CN106202304A - Method of Commodity Recommendation based on video and device - Google Patents
Method of Commodity Recommendation based on video and device Download PDFInfo
- Publication number
- CN106202304A CN106202304A CN201610511072.1A CN201610511072A CN106202304A CN 106202304 A CN106202304 A CN 106202304A CN 201610511072 A CN201610511072 A CN 201610511072A CN 106202304 A CN106202304 A CN 106202304A
- Authority
- CN
- China
- Prior art keywords
- user
- article
- video
- merchandise news
- interested
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/7867—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title and artist information, manually generated time, location and usage information, user ratings
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Multimedia (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to Method of Commodity Recommendation based on video and device.The method includes: obtain the frame of video of video file；It is identified this frame of video processing, to determine the article to be selected and the label of article to be selected that this frame of video includes；Obtain the merchandise news that user is interested；From article to be selected, filter out user's relative article according to the merchandise news that user is interested, and determine the user dependent merchandise information corresponding with user's relative article；Play this frame of video time, user's relative article is highlighted, and meet first pre-conditioned in the case of, show user's dependent merchandise information.Method of Commodity Recommendation based on video and device according to the present invention can recommend its merchandise news that may be interested according to the merchandise news that user is interested to user, substantially increase the efficiency of commercial product recommending based on video, and improve user and watch the experience of video.
Description
Technical field
The present invention relates to areas of information technology, particularly relate to a kind of Method of Commodity Recommendation based on video and device.
Background technology
" seeing while buy " refers to during user watches video, recommends the business relevant to video content to user
Product, allow users to immediately obtain the merchandise news relevant to video content.In order to realize " seeing while buy ", in prior art
In, need the commodity relevant to video content by operation personnel's human configuration, and the goods links of human configuration dependent merchandise.
In existing commercial product recommending technology based on video, for same video, all users are recommended identical
Commodity.The commodity interested due to different user are likely to difference, therefore according to this commercial product recommending skill based on video
Art, then be likely to recommend its uninterested commodity to user.If recommending its uninterested commodity, then user to user
Generally will not click on these commodity or buy this commodity, thus causing the inefficient of commercial product recommending, and easily affecting user's viewing
The experience of video.
Summary of the invention
Technical problem
In view of this, the technical problem to be solved in the present invention is, existing commercial product recommending technology based on video easily to
User recommends its uninterested commodity, causes the inefficient of commercial product recommending.
Solution
In order to solve above-mentioned technical problem, according to one embodiment of the invention, it is provided that a kind of commodity based on video push away
Recommend method, including:
Obtain the frame of video of video file；
It is identified described frame of video processing, to determine the article to be selected and described thing to be selected that described frame of video includes
The label of product；
Obtain the merchandise news that user is interested；
From described article to be selected, filter out user's relative article according to the merchandise news that described user is interested, and determine
The user dependent merchandise information corresponding with described user's relative article；
When playing described frame of video, described user's relative article is highlighted, and preset bar meeting first
In the case of part, show described user's dependent merchandise information.
For said method, in a kind of possible implementation, according to described user merchandise news interested from institute
State and article to be selected filter out user's relative article, and determine the user dependent merchandise letter corresponding with described user's relative article
Breath, including:
Label according to described article to be selected scans for, and obtains the first Search Results；
The merchandise news to be selected mated with described article to be selected is filtered out from described first Search Results；
Described user's phase that the merchandise news interested with described user is mated is filtered out from described merchandise news to be selected
Underlying commodity information,
Determine user's relative article corresponding with described user's dependent merchandise information in described article to be selected.
For said method, in a kind of possible implementation, according to described user merchandise news interested from institute
State and article to be selected filter out user's relative article, and determine the user dependent merchandise letter corresponding with described user's relative article
Breath, including:
From described article to be selected, user's relative article is filtered out according to the merchandise news that described user is interested；
Label according to described user's relative article scans for, and obtains the second Search Results；
The described user's dependent merchandise letter mated with described user's relative article is filtered out from described second Search Results
Breath.
For said method, in a kind of possible implementation, it is identified described frame of video processing, particularly as follows:
It is identified processing to described frame of video by the first artificial neural networks.
For said method, in a kind of possible implementation, by the first artificial neural networks to described video
Before frame is identified processing, described method also includes:
It is defined to the classification of the article of described the first artificial neural networks identification specify classification.
For said method, in a kind of possible implementation, obtain the merchandise news that user is interested, including:
Obtain cookie；
Obtain, from third party website, the merchandise news that described user is interested according to described cookie.
For said method, in a kind of possible implementation, according to described user merchandise news interested from institute
State and article to be selected filter out user's relative article, and determine the user dependent merchandise letter corresponding with described user's relative article
Breath, particularly as follows: the returning rate of described user meet second pre-conditioned in the case of, according to the commodity that described user is interested
Information filters out user's relative article from described article to be selected, and determines the user phase corresponding with described user's relative article
Underlying commodity information.
For said method, in a kind of possible implementation, the label of described article to be selected includes following at least one
:
The described article to be selected sectional drawing in described frame of video, the classification of described article to be selected, the money of described article to be selected
Formula, the color of described article to be selected and the decorative pattern of described article to be selected.
In order to solve above-mentioned technical problem, according to another embodiment of the present invention, it is provided that a kind of commodity based on video
Recommendation apparatus, including:
Frame of video acquisition module, for obtaining the frame of video of video file；
Identification module, for being identified process, to determine the article to be selected that described frame of video includes to described frame of video
And the label of described article to be selected；
The merchandise news acquisition module that user is interested, for obtaining the merchandise news that user is interested；
User's relative article and merchandise news determine module, for the merchandise news interested according to described user from described
Article to be selected filter out user's relative article, and determines the user dependent merchandise letter corresponding with described user's relative article
Breath；
Display module, for when playing described frame of video, highlighting described user's relative article, and full
In the case of foot first is pre-conditioned, show described user's dependent merchandise information.
For said apparatus, in a kind of possible implementation, described user's relative article and merchandise news determine mould
Block includes:
First search submodule, for scanning for according to the label of described article to be selected, obtains the first Search Results；
Merchandise news to be selected screening submodule, for filtering out and described article to be selected from described first Search Results
The merchandise news to be selected joined；
User's dependent merchandise information matches submodule, feels with described user for filtering out from described merchandise news to be selected
Described user's dependent merchandise information of the merchandise news coupling of interest；
User's relative article determines submodule, is used for determining in described article to be selected and described user's dependent merchandise information phase
Corresponding user's relative article.
For said apparatus, in a kind of possible implementation, described user's relative article and merchandise news determine mould
Block includes:
User's relative article screening submodule, for the merchandise news interested according to described user from described article to be selected
In filter out user's relative article；
Second search submodule, for scanning for according to the label of described user's relative article, obtains the second search knot
Really；
User's dependent merchandise information sifting submodule, for filtering out and described user's phase from described second Search Results
Close described user's dependent merchandise information of article coupling.
For said apparatus, in a kind of possible implementation, described identification module specifically for: artificial by first
Described frame of video is identified processing by neutral net.
