving more matchmakers After body album and target cover, which can be stored in memory, and show the target in user interface Cover is to recommend the multimedia album.As illustrated in figure 1 c, it determines to meet user from candidate cover set according to user's portrait The target cover of hobby is cover 2, then can be in user interface display cover 2 to attract user's click play multimedia to close Volume.

It should be noted that in other embodiments, the Multimedia Recommendation scheme also can be performed in terminal.Specifically, terminal A multimedia request can be sent to server, with request server transmitting multimedia album.It is sent getting server Multimedia album after, it may be determined that the candidate cover set of the multimedia album.Terminal can also obtain user's portrait, and root The target cover of multimedia album is determined from candidate cover set according to user's portrait；Then it is closed in user interface output multimedia Volume and target cover.

It can be seen that the embodiment of the present invention when recommending multimedia album, can first obtain user portrait, then according to Family portrait determines the target cover of multimedia album to be recommended from candidate cover set；Since user's portrait can be used for reflecting The multimedia of user is liked, therefore can meet the happiness of user to a certain extent according to target cover determined by user's portrait It is good, so that user's click play multimedia album can be attracted when user interface shows the target cover, Jin Erke To improve the clicking rate and playback volume of multimedia album.

Based on the description above, the embodiment of the present invention proposes a kind of exemplary flow of multimedia recommendation method in Fig. 2 Figure, the multimedia recommendation method can be realized by above-mentioned mentioned server or terminal.For ease of description, this hair Bright embodiment is by taking server executes the multimedia recommendation method as an example.The multimedia recommendation method may include steps of S201-S203:

S201 obtains the candidate cover set of multimedia album and multimedia album.

Multimedia album herein may include the multi-media segment of at least two different multimedias, candidate's cover set packet Include the corresponding cover of each multi-media segment in multimedia album.Wherein, the multi-media segment in multimedia album are as follows: pre- from meeting If the segment intercepted in the multimedia of condition, the multimedia for meeting preset condition herein be can include but is not limited to: sound Pleasure, video, audio-video and text etc.；Wherein, audio-video may include: MV (Music Video, music video)；In order to Convenient for description, the subsequent mentioned multimedia of the embodiment of the present invention is by taking MV as an example.

Preset condition may include any one of following: playback volume is greater than default playback volume threshold value, effective broadcasting time is greater than Preset times threshold value, clicking rate are greater than default clicking rate threshold value.Wherein, playback volume threshold value, preset times threshold value and pre- are preset If clicking rate threshold value can be arranged based on experience value；Effective broadcasting time refers to: multimedia number is completely played, it is so-called complete Broadcasting refers to that playing duration is equal to multimedia total duration.For example, the multimedia total duration of multimedia A is 300 seconds, if multimedia A Playing duration be 200 seconds, then it is assumed that this time do not play multimedia A completely, then this time play cannot be recorded as effectively It plays；If the playing duration of multimedia A is 300 seconds, then it is assumed that this time multimedia A is completely played, then this can be broadcast It puts and is recorded as primary effectively broadcasting.In other embodiments, complete play may also mean that: playing duration and multimedia total duration Difference be less than preset difference value, preset difference value such as preset difference value can be arranged based on experience value or actual demand setting It is 10 seconds.In one embodiment, multi-media segment may include: be truncated to from the multimedia for meet preset condition can be with For representing the segment of the multimedia theme for meeting preset condition, such as climax segment, wonderful；Alternatively, should Meet any segment in the multimedia of preset condition.

The cover of multi-media segment may include: static cover or dynamic cover.So-called static state cover refers to static map It, can be by a frame image construction as the cover of format；So-called dynamic cover refers to the cover of moving image format, can be by multiple image It constitutes.In one embodiment, the cover that static cover is constituted for the frame image that intercepts from multi-media segment, dynamic seal The cover that face is constituted for the multiple image that intercepts from multi-media segment.It should be noted that in other embodiments, it is static Cover can be the cover of a frame image construction associated with multi-media segment: for example static cover can be from multimedia piece The cover that the frame image intercepted in multimedia corresponding to section is constituted；Alternatively, static cover can be according to multimedia piece The cover that a frame image made by the theme of section is constituted.Similarly, in other embodiments, dynamic cover can be with The cover that the associated multiple image of multi-media segment is constituted.

Server can obtain multimedia album and multimedia after detecting the trigger event that multimedia album is recommended The candidate cover set of album.Wherein, trigger event can include but is not limited to: receive the thing of the multimedia request of terminal Part detects event, the event for recommending the period to reach of multimedia album that the multimedia database in server updates；Its In, it is multimedia that period arrival is recommended to refer to: between the time between current time and the last time for recommending multimedia album Every the event for being equal to prefixed time interval, which is arranged according to the recommendation period, such as can be set and recommend the period It is 7 days, then prefixed time interval is 7 days.

S202 obtains user's portrait, and determines the target of multimedia album from candidate cover set according to user's portrait Cover, user portrait is for reflecting the multimedia hobby of the indicated user of user's portrait.

The user identifier of the available target terminal of server obtains associated user according to the user identifier and draws a portrait； Target terminal herein may include: any terminal in the terminal set being connected with server, or send to server The terminal of multimedia request；User's portrait may include at least one of following information: the more matchmakers of history multimedia information, history Body searches for information and history multimedia collection information.

Since user's portrait can reflect the multimedia hobby of user, can be drawn a portrait according to user from candidate cover collection The target cover for meeting user preferences is determined in conjunction.It determines to meet user specifically, server can draw a portrait according to user Then the destination multimedia segment of hobby is determined to meet user's happiness from candidate cover set according to the destination multimedia segment Good target cover.

S203 exports multimedia album and target cover, so that when recommending multimedia album, it is aobvious in user interface Show target cover.