For said apparatus, in a kind of possible implementation, described device also includes:
Identify that classification limits module, for being defined to specify by the classification of the article of described the first artificial neural networks identification
Classification.
For said apparatus, in a kind of possible implementation, the merchandise news acquisition module that described user is interested
Including:
Cookie obtains submodule, is used for obtaining cookie；
The merchandise news that user is interested obtains submodule, for obtaining described from third party website according to described cookie
The merchandise news that user is interested.
For said apparatus, in a kind of possible implementation, described user's relative article and merchandise news determine mould
Block specifically for: the returning rate of described user meet second pre-conditioned in the case of, according to the business that described user is interested
Product information filters out user's relative article from described article to be selected, and determines the user corresponding with described user's relative article
Dependent merchandise information.
For said apparatus, in a kind of possible implementation, the label of described article to be selected includes following at least one
:
The described article to be selected sectional drawing in described frame of video, the classification of described article to be selected, the money of described article to be selected
Formula, the color of described article to be selected and the decorative pattern of described article to be selected.
Beneficial effect
Process, to determine the article to be selected and the mark of article to be selected that frame of video includes by frame of video is identified
Sign, from article to be selected, filter out user's relative article according to the merchandise news that user is interested, and determine and user's correlative
User's dependent merchandise information that condition is corresponding, when playing frame of video, highlights user's relative article, and is meeting
In the case of first is pre-conditioned, showing user's dependent merchandise information, commodity based on video according to embodiments of the present invention push away
Recommend method and device and can recommend its merchandise news that may be interested to user, significantly according to the merchandise news that user is interested
Improve the efficiency of commercial product recommending based on video, and improve user and watch the experience of video.
According to below with reference to the accompanying drawings detailed description of illustrative embodiments, the further feature of the present invention and aspect being become
Clear.
Accompanying drawing explanation
The accompanying drawing of the part comprising in the description and constituting description together illustrates the present invention's with description
Exemplary embodiment, feature and aspect, and for explaining the principle of the present invention.
Fig. 1 illustrates the flowchart of based on video according to an embodiment of the invention Method of Commodity Recommendation；
Fig. 2 illustrates the first artificial neural networks in based on video according to an embodiment of the invention Method of Commodity Recommendation
Schematic diagram；
Fig. 3 illustrates that based on video according to an embodiment of the invention Method of Commodity Recommendation step S104's is one exemplary
Implement flow chart；
Fig. 4 illustrates the another exemplary of based on video according to an embodiment of the invention Method of Commodity Recommendation step S104
Implement flow chart；
Fig. 5 illustrates that the one of based on video according to an embodiment of the invention Method of Commodity Recommendation exemplary realizes flow process
Figure；
Fig. 6 illustrates that based on video according to an embodiment of the invention Method of Commodity Recommendation step S103's is one exemplary
Implement flow chart；
Fig. 7 illustrates the structured flowchart of based on video the according to another embodiment of the present invention device for recommending the commodity；
Fig. 8 illustrates an exemplary structural frames of based on video the according to another embodiment of the present invention device for recommending the commodity
Figure；
Fig. 9 illustrates the structured flowchart of commercial product recommending equipment based on video according to another embodiment of the invention.
Detailed description of the invention
Various exemplary embodiments, feature and the aspect of the present invention is described in detail below with reference to accompanying drawing.In accompanying drawing identical
Reference represent the same or analogous element of function.Although the various aspects of embodiment shown in the drawings, but remove
Non-specifically is pointed out, it is not necessary to accompanying drawing drawn to scale.
The most special word " exemplary " means " as example, embodiment or illustrative ".Here as " exemplary "
Illustrated any embodiment should not necessarily be construed as preferred or advantageous over other embodiments.
It addition, in order to better illustrate the present invention, detailed description of the invention below gives numerous details.
It will be appreciated by those skilled in the art that do not have some detail, the present invention equally implements.In some instances, for
Method well known to those skilled in the art, means, element and circuit are not described in detail, in order to highlight the purport of the present invention.
Embodiment 1
Fig. 1 illustrates the flowchart of based on video according to an embodiment of the invention Method of Commodity Recommendation.Such as Fig. 1 institute
Showing, the method specifically includes that
In step S101, obtain the frame of video of video file.
As an example of the embodiment of the present invention, obtain the frame of video of video file, particularly as follows: in the class of video file
In the case of type is order video, the key frame of extraction video file.Wherein, key frame can refer to role or thing in video
That frame residing for key operations in body motion or change.Order video can include that on line, order video and local ordering regard
Frequently.For order video on line, can be by the video playback address acquisition video file of order video on this line；For this locality
Order video, can directly obtain video file from this locality.
As another example of the embodiment of the present invention, obtain the frame of video of video file, particularly as follows: at video file
In the case of type is live video, obtain the present frame of video file.
Wherein, live can finger, at the scene along with generation, the development process of event synchronize to make and release news, has double
To the information network published method of the process of circulation.Order video can refer to non-live video.
In step s 102, be identified this frame of video processing, with determine article to be selected that this frame of video includes and
The label of article to be selected.
Article to be selected in this embodiment, can be user may interested or may have a mind to buy any thing
The dress ornament occurred in product, such as this frame of video or furniture etc..The label of article to be selected can be to be capable of identify that this article to be selected
Information, or represent the information of the attribute of these article to be selected.Such as these article to be selected sectional drawing in this frame of video, this is to be selected
The classification of article, the style of these article to be selected, the color of these article to be selected and the decorative pattern etc. of these article to be selected.
As an example of the embodiment of the present invention, it is identified this frame of video processing, particularly as follows: artificial by first
This frame of video is identified processing by neutral net.
It will be understood by those skilled in the art that and prior art has various means all can realize identifying from this frame of video and treat
Selecting article, and determine the label corresponding with these article to be selected, wherein, artificial neural network is exactly one of these means.
As it is known by the man skilled in the art, artificial neural network (Artificial Neural Networks, is abbreviated as
ANNs) it is a kind of imitation animal nerve network behavior feature, carries out the algorithm mathematics model of distributed parallel information processing.This
Network relies on the complexity of system, by adjusting interconnective relation between internal great deal of nodes, thus reaches to process letter
The purpose of breath, and there is self study and adaptive ability, it is possible to the functions such as implementation pattern identification, coupling.
In this example, using the frame of video of acquisition as the most independent image, this frame of video is inputted first artificial
Neutral net, is identified processing to this frame of video by the first artificial neural networks.Here, the first artificial neural networks is permissible
For identifying the article to be selected and the label of these article to be selected that this frame of video includes.By artificial for this frame of video input first god
After network, the first artificial neural networks exports article to be selected and the label of these article to be selected that this frame of video includes.
Fig. 2 illustrates the first artificial neural networks in based on video according to an embodiment of the invention Method of Commodity Recommendation
Schematic diagram.Such as, the input of the first artificial neural networks can be this frame of video, and this can be regarded by the first artificial neural networks
Frequently each pixel of frame is converted to three-dimensional color vector, and is identified the three-dimensional color vector being converted to processing, with defeated
Go out article to be selected and the label of article to be selected that this frame of video includes.