Server, can be defeated by the multimedia album and target cover after the target cover of multimedia album has been determined Out to target terminal.Target terminal can show this in user interface after receiving the multimedia album and target cover Target cover, to attract user's click play multimedia album.If terminal is detected to the target cover of multimedia album Clicking operation then plays the multimedia album.From the foregoing it will be appreciated that target terminal can be the terminal set being connected with server In any terminal, be also possible to server send multimedia request terminal.Therefore, server can be actively by multimedia Album and target cover are sent to any terminal, allow any terminal to show the target cover to recommend the multimedia to close Volume；Multimedia request can also received and then be sent to multimedia album and target cover more to server transmission The terminal of media request, so that the terminal can show the target cover to recommend the multimedia album.

In one embodiment, server is when exporting multimedia album and target cover to target terminal, Ke Yiwei Multimedia album creates an album mark, and by the multimedia album and album mark associated storage in multimedia database In；Then the album mark and target cover of multimedia album are exported to target terminal.Correspondingly, target terminal is receiving After album mark and target cover, the target cover can be shown in user interface.If detecting, the click to target cover is grasped Make, then multimedia album is obtained from server according to album mark, and play the multimedia album.Due in album mark It is less to deposit occupancy, therefore when recommending multimedia album, multimedia album is stored in the multimedia database of server, Album mark is only sent to target terminal, the memory of target terminal can be saved, improve the speed of service of target terminal, thus The playing efficiency that target terminal plays multimedia album is improved, user experience is improved.

Based on the description above, the embodiment of the present invention proposes the signal stream of another multimedia recommendation method in Fig. 3 Cheng Tu, the multimedia recommendation method can be realized by above-mentioned mentioned server or terminal.For ease of description, originally Inventive embodiments are by taking server executes the multimedia recommendation method as an example.The multimedia recommendation method may include steps of S301-S306:

S301 obtains the candidate cover set of multimedia album and multimedia album.

It in the specific implementation process, include multiple to be recommended it is possible, firstly, to obtain multimedia collections, in the multimedia collections Multimedia.Server can count each multimedia playback volume in preset time period, and preset time period can be according to business Demand setting, such as setting preset time period are as follows: on the basis of current time, calculate one week period constituted forward, such as Current time is December 10, then preset time period is December 3 to December 9；It is true according to each multimedia playback volume Fixed candidate's multimedia, the multimedia playback volume of candidate are greater than default playback volume threshold value；Using candidate multimedia as to be recommended Multimedia；Multimedia collections are constructed using the multimedia to be recommended.In one embodiment, if candidate multimedia quantity compared with It is more, then the candidate multimedia conduct of preset quantity can also be chosen from candidate multimedia according to the sequence of playback volume from high to low Multimedia to be recommended.It should be understood that can also be determined according to multimedia effective broadcasting time to be recommended or clicking rate Candidate multimedia, principle determine that candidate multimedia principle is similar with according to playback volume, and details are not described herein.It needs to illustrate It is in other embodiments, multimedia to be recommended, such as preset multimedia can also to be determined according to preset Multimedia Style Style is " rock and roll ", then available Multimedia Style is the multimedia of " rock and roll " as multimedia to be recommended；In other realities It applies in example, the multimedia for belonging to same author (or singer) can also be obtained as multimedia to be recommended；In other embodiments, Can also according to actual needs, using about the multimedia of red-letter day current events as multimedia to be recommended；It, can such as during the Mid-autumn Festival To obtain the multimedia about themes such as " mid-autumn " " admiring the full moon " " reunions " as multimedia to be recommended.

Secondly, corresponding multi-media segment can be intercepted out from each multimedia to be recommended, and from each multimedia piece The cover of an at least each multi-media segment of frame image construction is intercepted out in section.Specifically, server can be first using preset Labeling method makes marks to the temporal information of the multi-media segment in each multimedia to be recommended, and temporal information may include: more At the beginning of media fragment and the end time；Preset labeling method may include: handmarking's method and/or based on depth The labeling method of neural network.The temporal information for being then based on multi-media segment intercepts out more matchmakers from each multimedia to be recommended Body segment.From the foregoing it will be appreciated that the cover of multi-media segment can include: static cover or dynamic cover；So from each more When intercepting out at least cover of an each multi-media segment of frame image construction in media fragment, any multimedia piece can be directed to Section, intercepts a frame image, and using a frame image of interception as any multi-media segment from any multi-media segment Static cover；Alternatively, intercepting multiple image from any multi-media segment, and it is any more to use the multiple image of interception to constitute this The dynamic cover of media fragment.

It is then possible to construct multimedia album using each multi-media segment, and using the cover of each multi-media segment Construct candidate cover set.When using each multi-media segment building multimedia album, server can use multimedia spelling Bonding tool splices each multi-media segment, obtains spliced multimedia, and multimedia splicing tool may include: The library FFmpeg (Fast Forward Mpeg, it is a set of be used to record, converted digital audio, video, and stream can be translated into Open source computer program)；Using spliced multimedia as multimedia album.In one embodiment, it can also obtain general Head multimedia and run-out multimedia, before head multimedia is added to spliced multimedia, run-out multimedia is added It adds to after spliced multimedia, obtains multimedia album, i.e., include: by the obtained multimedia album of this embodiment Head multimedia, spliced multimedia and run-out multimedia.

In one embodiment, each multi-media segment sequential in multimedia album；Getting multimedia album Later, the multi-media segment pair in multimedia album can also be obtained, which includes the first multi-media segment With the second multi-media segment, first multi-media segment and the second multi-media segment are adjacent；Transition animation is obtained, and transition is moved Picture is inserted between the first multi-media segment and second multi-media segment.Transition animation herein can be preset animation Any animation in library is also possible to animation relevant to the first multi-media segment and the second multi-media segment.It is as shown in Figure 4: Multi-media segment centering includes multi-media segment b (the first multi-media segment) and multi-media segment c (the second multi-media segment)； Transition animation is obtained, then transition animation is inserted between multi-media segment b and multi-media segment c.The embodiment of the present invention passes through Transition animation is inserted into two adjacent multi-media segments (the first multi-media segment and the second multi-media segment), can make from The transition that first multi-media segment goes to the second multi-media segment is more natural, smooth.

S302 obtains user's portrait.

S303 draws a portrait according to user and determines destination multimedia segment, and it is big that the first of the destination multimedia segment plays probability In probability threshold value.

Specifically, multimedia prediction first can be carried out according to the history multimedia messages of user's portrait, target is determined Multi-media segment, history multimedia messages may include following any: history multimedia information, history multimedia search Information and history multimedia collection information.Then, it intercepts to obtain destination multimedia segment from destination multimedia, the more matchmakers of the target Body segment includes: that the segment of the theme of destination multimedia is represented in destination multimedia, such as the climax in destination multimedia Segment, wonderful or significant segment；Alternatively, any segment in destination multimedia.It should be noted that due to target Multi-media segment is to intercept to obtain from destination multimedia, therefore the first of destination multimedia segment plays probability and subsequent institute It is identical that the first of the destination multimedia referred to plays probability.

In one embodiment, history multimedia messages include history multimedia information, correspondingly, according to user The history multimedia messages of portrait carry out multimedia prediction, determine that the specific embodiment of destination multimedia may is that head First, determine that the history multimedia of history broadcasting and the multimedia single of each history are gone through according to history multimedia information History playing duration and history broadcasting time.Then, it is broadcast according to the multimedia single history playing duration of each history and history It puts number and calculates that each history is multimedia to be liked being worth；It in one embodiment, can be according to the multimedia single of each history History playing duration and history broadcasting time calculate separately the multimedia history of each history and play total duration, and each history is more The history of media plays total duration as each history is multimedia and likes being worth.Finally, liking according to each history is multimedia Value determines that each history multimedia first plays probability, and plays probability for first and make greater than the history multimedia of preset threshold For destination multimedia.Specifically, available the first mapping table liked value and play probability, from the first mapping table It is middle to search that each history is multimedia to like being worth corresponding broadcasting probability respectively, and using the broadcasting probability found as Each history multimedia first plays probability.Wherein, like that the first mapping table of value and broadcasting probability can be such as 1 institute of table Show, likes that value is directly proportional with probability is played.

Table 1

Like being worth	Play probability
		More than or equal to 2000	90%
[1500,2000)	80%
		[1000,1500)	70%
……	……

In another embodiment, history multimedia messages include history multimedia search information, correspondingly, according to user The history multimedia messages of portrait carry out multimedia prediction, determine that the specific embodiment of destination multimedia may is that root The searching multimedia of historical search and the historical search letter of each searching multimedia are determined according to history multimedia search information Breath, which includes historical search number and/or historical search time.Then, according to each searching multimedia Historical search information determines that each searching multimedia is liked being worth.It in one embodiment, can be by each searching multimedia Historical search number is liked being worth respectively as each searching multimedia；In further embodiment, each search can be calculated separately Search time difference between rope multimedia historical search time and current time；It obtains search time difference and likes value Second mapping table is searched from the second mapping table respectively corresponding to the search time difference of each searching multimedia Like being worth, and likes what is found being worth as each searching multimedia and like being worth.Finally, according to each more matchmakers of search Body likes value to determine that the first of each searching multimedia plays probability, and plays the search that probability is greater than preset threshold for first Multimedia is as destination multimedia.Wherein, search time difference and like that the second mapping table of value can be as shown in table 2, Search time difference and value is liked to be inversely proportional.

Table 2

Search time difference	Like being worth
		Less than 24 hours	100
[24 hours, 48 hours)	90
		[48 hours, 72 hours)	80
……	……

In another embodiment, history multimedia messages include history multimedia collection information, correspondingly, according to user The history multimedia messages of portrait carry out multimedia prediction, determine that the specific embodiment of destination multimedia may is that root When determining that the collection multimedia of history collection and the multimedia history of each collection are collected according to history multimedia collection information Between.Then, it determines that each collection is multimedia according to each collection multimedia history collection time to like being worth.Specifically, can To calculate separately the collection time difference between each collection multimedia history collection time and current time；When obtaining collection Between difference and like the third mapping table of value, searched respectively from third mapping table and each collect multimedia collection Like being worth corresponding to time difference, and likes value to like being worth as each collection is multimedia for what is found.Finally, Value is liked to determine that each collection multimedia first plays probability according to each collection is multimedia, and the first broadcasting probability is big In preset threshold collection multimedia as destination multimedia.

S304, detection destination multimedia segment is with the presence or absence of in multimedia album.

S305, if destination multimedia segment there are in multimedia album, the more matchmakers of selection target in candidate cover set Target cover of the corresponding cover of body segment as multimedia album.

In step S304-S304, since the first of destination multimedia segment plays probability greater than preset threshold, illustrate to use Family is interested in the destination multimedia segment, therefore after destination multimedia segment has been determined, it is more to can detecte target Media fragment is with the presence or absence of in multimedia album.If it exists, then the envelope of destination multimedia segment is searched from candidate cover set The cover of the destination multimedia segment found is attracted user to click multimedia by face as the target cover of multimedia album Album.As shown in Figure 5 a, include 3 multi-media segments in multimedia album, be respectively: multi-media segment a, multi-media segment b And multi-media segment c；Correspondingly, the cover of multi-media segment a be cover 1, multi-media segment b cover be cover 2 and The cover of multi-media segment c is cover 3.Destination multimedia segment is multi-media segment b, there are in multimedia album, then will be sealed Face 2 is used as target cover.

It should be noted that with the presence of in multiple destination multimedia segment multimedia albums, then it can be from multiple target A destination multimedia segment is arbitrarily selected in multi-media segment, and the corresponding cover of selected destination multimedia segment is made For the target cover of multimedia album；Alternatively, selection first plays the mesh of maximum probability from multiple destination multimedia segment Multi-media segment is marked, and plays the corresponding cover of destination multimedia segment of maximum probability as multimedia for selected first The target cover of album；Or the cover of multiple destination multimedia segment is combined splicing and obtains target cover.