It should be noted that the first artificial neural networks in this example can utilize existing artificial neural network technology
Realize, no longer the operation principle of the first artificial neural networks is repeated at this.
In a kind of possible implementation, the method can also include: by the mark of the information of frame of video Yu article to be selected
Mapping relations between label store in the first memory.Wherein, the information of frame of video can include that this frame of video is at this video
Time point in file, it is also possible to include the unique identifier of this video file.In this implementation, can deposit by first
Reservoir stores the recognition result of the first artificial neural networks, i.e. can be by the information of first memory storage frame of video with to be selected
Mapping relations between the label of article.
In step s 103, the merchandise news that user is interested is obtained.
As an example of the embodiment of the present invention, can be before playing video or during playing video, please
Ask user to select its merchandise classification interested from multiple merchandise classifications, and the merchandise classification selected according to user determines that it is felt
The merchandise news of interest.
In step S104, from article to be selected, filter out user's relative article according to the merchandise news that user is interested,
And determine the user dependent merchandise information corresponding with user's relative article.
Wherein, user's dependent merchandise information can include the link of user's dependent merchandise, it is also possible to includes that user is correlated with business
The title of product and/or picture.User's dependent merchandise information can set flexibly according to operation demand, such as, and user's dependent merchandise
Information can also include platform information, Business Information, sales volume and/or comment amount etc., in this no limit.
In step S105, when playing this frame of video, user's relative article is highlighted, and meeting first
In the case of pre-conditioned, show user's dependent merchandise information.
As an example of the embodiment of the present invention, user's relative article is highlighted and includes: be correlated with user
The preset icon displayed above of article region；And/or show user's relative article with dynamic effect in this frame of video.Its
In, preset icon can be a dot, in this no limit.Displayed above in user's relative article region is preset
Icon can be: shows preset icon with float layer above user's relative article region.With dynamic in this frame of video
State effect display user's relative article can be: according to the border of the user's relative article identified, enters user's relative article
Line flicker shows.Preset icon displayed above in user's relative article region can be: at user's relative article place
The top in region shows preset icon with dynamic effect, and wherein, dynamic effect can include flicker effect and/or shake effect,
In this no limit.
As an example of the embodiment of the present invention, display user's dependent merchandise information can be: in the feelings of played in full screen
Under condition, from the left side of screen, the right, above or below eject merchandise news frame, with by this merchandise news frame show user
Dependent merchandise information.Such as, in the case of mobile phones transverse screen played in full screen, eject merchandise news frame from the left side of screen.At this
In example, can guide and the user discover that its commodity that may be interested in the case of user's full frame viewing video.
As another example of the embodiment of the present invention, display user's dependent merchandise information can also be: full frame broadcasts non-
In the case of putting, position display user's dependent merchandise information beyond broadcast window in screen.In this example, can with
The interference to user is reduced during the viewing video of family.
As an example of the embodiment of the present invention, meet the first pre-conditioned may include that and detect that cursor moves to
The top of preset icon；And/or detect that cursor moves to the top of user's relative article region.
As another example of the embodiment of the present invention, meet first and pre-conditioned can also include: detect that cursor is sliding
Cross preset icon；And/or, detect that cursor slips over user's relative article region.
As another example of the embodiment of the present invention, meet first and pre-conditioned can also include: user's point detected
Hit preset icon；And/or, detect that user clicks on user's relative article region.
As an example of the embodiment of the present invention, meet first pre-conditioned in the case of, no longer user is correlated with
Article highlight.
In a kind of possible implementation, after display user's dependent merchandise information, user can be by clicking on user
Dependent merchandise information inspection commodity details or purchase commodity.
Fig. 3 illustrates that based on video according to an embodiment of the invention Method of Commodity Recommendation step S104's is one exemplary
Implement flow chart.As it is shown on figure 3, filter out user's correlative from article to be selected according to the merchandise news that user is interested
Product, and determine the user dependent merchandise information corresponding with user's relative article, including:
In step S301, scan for according to the label of article to be selected, obtain the first Search Results.
Such as, scan for according to the label of article to be selected, can be according to the article to be selected sectional drawing in this frame of video
Scan for；Can also scan for according to the classification of the article to be selected sectional drawing in this frame of video and article to be selected；Can also
For scanning for according to the style of the article to be selected sectional drawing in this frame of video and article to be selected；Or can be according to thing to be selected
The style of product sectional drawing, the classification of article to be selected and article to be selected in this frame of video scans for.
In this example, scan for according to the label of article to be selected, Ke Yiwei: according to the label of article to be selected any one
Individual or merchandising database corresponding to multiple electricity business website scans for.
In step s 302, from the first Search Results, filter out the merchandise news to be selected mated with article to be selected.
It will be understood by those skilled in the art that having various means all can realize from the first Search Results in prior art screens
Going out the merchandise news to be selected mated with article to be selected, wherein, artificial neural network is exactly one of these means.
As an example of the embodiment of the present invention, filter out to be selected with what article to be selected mated from the first Search Results
Merchandise news can be: filtered out to be selected with what article to be selected mated from the first Search Results by the second artificial neural network
Merchandise news.In this example, the second artificial neural network may be used for mating article to be selected and the first Search Results
Process, to determine similarity between the two, and then from the first Search Results, filter out the business to be selected mated with article to be selected
Product information.Such as, several pictures of the commodity during the input of the second artificial neural network can be the first Search Results and treating
Selecting article sectional drawing in this frame of video, output result can be that the commodity in the first Search Results are similar to article to be selected
Degree.In this example, can by the first Search Results with the similarity of article to be selected more than the merchandise news of the first preset value
It is defined as the merchandise news to be selected mated with article to be selected, i.e. retains the business more than the first preset value of the similarity with article to be selected
Product information, rejects the merchandise news less than or equal to the first preset value of the similarity with article to be selected.Such as, the first preset value can
Think 80%, in this no limit.It should be noted that the first preset value can be flexible according to the merchandise news number to be selected of coupling
Adjust, such as, if carrying out, by the second artificial neural network, the commodity to be selected mated with the article to be selected letter that matching treatment obtains
Breath number is relatively big, then can heighten the first preset value；If by the second artificial neural network carry out that matching treatment obtains with to be selected
The merchandise news number to be selected of article coupling is less, then can turn down the first preset value.First preset value can also be according to coupling
Precision demand is adjusted flexibly, and such as, if desired carries out strong correlation coupling, then can heighten the first preset value；If desired
Carry out weak relevant matches, then can turn down the first preset value.
It should be noted that the second artificial neural network in the embodiment of the present invention can utilize existing ANN
Network technology realizes, and no longer repeats the operation principle of the second artificial neural network at this.
In a kind of possible implementation, the method can also include: by the information of frame of video, the label of article to be selected
And the mapping relations between merchandise news three to be selected are stored in second memory.In this implementation, can be by the
Two memorizeies store the matching result of the second artificial neural network, i.e. can by the information of second memory storage frame of video,
Mapping relations between label and the merchandise news three to be selected of article to be selected.
It should be noted that owing to the merchandise news such as goods links of electricity business website is it may happen that change, therefore, it can
Scheduled task is set, regular update merchandise news to be selected, and according to the merchandise news to be selected updated, time update second memory
Mapping relations between the information of frame of video, the label of article to be selected and the merchandise news three to be selected of storage.
In step S303, from merchandise news to be selected, filter out user's phase that the merchandise news interested with user is mated
Underlying commodity information.
It will be understood by those skilled in the art that having various means all can realize from merchandise news to be selected in prior art screens
Going out user's dependent merchandise information that the merchandise news interested with user is mated, wherein, artificial neural network is exactly these means
One of.
As an example of the embodiment of the present invention, from merchandise news to be selected, filter out the commodity letter interested with user
User's dependent merchandise information of breath coupling can be: is filtered out from merchandise news to be selected by the 3rd artificial neural network and uses
User's dependent merchandise information of the merchandise news coupling that family is interested.In this example, the 3rd artificial neural network may be used for
The merchandise news that merchandise news to be selected is interested with user is carried out matching treatment, to determine similarity between the two, and then
User's dependent merchandise information that the merchandise news interested with user is mated is filtered out from merchandise news to be selected.Such as, the 3rd
The input of artificial neural network can be the figure that picture corresponding to the merchandise news to be selected merchandise news interested with user is corresponding
Sheet, output result can be the similarity of the merchandise news to be selected merchandise news interested with user.In this example, can be by
The similarity of merchandise news interested with user in merchandise news to be selected determines more than the merchandise news to be selected of the second preset value
For user's dependent merchandise information.Such as, the second preset value can be 70%, in this no limit.It should be noted that second is pre-
If value can be adjusted flexibly, such as, if being carried out by the 3rd artificial neural network according to user's dependent merchandise Information Number of coupling
User's dependent merchandise Information Number that the merchandise news interested with user that matching treatment obtains is mated is relatively big, then can heighten the
Two preset values；If carrying out what the merchandise news interested with user that matching treatment obtains was mated by the 3rd artificial neural network
User's dependent merchandise Information Number is less, then can turn down the second preset value.Second preset value can also be according to the precision of coupling
Demand is adjusted flexibly, and such as, if desired carries out strong correlation coupling, then can heighten the second preset value；If desired carry out weak
Relevant matches, then can turn down the second preset value.
In this example, user's dependent merchandise information can be ranked up, to facilitate when front end applications, according to difference
Operation demand, represent the commodity of different emphasis.For example, it is possible to user is correlated with business according to similarity descending order
Product information is ranked up；Or, can be in conjunction with similarity and shop grade, shop authentication information to user's dependent merchandise information
It is ranked up.
It should be noted that the 3rd artificial neural network in the embodiment of the present invention can utilize existing ANN
Network technology realizes, and no longer repeats the operation principle of the 3rd artificial neural network at this.
In step s 304, user's relative article corresponding with user's dependent merchandise information in article to be selected is determined.
It should be noted that in this example, by first determining merchandise news to be selected, then screen from merchandise news to be selected
Go out user's dependent merchandise information, in the case of customer volume is relatively big, be conducive to improving matching efficiency.
Fig. 4 illustrates the another exemplary of based on video according to an embodiment of the invention Method of Commodity Recommendation step S104
Implement flow chart.As shown in Figure 4, from article to be selected, filter out user according to the merchandise news that user is interested to be correlated with
Article, and determine the user dependent merchandise information corresponding with user's relative article, including:
In step S401, from article to be selected, filter out user's relative article according to the merchandise news that user is interested.
It will be understood by those skilled in the art that various means of the prior art all can realize the commodity interested according to user
Information filters out user's relative article from article to be selected, and wherein, artificial neural network is exactly one of these means.
As an example of the embodiment of the present invention, filter out from article to be selected according to the merchandise news that user is interested
User's relative article can be: is sieved from article to be selected according to the merchandise news that user is interested by the 4th artificial neural network
Select user's relative article.In this example, the 4th artificial neural network may be used for merchandise news interested in user with
Article to be selected carry out matching treatment, to determine similarity between the two, and then according to user's merchandise news interested from treating
Select and article filter out user's relative article.Such as, the input of the 4th artificial neural network can be the commodity that user is interested
Several pictures of information and the article to be selected sectional drawing in this frame of video, output result can be the commodity letter that user is interested
Breath and the similarity of article to be selected.In this example, can similar by merchandise news interested to user in article to be selected
Degree is defined as user's relative article more than the article to be selected of the 3rd preset value.Such as, the 3rd preset value can be 70%, at this not
It is construed as limiting.It should be noted that the 3rd preset value can be adjusted flexibly according to the precision demand of coupling, such as, if needing
Strong correlation coupling to be carried out, then can heighten the 3rd preset value；If desired carry out weak relevant matches, then can turn down the 3rd and preset
Value.
It should be noted that the 4th artificial neural network in the embodiment of the present invention can utilize existing ANN
Network technology realizes, and no longer repeats the operation principle of the 4th artificial neural network at this.
In step S402, scan for according to the label of user's relative article, obtain the second Search Results.
In this example, scan for according to the label of user's relative article, Ke Yiwei: according to the mark of user's relative article
Sign and scan in the merchandising database that any one or more electricity business websites are corresponding.
In step S403, from the second Search Results, filter out the user's dependent merchandise letter mated with user's relative article
Breath.
It will be understood by those skilled in the art that various means of the prior art all can realize from the second Search Results to screen
Going out the user's dependent merchandise information mated with user's relative article, wherein, artificial neural network is exactly one of these means.
As an example of the embodiment of the present invention, filter out from the second Search Results and mate with user's relative article
User's dependent merchandise information can be: is filtered out from the second Search Results and user's correlative by the 5th artificial neural network
User's dependent merchandise information of product coupling.In this example, the 5th artificial neural network may be used for user's relative article with
Second Search Results carries out matching treatment, to determine similarity between the two, and then filter out from the second Search Results with
User's dependent merchandise information of user's relative article coupling.Such as, the input of the 5th artificial neural network can be the second search
Several pictures of commodity in result and user's relative article sectional drawing in this frame of video, output result can be second to search
Commodity in hitch fruit and the similarity of user's relative article.In this example, can by the second Search Results with user's phase
The similarity closing article is defined as user's dependent merchandise information more than the merchandise news of the 4th preset value.Such as, the 4th preset value
Can be 80%, in this no limit.It should be noted that the 4th preset value can be according to user's dependent merchandise information of coupling
Number is adjusted flexibly, such as, if by the 5th artificial neural network mating with user's relative article of carrying out that matching treatment obtains
User's dependent merchandise Information Number is relatively big, then can heighten the 4th preset value；If being carried out at coupling by the 5th artificial neural network
The user's dependent merchandise Information Number mated with user's relative article that reason obtains is less, then can turn down the 4th preset value.4th
Preset value can also be adjusted flexibly according to the precision demand of coupling, such as, if desired carries out strong correlation coupling, the most permissible
Heighten the 4th preset value；If desired carry out weak relevant matches, then can turn down the 4th preset value.