It in one embodiment, will be in candidate cover set if destination multimedia segment is not present in multimedia album Each cover be combined splicing, will be combined and spliced after cover as the target cover of multimedia album.Specifically, by candidate The specific embodiment that each cover in cover set is combined splicing may is that using each cover in candidate cover set Spliced, obtains splicing base map；Obtain the information that illustrates of background color mask and multimedia album, the background color mask it is transparent Degree is greater than default transparency threshold；By the background color mask and it is described illustrate that information is covered on the splicing base map, obtain group Close spliced cover.As shown in Figure 5 b, in candidate cover set include 5 covers, be respectively: cover 1, cover 2, cover 3, Cover 4 and cover 5 are spliced using this 5 covers, available splicing base map p.Obtain background color mask m and more matchmakers Body album illustrates information, by background color mask m and illustrate information be covered on splicing base map p on, obtain it is combined and spliced after cover S, cover s are target cover.

In further embodiment, if destination multimedia segment can first calculate the more matchmakers of target there is no in multimedia album The similarity of each multi-media segment in body segment and multimedia album.Specifically, firstly, eigenmatrix algorithm (example can be used Such as Fourier Transform Algorithm, mel cepstrum coefficients algorithm) calculate the eigenvectors matrix of destination multimedia segment and more The eigenvectors matrix of each multi-media segment in media album；Then using similarity algorithm (such as cosine similarity algorithm, Euclidean distance algorithm etc.) calculate separately destination multimedia segment eigenvectors matrix and multimedia album in each multimedia piece Similarity between section, obtains the similarity of each multi-media segment in destination multimedia segment and multimedia album.? After the similarity of each multi-media segment into destination multimedia segment and multimedia album, it can be sealed according to similarity from candidate The target cover of multimedia album is determined in the set of face.Wherein, determine that multimedia is closed from candidate cover set according to similarity The specific embodiment for the target cover collected may is that is determined as similar multimedia piece according to similarity from multimedia album Section, the similar multi-media segment are the maximum multi-media segment of similarity in multimedia album；It is searched from candidate cover set The cover of similar multi-media segment, and sealed the cover of the similar multi-media segment found as the target of multimedia album Face.

In further embodiment, if destination multimedia segment can draw a portrait pre- there is no in multimedia album according to user Survey multimedia album in each multi-media segment second play probability, and according to the second of each multi-media segment play probability from The target cover of multimedia album is determined in candidate cover set.Wherein, according to the second of each multi-media segment play probability from Determine that the specific embodiment of the target cover of multimedia album may is that according to each multi-media segment in candidate cover set Second plays the multi-media segment that determine the probability second plays maximum probability；Second is searched from candidate cover set plays probability The cover of maximum multi-media segment, and the cover for the multi-media segment for playing maximum probability for find second is as more matchmakers The target cover of body album.

S306 exports multimedia album and target cover, so that when recommending multimedia album, it is aobvious in user interface Show target cover.

Server, can be defeated by the multimedia album and target cover after the target cover of multimedia album has been determined Out to terminal.Terminal can show the target cover in user interface after receiving the multimedia album and target cover, To attract user's click play multimedia album.If terminal detects the clicking operation to the target cover of multimedia album, Then play the multimedia album.In one embodiment, server can also issue the displaying information of each multi-media segment, with For terminal when playing each multi-media segment in multimedia album, shows in the information display board of multi-media segment and show phase The displaying information answered, the displaying information can include but is not limited to: multimedia title, corresponding work belonging to multi-media segment Multimedia ranking belonging to regional information and the multi-media segment where person (such as singer), author, etc..

Based on the description above, the embodiment of the present invention proposes the signal stream of another multimedia recommendation method in Fig. 6 Cheng Tu, the multimedia recommendation method can be realized by above-mentioned mentioned server or terminal.For ease of description, originally Inventive embodiments are by taking server executes the multimedia recommendation method as an example.The multimedia recommendation method may include steps of S601-S604:

S601 obtains the candidate cover set of multimedia album and multimedia album.

Multimedia album may include the multi-media segment of at least two different multimedias, which may include institute State the corresponding cover of each multi-media segment in multimedia album.Wherein, the multi-media segment in multimedia album are as follows: pre- from meeting If the segment intercepted in the multimedia of condition, which includes any one of following: playback volume is greater than default playback volume Threshold value, effective broadcasting time are greater than preset times threshold value, clicking rate is greater than default clicking rate threshold value.The corresponding envelope of multi-media segment Face includes: static cover or dynamic cover；Wherein, which is the frame image institute structure that intercepts from multi-media segment At cover, the cover that dynamic cover is constituted for the multiple image that intercepts from the multi-media segment.

Optionally, each multi-media segment sequential in multimedia album；After getting multimedia album, may be used also To obtain the multi-media segment pair in multimedia album, which includes the first multi-media segment and more than second matchmaker Body segment, first multi-media segment and the second multi-media segment are adjacent；Transition animation is obtained, and transition animation is inserted into the Between one multi-media segment and second multi-media segment.

It should be noted that the specific embodiment of step S601 may refer to the step S201 in foregoing invention embodiment Or step S301, details are not described herein.

S602 obtains user's portrait, and is drawn a portrait according to user and determine object element, and the object element is for reflecting the use The multimedia hobby of the indicated user of family portrait.

User's portrait may include history multimedia operations record, which may include but be not limited to media manipulation record: History multimedia record, history multimedia search record, history multimedia collection record, etc..It is drawn getting user As after, can be drawn a portrait according to user in history multimedia operations record take out object element；Object element herein can To include but is not limited to: destination multimedia style, destination multimedia author (such as target video author, target song singer Deng).

In one embodiment, history multimedia operations record includes that history multimedia records；Correspondingly, according to The specific embodiment that history multimedia operations record in user's portrait takes out object element may is that according to the more matchmakers of history Body plays the history multimedia for recording and determining that history plays, and each multimedia history of history plays total duration and each history Multimedia element；History is played into the multimedia element of the maximum history of total duration as object element.

In another embodiment, history multimedia operations record includes that history multimedia search records；Correspondingly, according to The specific embodiment that history multimedia operations record in user's portrait takes out object element may is that according to the more matchmakers of history Body search records the searching multimedia for determining historical search, and the historical search number of each searching multimedia and each search are more The element of media；Using the element of the maximum searching multimedia of historical search number as object element.

In another embodiment, history multimedia operations record includes that history multimedia collection records；Correspondingly, according to The specific embodiment that history multimedia operations record in user's portrait takes out object element may is that according to the more matchmakers of history Body collection record determines the collection multimedia of history collection, and each collection multimedia history collection time and each collection are more The element of media；The collection time difference between each collection multimedia history collection time and current time is calculated, will be received The smallest multimedia element of collection of time difference is hidden as object element.

S603 chooses the destination multimedia segment to match with object element from multimedia album, and in candidate cover Target cover of the corresponding cover of selection target multi-media segment as multimedia album in set.

When choosing the destination multimedia segment to match with object element from multimedia album, multimedia conjunction can be obtained The corresponding element of each multi-media segment in volume；If including object element in the corresponding element of each multi-media segment, by target element The corresponding multi-media segment of element is as destination multimedia segment.If not including target element in the corresponding element of each multi-media segment Element then can first calculate the similarity between the corresponding element of each multi-media segment and object element；Specifically, cosine can be used The similarity algorithms such as similarity algorithm, Euclidean distance algorithm, Hamming distance algorithm calculate the corresponding element of each multi-media segment with Similarity between object element.Then, destination multimedia segment can be determined according to similarity, which is phase It is greater than the corresponding multi-media segment of element of similarity threshold like degree；Similarity threshold herein can according to practical business demand or The setting of person's empirical value.The similarity of multiple multi-media segments is greater than similarity threshold if it exists, then can be greater than similarity threshold from this Select the maximum multi-media segment of similarity as destination multimedia segment in multiple multi-media segments of value；Alternatively, big from this Any multi-media segment is chosen in multiple multi-media segments of similarity threshold as destination multimedia segment

Optionally, if choosing destination multimedia piece segment fault, each cover in candidate cover set can be combined Splicing, will be combined and spliced after cover as the target cover of multimedia album；Specific embodiment may refer to shown in Fig. 5 b, Details are not described herein.

S604 exports multimedia album and the target cover, so that when recommending multimedia album, in user circle Face displaying target cover.

For the embodiment of the present invention when recommending multimedia album, available user's portrait simultaneously determines target according to user's portrait Element；Then destination multimedia segment, and the selection target multimedia in the candidate cover set of institute can be determined according to object element Target cover of the corresponding cover of segment as multimedia album；Since object element is for reflecting the indicated user of user's portrait Multimedia hobby, therefore the destination multimedia segment according to determined by object element can meet to a certain extent user's Hobby is broadcast so that user can be attracted to click when user interface shows the corresponding target cover of the destination multimedia segment The multimedia album is put, and then the clicking rate and playback volume of multimedia album can be improved.

Description based on above method embodiment, in one embodiment, the embodiment of the invention also provides a kind of such as Fig. 7 Shown in Multimedia Recommendation device structural schematic diagram.As shown in fig. 7, the Multimedia Recommendation device in the embodiment of the present invention can wrap It includes:

Acquiring unit 101 is described more for obtaining the candidate cover set of multimedia album and the multimedia album Media album includes at least two multi-media segments, and candidate's cover set includes each multimedia piece in the multimedia album The corresponding cover of section；

The acquiring unit 101 is drawn a portrait for obtaining user's portrait, and according to the user from the candidate cover set The target cover of the middle determination multimedia album；

Output unit 102, for exporting the multimedia album and the target cover, so that described more in recommendation When media album, the target cover is shown in user interface.

In one embodiment, acquiring unit 101 according to the user for drawing a portrait from the candidate cover set When determining the target cover of the multimedia album, it can be specifically used for:

It is drawn a portrait according to the user and determines destination multimedia segment, it is big that the first of the destination multimedia segment plays probability In probability threshold value；

If the destination multimedia segment selects institute in the candidate cover set there are in the multimedia album State target cover of the corresponding cover of destination multimedia segment as the multimedia album.

In further embodiment, acquiring unit 101 for according to the user draw a portrait determine destination multimedia segment when, It can be specifically used for:

Multimedia prediction is carried out according to the history multimedia messages of user portrait, determines destination multimedia, institute It includes following any for stating history multimedia messages: history multimedia information, history multimedia search information and history Multimedia collection information；

It intercepts to obtain destination multimedia segment from the destination multimedia.

In further embodiment, acquiring unit 101 can also be used in:

If the destination multimedia segment there is no in the multimedia album, will be each in the candidate cover set Cover is combined splicing, will be combined and spliced after cover as the target cover of the multimedia album；Alternatively,

If the destination multimedia segment there is no in the multimedia album, calculate the destination multimedia segment and The similarity of each multi-media segment in the multimedia album is determined from the candidate cover set according to the similarity The target cover of the multimedia album；Alternatively,

If the destination multimedia segment is drawn a portrait described in prediction there is no in the multimedia album according to the user Second of each multi-media segment in multimedia album plays probability, and plays probability according to the second of each multi-media segment The target cover of the multimedia album is determined from the candidate cover set.

In further embodiment, acquiring unit 101 for it is described according to the similarity from the candidate cover set When the target cover of the middle determination multimedia album, it can be specifically used for:

It is determined as similar multi-media segment, the similar multimedia piece from the multimedia album according to the similarity Section is the maximum multi-media segment of similarity in the multimedia album；

The cover of the similar multi-media segment is searched from the candidate cover set, and described similar by what is found Target cover of the cover of multi-media segment as the multimedia album.

Multi-media segment in further embodiment, in the multimedia album are as follows: from the multimedia for meeting preset condition Middle to intercept obtained segment, the preset condition includes any one of following: playback volume is greater than default playback volume threshold value, effectively plays Number is greater than preset times threshold value, clicking rate is greater than default clicking rate threshold value；

The cover of the multi-media segment includes: static cover or dynamic cover；Wherein, the static cover is from institute The cover that the frame image intercepted in multi-media segment is constituted is stated, the dynamic cover is to intercept from the multi-media segment The cover that is constituted of multiple image.