It should be noted that the 5th artificial neural network in the embodiment of the present invention can utilize existing ANN
Network technology realizes, and no longer repeats the operation principle of the 5th artificial neural network at this.
Fig. 5 illustrates that the one of based on video according to an embodiment of the invention Method of Commodity Recommendation exemplary realizes flow process
Figure.As it is shown in figure 5, the method includes:
In step S501, obtain the frame of video of video file；
In step S502, it is defined to the classification of the article of the first artificial neural networks identification specify classification；
In step S503, it is identified processing, to determine this video to this frame of video by the first artificial neural networks
Article to be selected that frame includes and the label of article to be selected；
In step S504, obtain the merchandise news that user is interested；
In step S505, from article to be selected, filter out user's relative article according to the merchandise news that user is interested,
And determine the user dependent merchandise information corresponding with user's relative article；
In step S506, when playing this frame of video, user's relative article is highlighted, and meeting first
In the case of pre-conditioned, show user's dependent merchandise information.
As an example of the embodiment of the present invention, it is being identified locating to this frame of video by the first artificial neural networks
Before reason, it is defined to the classification of the article of the first artificial neural networks identification specify classification.Such as, it is intended that classification can include
Clothing and accessories.In this example, by the classification of the article of the first artificial neural networks identification is defined to specify classification, fall
The low complexity of the first artificial neural networks, thus improve the recognition efficiency to this frame of video.
Although it should be noted that describe appointment classification as above as example using clothing and accessories, but art technology
Personnel are it is understood that the present invention should be not limited to this.It is true that user completely can be according to personal like and/or actual application scenarios
Set flexibly and specify classification.Such as, it is intended that classification can also be snacks beverage or furniture etc..
Fig. 6 illustrates that based on video according to an embodiment of the invention Method of Commodity Recommendation step S103's is one exemplary
Implement flow chart.As shown in Figure 6, obtain the merchandise news that user is interested, including:
In step s 601, cookie is obtained；
In step S602, obtain, from third party website, the merchandise news that user is interested according to cookie.
As an example of the embodiment of the present invention, the merchandise news obtaining user interested can be: obtains cookie
(being stored in the data on user local terminal), obtains, from third party website, the commodity letter that user is interested further according to cookie
Breath.Wherein, third party website can be electricity business website or other websites, in this no limit.The commodity that user is interested can
To include commodity that commodity that user collects, user bought or the browsed commodity of user.The commodity letter that user is interested
Breath can include purchase link or the picture of the commodity that user is interested.
It should be noted that the merchandise news that existing commercial productainterests Analysis Service analysis user is interested can be utilized,
The operation principle of the merchandise news no longer interested in analysis user at this repeats.
In a kind of possible implementation, from article to be selected, filter out user according to the merchandise news that user is interested
Relative article, and determine the user dependent merchandise information corresponding with user's relative article, particularly as follows: the returning rate user is full
In the case of foot second is pre-conditioned, from article to be selected, filter out user's correlative according to the merchandise news that user is interested
Product, and determine the user dependent merchandise information corresponding with user's relative article.In this implementation, the returning rate of user can
Think that user pays a return visit the returning rate of the video website playing this video file.Meeting second pre-conditioned can be: user returns
Visit rate is more than the 5th preset value.Meeting second pre-conditioned can also be: according to returning rate order from high to low to user's
Returning rate is ranked up, if the returning rate sequence of a certain user is forward, then to meet second pre-conditioned for the returning rate of this user.Example
As, sorting forward can be that sequence is front 30%, in this no limit.In this implementation, returning rate is met second pre-
If the user of condition provides the service that preferentially identifies, thus ensure that the response efficiency of the user of high liveness.
In a kind of possible implementation, the returning rate at family can be employed by cookie memory recording.
In a kind of possible implementation, the label of article to be selected include following at least one: article to be selected regard at this
Frequently the sectional drawing in frame, the classification of article to be selected, the style of article to be selected, the color of article to be selected and the decorative pattern of article to be selected.
Wherein, the article to be selected sectional drawing in this frame of video can determine according to the first rectangle, it is also possible to according to the second square
Shape determines, it is also possible to determine according to the article to be selected border in this frame of video, in this no limit.Wherein, the first rectangle can
Think the boundary rectangle of parallel with the four edges of this frame of video respectively these article to be selected of four edges.Second rectangle can be that this is treated
Select article minimum enclosed rectangle in this frame of video.
Such as, the classification of article to be selected can be jacket, trousers, shoes, bag or necklace etc., and the style of article to be selected can
Thinking shirt, western-style clothes, one-piece dress or jeans etc., the decorative pattern of article to be selected can be horizontal stripe, nicking, pure color, letter
Or round dot etc..
In a kind of possible implementation, the label of article to be selected can also include that article to be selected are in this frame of video
Positional information.The article to be selected positional information in this frame of video can include that the geometric center of article to be selected is in this frame of video
Coordinate and the width of the article to be selected sectional drawing in this frame of video and height.The article to be selected positional information in this frame of video also may be used
To include the coordinate of the article to be selected sectional drawing in this frame of video.Such as, it is square at the article to be selected sectional drawing in this frame of video
In the case of shape, the coordinate of the article to be selected sectional drawing in this frame of video can be the coordinate on four summits of this rectangle.According to
The geometric center of article to be selected coordinate in this frame of video and the width of the article to be selected sectional drawing in this frame of video and height, permissible
Determine the article to be selected position in this frame of video.Coordinate according to the article to be selected sectional drawing in this frame of video, it is also possible to really
The fixed article to be selected position in this frame of video.
In a kind of possible implementation, the preset icon displayed above in user's relative article region is permissible
For: according to user's relative article positional information in this frame of video, displayed above in this user's relative article region
Preset icon.For example, it is possible to show preset icon on the geometric center of this user's relative article coordinate in this frame of video.
So, process, to determine the article to be selected and thing to be selected that this frame of video includes by frame of video is identified
The label of product, filters out user's relative article from article to be selected according to the merchandise news that user is interested, and determines and user
User's dependent merchandise information that relative article is corresponding, when playing this frame of video, highlights user's relative article,
And meet first pre-conditioned in the case of, show user's dependent merchandise information, according to embodiments of the present invention based on video
Method of Commodity Recommendation can recommend its merchandise news that may be interested according to user's merchandise news interested to user, greatly
Improve greatly the efficiency of commercial product recommending based on video, and improve user and watch the experience of video.
Embodiment 2
Fig. 7 illustrates the structured flowchart of based on video the according to another embodiment of the present invention device for recommending the commodity.This device
May be used for the Method of Commodity Recommendation based on video shown in service chart 1,3-6.For convenience of description, illustrate only and the present invention
The part that embodiment is relevant.