Each multi-media segment sequential in further embodiment, in the multimedia album；Correspondingly, acquiring unit 101 can also be used in:

The multi-media segment pair in the multimedia album is obtained, the multi-media segment centering includes the first multimedia piece Section and the second multi-media segment, first multi-media segment and second multi-media segment are adjacent；

Transition animation is obtained, and the transition animation is inserted into first multi-media segment and second multimedia Between segment.

According to one embodiment of present invention, each step involved in Fig. 2 and method shown in Fig. 3 may each be by scheming Each unit in Multimedia Recommendation device shown in 7 is performed.For example, step S201 and S202 shown in Fig. 2 can be with It is executed by acquiring unit 101 shown in fig. 7, step S203 can be executed by output unit 102 shown in fig. 7；Again Such as, step S301-S305 shown in Fig. 3 can be executed by acquiring unit 101 shown in fig. 7, and step S306 can be by Output unit 102 shown in fig. 7 executes.

According to another embodiment of the invention, each unit in Multimedia Recommendation device shown in Fig. 7 can be distinguished Or one or several other units all are merged into constitute or some (a little) unit therein can also be split as again Functionally smaller multiple units are constituted, this may be implemented similarly to operate, the technology without influencing the embodiment of the present invention The realization of effect.Said units are logic-based function divisions, and in practical applications, the function of a unit can also be by more A unit is realized or the function of multiple units is realized by a unit.In other embodiments of the invention, multimedia pushes away Recommending device also may include other units, and in practical applications, these functions can also be assisted to realize by other units, and can It is realized with being cooperated by multiple units.

It according to another embodiment of the invention, can be by including central processing unit (CPU), random access memory It is transported on the universal computing device of such as computer of the processing elements such as medium (RAM), read-only storage medium (ROM) and memory element Row is able to carry out the computer program (including program code) of each step involved in correlation method as shown in Figures 2 and 3, Construct Multimedia Recommendation appliance arrangement as shown in Figure 7, and come the multimedia recommendation method of realizing the embodiment of the present invention. The computer program can be recorded in such as computer readable recording medium, and be loaded by computer readable recording medium In above-mentioned calculating equipment, and run wherein.

Description based on above method embodiment, in one embodiment, the embodiment of the invention also provides a kind of such as Fig. 8 Shown in Multimedia Recommendation device structural schematic diagram.As shown in figure 8, the Multimedia Recommendation device in the embodiment of the present invention can wrap It includes:

Acquiring unit 201 is described more for obtaining the candidate cover set of multimedia album and the multimedia album Media album includes the multi-media segment of at least two different multimedias, and candidate's cover set includes the multimedia album In the corresponding cover of each multi-media segment；

Acquiring unit 201 for obtaining user's portrait, and draws a portrait according to the user and determines object element, the target Element is used to reflect the multimedia hobby of the indicated user of user's portrait；

Selection unit 202, for choosing the more matchmakers of target to match with the object element from the multimedia album Body segment, and select the corresponding cover of the destination multimedia segment as the multimedia in the candidate cover set and close The target cover collected；

Output unit 203, for exporting the multimedia album and the target cover, so that described more in recommendation When media album, the target cover is shown in user interface.

In one embodiment, selection unit 202 from the multimedia album for choosing and the target element When the destination multimedia segment that element matches, it can be specifically used for:

Obtain the corresponding element of each multi-media segment in the multimedia album；

It is if including the object element in the corresponding element of each multi-media segment, the object element is corresponding Multi-media segment is as destination multimedia segment；

If not including the object element in the corresponding element of each multi-media segment, each multi-media segment pair is calculated The similarity between element and the object element answered, and determine that destination multimedia segment, the target are more according to similarity Media fragment is the corresponding multi-media segment of element that similarity is greater than similarity threshold.

In another embodiment, selection unit 202 can also be used in:

If choosing the destination multimedia piece segment fault, each cover in the candidate cover set is combined spelling Connect, will be combined and spliced after cover as the target cover of the multimedia album.

Multi-media segment in another embodiment, in the multimedia album are as follows: from the more matchmakers for meeting preset condition The segment intercepted in body, the preset condition include any one of following: playback volume is greater than default playback volume threshold value, effectively broadcasts It puts number and is greater than default clicking rate threshold value greater than preset times threshold value, clicking rate；

Each multi-media segment sequential in another embodiment, in the multimedia album；Correspondingly, obtaining single Member 201 can also be used in:

According to one embodiment of present invention, each step involved in method shown in fig. 6 may each be as shown in Figure 8 Multimedia Recommendation device in each unit be performed.For example, step S601 and S602 can be by Fig. 8 shown in Fig. 6 Shown in acquiring unit 201 execute, step S603 can be executed by selection unit 202 shown in fig. 8, step S604 It can be executed by output unit 203 shown in fig. 8.

According to another embodiment of the invention, each unit in Multimedia Recommendation device shown in Fig. 8 can be distinguished Or one or several other units all are merged into constitute or some (a little) unit therein can also be split as again Functionally smaller multiple units are constituted, this may be implemented similarly to operate, the technology without influencing the embodiment of the present invention The realization of effect.Said units are logic-based function divisions, and in practical applications, the function of a unit can also be by more A unit is realized or the function of multiple units is realized by a unit.In other embodiments of the invention, multimedia pushes away Recommending device also may include other units, and in practical applications, these functions can also be assisted to realize by other units, and can It is realized with being cooperated by multiple units.

It according to another embodiment of the invention, can be by including central processing unit (CPU), random access memory It is transported on the universal computing device of such as computer of the processing elements such as medium (RAM), read-only storage medium (ROM) and memory element Row is able to carry out the computer program (including program code) of each step involved in correlation method as shown in Figure 6, carrys out structure Make Multimedia Recommendation appliance arrangement as shown in Figure 8, and come the multimedia recommendation method of realizing the embodiment of the present invention.It is described Computer program can be recorded in such as computer readable recording medium, and be loaded by computer readable recording medium It states and calculates in equipment, and run wherein.