As it is shown in fig. 7, this device includes: frame of video acquisition module 71, for obtaining the frame of video of video file；Identify mould
Block 72, processes for being identified described frame of video, to determine article to be selected that described frame of video includes and described to be selected
The label of article；The merchandise news acquisition module 73 that user is interested, for obtaining the merchandise news that user is interested；User's phase
Close article and merchandise news determines module 74, for sieving from described article to be selected according to the merchandise news that described user is interested
Select user's relative article, and determine the user dependent merchandise information corresponding with described user's relative article；Display module 75,
For when playing described frame of video, described user's relative article being highlighted and pre-conditioned meeting first
In the case of, show described user's dependent merchandise information.
Fig. 8 illustrates an exemplary structural frames of based on video the according to another embodiment of the present invention device for recommending the commodity
Figure.The assembly that in Fig. 8, label is identical with Fig. 7 has identical function, for simplicity's sake, omits to these assemblies specifically
Bright.For convenience of description, illustrate only the part relevant to the embodiment of the present invention.As shown in Figure 8:
In a kind of possible implementation, described user's relative article and merchandise news determine that module 74 includes: first
Search submodule 741, for scanning for according to the label of described article to be selected, obtains the first Search Results；Commodity to be selected are believed
Breath screening submodule 742, for filtering out the commodity to be selected letter mated with described article to be selected from described first Search Results
Breath；User's dependent merchandise information matches submodule 743, feels emerging for filtering out from described merchandise news to be selected with described user
Described user's dependent merchandise information of the merchandise news coupling of interest；User's relative article determines submodule 744, is used for determining described
User's relative article corresponding with described user's dependent merchandise information in article to be selected.
In a kind of possible implementation, described user's relative article and merchandise news determine that module 74 includes: user
Relative article screening submodule 745, for filtering out from described article to be selected according to the merchandise news that described user is interested
User's relative article；Second search submodule 746, for scanning for according to the label of described user's relative article, obtains the
Two Search Results；User's dependent merchandise information sifting submodule 747, for filtering out with described from described second Search Results
Described user's dependent merchandise information of user's relative article coupling.
In a kind of possible implementation, described identification module 72 specifically for: by the first artificial neural networks pair
Described frame of video is identified processing.
In a kind of possible implementation, described device also includes: identify that classification limits module 76, for by described the
The classification of the article of one artificial neural network identification is defined to specify classification.
In a kind of possible implementation, the merchandise news acquisition module 73 that described user is interested includes: Cookie
Obtain submodule 731, be used for obtaining cookie；The merchandise news that user is interested obtains submodule 732, for according to described
Cookie obtains, from third party website, the merchandise news that described user is interested.
In a kind of possible implementation, described user's relative article and merchandise news determine module 74 specifically for:
The returning rate of described user meet second pre-conditioned in the case of, according to described user merchandise news interested from described
Article to be selected filter out user's relative article, and determines the user dependent merchandise letter corresponding with described user's relative article
Breath.
In a kind of possible implementation, the label of described article to be selected include following at least one: described thing to be selected
Product sectional drawing in described frame of video, the classification of described article to be selected, the style of described article to be selected, the face of described article to be selected
The decorative pattern of article to be selected described in normal complexion.
It should be noted that so, by frame of video being identified process, to determine the thing to be selected that this frame of video includes
Product and the label of article to be selected, filter out user's relative article according to the merchandise news that user is interested from article to be selected,
And determine the user dependent merchandise information corresponding with user's relative article, when playing this frame of video, to user's relative article
Highlight, and meet first pre-conditioned in the case of, show user's dependent merchandise information, according to the present invention implement
The device for recommending the commodity based on video of example can recommend it possible interested according to the merchandise news that user is interested to user
Merchandise news, substantially increase the efficiency of commercial product recommending based on video, and improve user and watch the experience of video.
Embodiment 3
Fig. 9 illustrates the structured flowchart of commercial product recommending equipment based on video according to another embodiment of the invention.Institute
State commercial product recommending equipment 1100 based on video can be possess the host server of computing capability, personal computer PC or
Portable portable computer or terminal etc..Calculating node is not implemented and limits by the specific embodiment of the invention.
Described commercial product recommending equipment 1100 based on video includes processor (processor) 1110, communication interface
(Communications Interface) 1120, memorizer (memory) 1130 and bus 1140.Wherein, processor 1110,
Communication interface 1120 and memorizer 1130 complete mutual communication by bus 1140.
Communication interface 1120 is used for and network device communications, and wherein the network equipment includes such as Virtual Machine Manager center, is total to
Enjoy storage etc..
Processor 1110 is used for performing program.Processor 1110 is probably a central processor CPU, or special collection
Become circuit ASIC (Application Specific Integrated Circuit), or be configured to implement the present invention
One or more integrated circuits of embodiment.
Memorizer 1130 is used for depositing file.Memorizer 1130 may comprise high-speed RAM memorizer, it is also possible to also includes non-
Volatile memory (non-volatile memory), for example, at least one disk memory.Memorizer 1130 can also be to deposit
Memory array.Memorizer 1130 is also possible to by piecemeal, and described piece can be by certain rule sets synthesis virtual volume.
In a kind of possible embodiment, said procedure can be the program code including computer-managed instruction.This journey
Sequence is particularly used in: realize the operation of each step in embodiment 1.
Those of ordinary skill in the art are it is to be appreciated that each exemplary cell in embodiment described herein and algorithm
Step, it is possible to being implemented in combination in of electronic hardware or computer software and electronic hardware.These functions are actually with hardware also
It is that software form realizes, depends on application-specific and the design constraint of technical scheme.Professional and technical personnel can be for
Specific application selects different methods to realize described function, but this realization is it is not considered that exceed the model of the present invention
Enclose.
If using the form of computer software realize described function and as independent production marketing or use time, then exist
To a certain extent it is believed that all or part of (part such as contributed prior art) of technical scheme is
Embody in form of a computer software product.This computer software product is generally stored inside the non-volatile of embodied on computer readable
In storage medium, including some instructions with so that computer equipment (can be that personal computer, server or network set
Standby etc.) perform all or part of step of various embodiments of the present invention method.And aforesaid storage medium include USB flash disk, portable hard drive,
Read only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic
The various medium that can store program code such as dish or CD.
The above, the only detailed description of the invention of the present invention, but protection scope of the present invention is not limited thereto, and any
Those familiar with the art, in the technical scope that the invention discloses, can readily occur in change or replace, should contain
Cover within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with described scope of the claims.
Claims (16)
1. a Method of Commodity Recommendation based on video, it is characterised in that including:
Obtain the frame of video of video file；
It is identified described frame of video processing, to determine article to be selected that described frame of video includes and described article to be selected
Label；
Obtain the merchandise news that user is interested；
From described article to be selected, filter out user's relative article according to the merchandise news that described user is interested, and determine and institute
State user's dependent merchandise information that user's relative article is corresponding；
When playing described frame of video, described user's relative article is highlighted, and pre-conditioned meeting first
In the case of, show described user's dependent merchandise information.
Method the most according to claim 1, it is characterised in that treat from described according to the merchandise news that described user is interested
Select and article filter out user's relative article, and determine the user dependent merchandise information corresponding with described user's relative article,
Including:
Label according to described article to be selected scans for, and obtains the first Search Results；
The merchandise news to be selected mated with described article to be selected is filtered out from described first Search Results；
The relevant business of described user of the merchandise news coupling interested to described user is filtered out from described merchandise news to be selected
Product information,
Determine user's relative article corresponding with described user's dependent merchandise information in described article to be selected.
Method the most according to claim 1, it is characterised in that treat from described according to the merchandise news that described user is interested
Select and article filter out user's relative article, and determine the user dependent merchandise information corresponding with described user's relative article,
Including:
From described article to be selected, user's relative article is filtered out according to the merchandise news that described user is interested；
Label according to described user's relative article scans for, and obtains the second Search Results；
The described user's dependent merchandise information mated with described user's relative article is filtered out from described second Search Results.
Method the most according to claim 1, it is characterised in that be identified described frame of video processing, particularly as follows: pass through
Described frame of video is identified processing by the first artificial neural networks.
Method the most according to claim 4, it is characterised in that described frame of video is being entered by the first artificial neural networks
Before row identifying processing, described method also includes:
It is defined to the classification of the article of described the first artificial neural networks identification specify classification.
Method the most according to claim 1, it is characterised in that obtain the merchandise news that user is interested, including:
Obtain cookie；
Obtain, from third party website, the merchandise news that described user is interested according to described cookie.
Method the most according to claim 1, it is characterised in that treat from described according to the merchandise news that described user is interested
Select and article filter out user's relative article, and determine the user dependent merchandise information corresponding with described user's relative article,
Particularly as follows:
The returning rate of described user meet second pre-conditioned in the case of, according to described user merchandise news interested from
Described article to be selected filter out user's relative article, and determines the user dependent merchandise corresponding with described user's relative article
Information.
Method the most according to claim 1, it is characterised in that the label of described article to be selected include following at least one:
The described article to be selected sectional drawing in described frame of video, the classification of described article to be selected, the style of described article to be selected, institute
State color and the decorative pattern of described article to be selected of article to be selected.
9. a device for recommending the commodity based on video, it is characterised in that including:
Frame of video acquisition module, for obtaining the frame of video of video file；
Identification module, processes for being identified described frame of video, with determine article to be selected that described frame of video includes and
The label of described article to be selected；
The merchandise news acquisition module that user is interested, for obtaining the merchandise news that user is interested；
User's relative article and merchandise news determine module, for the merchandise news interested according to described user from described to be selected
Article filter out user's relative article, and determines the user dependent merchandise information corresponding with described user's relative article；
Display module, for when playing described frame of video, highlighting described user's relative article, and is meeting the
In the case of one is pre-conditioned, show described user's dependent merchandise information.
Device the most according to claim 9, it is characterised in that described user's relative article and merchandise news determine module
Including:
First search submodule, for scanning for according to the label of described article to be selected, obtains the first Search Results；
Merchandise news to be selected screening submodule, mates with described article to be selected for filtering out from described first Search Results
Merchandise news to be selected；
User's dependent merchandise information matches submodule, interested with described user for filtering out from described merchandise news to be selected
Merchandise news coupling described user's dependent merchandise information；
User's relative article determines submodule, is used for determining in described article to be selected corresponding with described user's dependent merchandise information
User's relative article.
11. devices according to claim 9, it is characterised in that described user's relative article and merchandise news determine module
Including:
User's relative article screening submodule, for sieving from described article to be selected according to the merchandise news that described user is interested
Select user's relative article；
Second search submodule, for scanning for according to the label of described user's relative article, obtains the second Search Results；
User's dependent merchandise information sifting submodule, for filtering out and described user's correlative from described second Search Results
Described user's dependent merchandise information of product coupling.
12. devices according to claim 9, it is characterised in that described identification module specifically for: by the first artificial god
It is identified processing to described frame of video through network.
13. devices according to claim 12, it is characterised in that described device also includes:
Identify that classification limits module, for being defined to specify class by the classification of the article of described the first artificial neural networks identification
Not.
14. devices according to claim 9, it is characterised in that the merchandise news acquisition module bag that described user is interested
Include:
Cookie obtains submodule, is used for obtaining cookie；
The merchandise news that user is interested obtains submodule, for obtaining described user according to described cookie from third party website
Merchandise news interested.
15. devices according to claim 9, it is characterised in that described user's relative article and merchandise news determine module
Specifically for:
The returning rate of described user meet second pre-conditioned in the case of, according to described user merchandise news interested from
Described article to be selected filter out user's relative article, and determines the user dependent merchandise corresponding with described user's relative article
Information.
16. devices according to claim 9, it is characterised in that the label of described article to be selected include following at least one:
The described article to be selected sectional drawing in described frame of video, the classification of described article to be selected, the style of described article to be selected, institute
State color and the decorative pattern of described article to be selected of article to be selected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610511072.1A CN106202304A (en) | 2016-07-01 | 2016-07-01 | Method of Commodity Recommendation based on video and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610511072.1A CN106202304A (en) | 2016-07-01 | 2016-07-01 | Method of Commodity Recommendation based on video and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106202304A true CN106202304A (en) | 2016-12-07 |
Family
ID=57464551
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610511072.1A Pending CN106202304A (en) | 2016-07-01 | 2016-07-01 | Method of Commodity Recommendation based on video and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106202304A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107833082A (en) * | 2017-09-15 | 2018-03-23 | 广州唯品会研究院有限公司 | A kind of recommendation method and apparatus of commodity picture |