Description based on above method embodiment and Installation practice, the embodiment of the present invention also provide a kind of server. Fig. 9 is referred to, which includes at least processor 301, transceiver 302 and memory 303.Wherein, the place in server Managing device 301, transceiver 302 and memory 303 can be connected by bus or other modes.Memory 303 is calculated for storing Machine program, the computer program include program instruction, and the processor 301 is configured for calling described program instruction execution Above-mentioned mentioned multimedia recommendation method.Processor 301 (or CPU (Central Processing Unit, central processing Device)) be server calculating core and control core, be adapted for carrying out one or one or more instruction, be particularly adapted to load And execution one or one or more instruction are to realize correlation method process or corresponding function.

In one embodiment, processor 301 described in the embodiment of the present invention can be used for carrying out a series of multimedia Recommendation process, comprising: obtain the candidate cover set of multimedia album and the multimedia album, the multimedia album packet The multi-media segment of at least two different multimedias is included, candidate's cover set includes each multimedia in the multimedia album The corresponding cover of segment；User's portrait is obtained, and determination is described more from the candidate cover set according to user portrait The target cover of media album, user's portrait is for reflecting the multimedia hobby of the indicated user of user's portrait；It is defeated The multimedia album and the target cover out, so that being shown when recommending the multimedia album in user interface Described target cover, etc..Alternatively, processor 301 described in the embodiment of the present invention can be used for carrying out a series of multimedia Recommendation process, comprising: obtain the candidate cover set of multimedia album and the multimedia album, the multimedia album packet The multi-media segment of at least two different multimedias is included, candidate's cover set includes each multimedia in the multimedia album The corresponding cover of segment；User's portrait is obtained, and is drawn a portrait according to the user and determines object element, the object element is for anti- Reflect the multimedia hobby of the indicated user of user's portrait；It is chosen and the object element phase from the multimedia album The destination multimedia segment matched, and select in the candidate cover set the corresponding cover of the destination multimedia segment as The target cover of the multimedia album；The multimedia album and the target cover are exported, so that described in the recommendation When multimedia album, described target cover, etc. is shown in user interface.

The embodiment of the invention also provides a kind of computer storage medium (Memory), the computer storage medium is clothes The memory device being engaged in device, for storing program and data.It is understood that computer storage medium herein both can wrap Include the built-in storage medium in server, naturally it is also possible to the expansion storage medium supported including server.Computer storage Medium provides memory space, which stores the operating system of server.Also, it is also housed in the memory space Suitable for by one or more than one instructions that processor 301 loads and executes, these instructions can be one or more Computer program (including program code).It should be noted that computer storage medium herein can be high-speed RAM storage Device is also possible to non-labile memory (non-volatile memory), for example, at least a magnetic disk storage；It is optional Can also be that at least one is located remotely from the computer storage medium of aforementioned processor.

In one embodiment, it can be loaded by processor 301 and execute first stored in computer storage medium and calculated Machine program instruction, to realize the corresponding steps in relation to the method in Multimedia Recommendation embodiment in above-mentioned Fig. 2 or Fig. 3；It is specific real In existing, the first computer program instructions in computer storage medium are loaded by processor 301 and execute following steps:

In one embodiment, the multimedia is being determined from the candidate cover set according to user portrait When the target cover of album, first computer program instructions are loaded by processor 301 and are specifically executed:

In another embodiment, when determining destination multimedia segment according to user portrait, described first is calculated Machine program instruction is loaded by processor 301 and is specifically executed:

In another embodiment, first computer program instructions can also be loaded and be executed by processor 301:

In another embodiment, determining that the multimedia is closed from the candidate cover set according to the similarity When the recommendation cover collected, first computer program instructions are loaded by processor 301 and are specifically executed:

The cover of the similar multi-media segment is searched from the candidate cover set, and described similar by what is found Recommendation cover of the cover of multi-media segment as the multimedia album.

Each multi-media segment sequential in another embodiment, in the multimedia album；Correspondingly, described One computer program instructions can also be loaded and be executed by processor 301:

In further embodiment, it can also be loaded by processor 301 and execute the second meter stored in computer storage medium Calculation machine program instruction, to realize the corresponding steps in relation to the method in Multimedia Recommendation embodiment in above-mentioned Fig. 6；Specific implementation In, the second computer program instruction in computer storage medium is loaded by processor 301 and executes following steps:

In one embodiment, more in the target chosen from the multimedia album and the object element matches When media fragment, the second computer program instruction can also be loaded and be executed by processor 301:

In another embodiment, the second computer program instruction can also be loaded and be executed by processor 301:

Each multi-media segment sequential in another embodiment, in the multimedia album；Correspondingly, described Two computer program instructions can also be loaded and be executed by processor 301:

The above disclosure is only the preferred embodiments of the present invention, cannot limit the right model of the present invention with this certainly It encloses, therefore equivalent changes made in accordance with the claims of the present invention, is still within the scope of the present invention.

Claims

1. a kind of multimedia recommendation method characterized by comprising

The candidate cover set of multimedia album and the multimedia album is obtained, the multimedia album includes at least two The multi-media segment of different multimedia, it is described candidate cover set include in the multimedia album each multi-media segment it is corresponding Cover；

User's portrait is obtained, and determines the mesh of the multimedia album from the candidate cover set according to user portrait Cover is marked, user's portrait is for reflecting the multimedia hobby of the indicated user of user's portrait；

The multimedia album and the target cover are exported, so that when recommending the multimedia album, in user circle Face shows the target cover.

2. the method as described in claim 1, which is characterized in that described to be drawn a portrait according to the user from the candidate cover set The target cover of the middle determination multimedia album, comprising:

It is drawn a portrait according to the user and determines destination multimedia segment, the first of the destination multimedia segment plays probability and be greater than generally Rate threshold value；

If the destination multimedia segment selects the mesh in the candidate cover set there are in the multimedia album Mark target cover of the corresponding cover of multi-media segment as the multimedia album.

3. method according to claim 2, which is characterized in that described drawn a portrait according to the user determines destination multimedia piece Section, comprising:

Multimedia prediction is carried out according to the history multimedia messages of user portrait, determines destination multimedia, it is described to go through History multimedia messages include following any: history multimedia information, history multimedia search information and the more matchmakers of history Body Information on Collection；

4. method according to claim 2, which is characterized in that the method also includes:

If the destination multimedia segment is not present in the multimedia album, by each cover in the candidate cover set Be combined splicing, will be combined and spliced after cover as the target cover of the multimedia album；Alternatively,

If there is no in the multimedia album, calculate the destination multimedia segment and described for the destination multimedia segment The similarity of each multi-media segment in multimedia album, according to similarity determination from the candidate cover set The target cover of multimedia album；Alternatively,

If the destination multimedia segment draws a portrait according to the user there is no in the multimedia album and predicts more matchmakers Second of each multi-media segment in body album plays probability, and plays probability from institute according to the second of each multi-media segment State the target cover that the multimedia album is determined in candidate cover set.

5. method according to any of claims 1-4, which is characterized in that the method also includes:

Multi-media segment in the multimedia album are as follows: the segment intercepted from the multimedia for meet preset condition, institute It includes any one of following for stating preset condition: playback volume is greater than default playback volume threshold value, effective broadcasting time is greater than preset times threshold Value, clicking rate are greater than default clicking rate threshold value；

The cover of the multi-media segment includes: static cover or dynamic cover；Wherein, the static cover is from described more The cover that the frame image intercepted in media fragment is constituted, the dynamic cover are more to intercept from the multi-media segment The cover that frame image is constituted.

6. method as claimed in claim 5, which is characterized in that each multi-media segment in the multimedia album is sequentially arranged Column；The method also includes:

Obtain the multi-media segment pair in the multimedia album, the multi-media segment centering include the first multi-media segment and Second multi-media segment, first multi-media segment and second multi-media segment are adjacent；

Transition animation is obtained, and the transition animation is inserted into first multi-media segment and second multi-media segment Between.

7. a kind of multimedia recommendation method characterized by comprising

User's portrait is obtained, and is drawn a portrait according to the user and determines object element, the object element is for reflecting the user The multimedia hobby of user indicated by drawing a portrait；

The destination multimedia segment to match with the object element is chosen from the multimedia album, and in the candidate envelope Target cover of the corresponding cover of the destination multimedia segment as the multimedia album is selected in the set of face；

8. the method for claim 7, which is characterized in that described to be chosen and the target element from the multimedia album The destination multimedia segment that element matches, comprising:

If including the object element in the corresponding element of each multi-media segment, by the corresponding more matchmakers of the object element Body segment is as destination multimedia segment；

If not including the object element in the corresponding element of each multi-media segment, it is corresponding to calculate each multi-media segment Similarity between element and the object element, and destination multimedia segment, the destination multimedia are determined according to similarity Segment is the corresponding multi-media segment of element that similarity is greater than similarity threshold.

9. the method for claim 7, which is characterized in that the method also includes:

If choosing the destination multimedia piece segment fault, each cover in the candidate cover set is combined splicing, Cover after will be combined and spliced is as the target cover of the multimedia album.

10. such as the described in any item methods of claim 7-9, which is characterized in that

The corresponding cover of the multi-media segment includes: static cover or dynamic cover；Wherein, the static cover is from institute The cover that the frame image intercepted in multi-media segment is constituted is stated, the dynamic cover is to intercept from the multi-media segment The cover that is constituted of multiple image.

11. method as claimed in claim 10, which is characterized in that each multi-media segment in the multimedia album is sequentially arranged Column；The method also includes:

12. a kind of Multimedia Recommendation device characterized by comprising

Acquiring unit, for obtaining the candidate cover set of multimedia album and the multimedia album, the multimedia is closed Collecting includes at least two multi-media segments, and candidate's cover set includes each multi-media segment correspondence in the multimedia album Cover；

The acquiring unit is determined from the candidate cover set for obtaining user's portrait, and according to user portrait The target cover of the multimedia album, user's portrait is for reflecting the multimedia happiness of the indicated user of user's portrait It is good；

Output unit, for exporting the multimedia album and the target cover, so that recommending the multimedia to close When collecting, the target cover is shown in user interface.

13. a kind of Multimedia Recommendation device characterized by comprising

Acquiring unit, for obtaining the candidate cover set of multimedia album and the multimedia album, the multimedia is closed The multi-media segment including at least two different multimedias is collected, candidate's cover set includes each more in the multimedia album The corresponding cover of media fragment；

Acquiring unit for obtaining user's portrait, and draws a portrait according to the user and determines that object element, the object element are used for Reflect the multimedia hobby of the indicated user of user's portrait；

Selection unit, for choosing the destination multimedia segment to match with the object element from the multimedia album, And mesh of the corresponding cover of the destination multimedia segment as the multimedia album is selected in the candidate cover set Mark cover；

14. a kind of server, which is characterized in that including processor, transceiver and memory, the processor, transceiver and deposit Reservoir is connected with each other, wherein the memory includes that the first program refers to for storing computer program, the computer program It enables, the processor is configured for calling first program instruction, executes more matchmakers as claimed in any one of claims 1 to 6 Body recommended method；Alternatively, the computer program includes the second program instruction, the processor is configured for calling described the Two program instructions execute such as the described in any item multimedia recommendation methods of claim 7-11.

15. a kind of computer storage medium, which is characterized in that the computer storage medium is stored with the first computer program and refers to It enables, first computer program instructions are suitable for being loaded by processor and executing more matchmakers as claimed in any one of claims 1 to 6 Body recommended method；Alternatively, the computer storage medium is stored with second computer program instruction, the second computer program Instruction is suitable for being loaded by processor and being executed such as the described in any item multimedia recommendation methods of claim 7-11.