| CN108509436A (en) * | 2017-02-24 | 2018-09-07 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and computer storage media of determining recommended |
| CN109389182A (en) * | 2018-10-31 | 2019-02-26 | 北京字节跳动网络技术有限公司 | Method and apparatus for generating information |
| CN110059256A (en) * | 2019-04-26 | 2019-07-26 | 北京沃东天骏信息技术有限公司 | For showing system, the method and device of information |
| CN110099301A (en) * | 2019-04-28 | 2019-08-06 | 深圳回收宝科技有限公司 | The method, apparatus and portable terminal of terminal detection are carried out based on interception image |
| CN110807664A (en) * | 2019-10-31 | 2020-02-18 | 深圳市云积分科技有限公司 | Cross-platform customer marketing method and device |
| CN111225250A (en) * | 2020-01-14 | 2020-06-02 | 腾讯科技（深圳）有限公司 | Video extended information processing method and device |
| CN111813944A (en) * | 2020-09-09 | 2020-10-23 | 北京神州泰岳智能数据技术有限公司 | Live comment analysis method and device, electronic equipment and storage medium |
| CN112653900A (en) * | 2020-12-21 | 2021-04-13 | Oppo广东移动通信有限公司 | Target display method, device and equipment in video live broadcast |
| WO2021093386A1 (en) * | 2019-11-14 | 2021-05-20 | 北京沃东天骏信息技术有限公司 | Article information processing method, article information processing apparatus, medium, and electronic device |
| CN113760158A (en) * | 2021-04-30 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Target object display method, object association method, device, medium and equipment |
| CN114501043A (en) * | 2021-12-24 | 2022-05-13 | 中国电信股份有限公司 | Video pushing method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115113A1 (en) * | 2001-12-13 | 2003-06-19 | Duncan Ross W. | Method and apparatus for making recommendations |
| CN102802055A (en) * | 2011-05-25 | 2012-11-28 | 阿里巴巴集团控股有限公司 | Data interaction method and device based on on-line video |
| CN103037262A (en) * | 2011-10-05 | 2013-04-10 | Lg电子株式会社 | Display device for displaying meta data according to command signal of remote controller and control method of the same |
| CN103235678A (en) * | 2013-03-22 | 2013-08-07 | 东莞宇龙通信科技有限公司 | Display method for background information of specific target in video and video terminal |
| CN104080007A (en) * | 2014-07-17 | 2014-10-01 | 合一网络技术(北京)有限公司 | Method and system for generating commodity trading information in video playing |
| CN104715023A (en) * | 2015-03-02 | 2015-06-17 | 北京奇艺世纪科技有限公司 | Commodity recommendation method and system based on video content |
-
2016
- 2016-07-01 CN CN201610511072.1A patent/CN106202304A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115113A1 (en) * | 2001-12-13 | 2003-06-19 | Duncan Ross W. | Method and apparatus for making recommendations |
| CN102802055A (en) * | 2011-05-25 | 2012-11-28 | 阿里巴巴集团控股有限公司 | Data interaction method and device based on on-line video |
| CN103037262A (en) * | 2011-10-05 | 2013-04-10 | Lg电子株式会社 | Display device for displaying meta data according to command signal of remote controller and control method of the same |
| CN103235678A (en) * | 2013-03-22 | 2013-08-07 | 东莞宇龙通信科技有限公司 | Display method for background information of specific target in video and video terminal |
| CN104080007A (en) * | 2014-07-17 | 2014-10-01 | 合一网络技术(北京)有限公司 | Method and system for generating commodity trading information in video playing |
| CN104715023A (en) * | 2015-03-02 | 2015-06-17 | 北京奇艺世纪科技有限公司 | Commodity recommendation method and system based on video content |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108509436B (en) * | 2017-02-24 | 2022-02-18 | 阿里巴巴集团控股有限公司 | Method and device for determining recommended object and computer storage medium |
| CN108509436A (en) * | 2017-02-24 | 2018-09-07 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and computer storage media of determining recommended |
| CN107833082A (en) * | 2017-09-15 | 2018-03-23 | 广州唯品会研究院有限公司 | A kind of recommendation method and apparatus of commodity picture |
| CN107833082B (en) * | 2017-09-15 | 2022-04-12 | 唯品会(海南)电子商务有限公司 | Commodity picture recommendation method and device |
| CN109389182A (en) * | 2018-10-31 | 2019-02-26 | 北京字节跳动网络技术有限公司 | Method and apparatus for generating information |
| CN110059256A (en) * | 2019-04-26 | 2019-07-26 | 北京沃东天骏信息技术有限公司 | For showing system, the method and device of information |
| CN110059256B (en) * | 2019-04-26 | 2023-11-07 | 北京沃东天骏信息技术有限公司 | System, method and device for displaying information |
| CN110099301A (en) * | 2019-04-28 | 2019-08-06 | 深圳回收宝科技有限公司 | The method, apparatus and portable terminal of terminal detection are carried out based on interception image |
| CN110807664A (en) * | 2019-10-31 | 2020-02-18 | 深圳市云积分科技有限公司 | Cross-platform customer marketing method and device |
| WO2021093386A1 (en) * | 2019-11-14 | 2021-05-20 | 北京沃东天骏信息技术有限公司 | Article information processing method, article information processing apparatus, medium, and electronic device |
| CN111225250B (en) * | 2020-01-14 | 2021-11-09 | 腾讯科技（深圳）有限公司 | Video extended information processing method and device |
| CN111225250A (en) * | 2020-01-14 | 2020-06-02 | 腾讯科技（深圳）有限公司 | Video extended information processing method and device |
| CN111813944A (en) * | 2020-09-09 | 2020-10-23 | 北京神州泰岳智能数据技术有限公司 | Live comment analysis method and device, electronic equipment and storage medium |
| CN112653900A (en) * | 2020-12-21 | 2021-04-13 | Oppo广东移动通信有限公司 | Target display method, device and equipment in video live broadcast |
| CN113760158A (en) * | 2021-04-30 | 2021-12-07 | 腾讯科技（深圳）有限公司 | Target object display method, object association method, device, medium and equipment |
| CN114501043A (en) * | 2021-12-24 | 2022-05-13 | 中国电信股份有限公司 | Video pushing method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106202304A (en) | Method of Commodity Recommendation based on video and device | |
| CN106202316A (en) | Merchandise news acquisition methods based on video and device | |
| CN106055710A (en) | Video-based commodity recommendation method and device | |
| US20210406987A1 (en) | Recommendation system, method and computer program product based on a user's physical features | |
| KR101836056B1 (en) | Image feature data extraction and use | |
| KR102664008B1 (en) | Method and device to recommend customer item based on visual information | |
| US8718369B1 (en) | Techniques for shape-based search of content | |
| US11538083B2 (en) | Cognitive fashion product recommendation system, computer program product, and method | |
| CN106202317A (en) | Method of Commodity Recommendation based on video and device | |
| US9727620B2 (en) | System and method for item and item set matching | |
| CN109635680A (en) | Multitask attribute recognition approach, device, electronic equipment and storage medium | |
| CN108713212A (en) | Method, system and medium for providing information based on grouping information | |
| JP2022501726A (en) | Style recommendation method, equipment and computer program | |
| CN111861678B (en) | Resource recommendation method and device, electronic equipment and storage medium | |
| CN107666435A (en) | A kind of method and device for shielding message | |
| CN108614832A (en) | A kind of user individual commercial articles searching implementation method and device | |
| CN107515868A (en) | Searching method and device | |
| CN111815407A (en) | Method and device for constructing user portrait | |
| CN107993103A (en) | A kind of wearing article based on augmented reality recommends method and apparatus | |
| CN108596646B (en) | Garment matching recommendation method integrating face attribute analysis | |
| Deshbhratar et al. | Augmented reality of online and physical retailing: A study of applications and its value | |
| CN109146606A (en) | A kind of brand recommended method, electronic equipment, storage medium and system | |
| KR102102572B1 (en) | System and method for providing online shopping mall | |
| CN110348925A (en) | Shops's system, article matching method, device and electronic equipment | |
| CN106651500B (en) | Online shopping system based on video image recognition technology and virtual reality technology of spatial feedback characteristics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20200521 Address after: 310052 room 508, floor 5, building 4, No. 699, Wangshang Road, Changhe street, Binjiang District, Hangzhou City, Zhejiang Province Applicant after: Alibaba (China) Co.,Ltd. Address before: 200241, room 2, floor 02, building 555, Dongchuan Road, Minhang District, Shanghai Applicant before: Transmission network technology (Shanghai) Co., Ltd |
|
| TA01 | Transfer of patent application right | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161207 |
|
| RJ01 | Rejection of invention patent application after publication |