CN107404657B - Advertisement recommendation method and device - Google Patents
Advertisement recommendation method and device Download PDFInfo
- Publication number
- CN107404657B CN107404657B CN201710533587.6A CN201710533587A CN107404657B CN 107404657 B CN107404657 B CN 107404657B CN 201710533587 A CN201710533587 A CN 201710533587A CN 107404657 B CN107404657 B CN 107404657B
- Authority
- CN
- China
- Prior art keywords
- candidate
- score
- advertisement
- videos
- video
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/254—Management at additional data server, e.g. shopping server, rights management server
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
- H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
- H04N21/4668—Learning process for intelligent management, e.g. learning user preferences for recommending movies for recommending content, e.g. movies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/81—Monomedia components thereof
- H04N21/812—Monomedia components thereof involving advertisement data
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Business, Economics & Management (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Computing Systems (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an advertisement recommendation method and device, wherein the method comprises the following steps: respectively calculating a relevance score between each candidate advertisement and the video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video respectively corresponds to a plurality of candidate advertisements; aiming at each video, selecting a relevance score from the relevance scores of a plurality of candidate advertisements of each video according to a first preset mode, and recording the relevance score as the advertisement candidate score of each video; selecting two target advertisement candidate scores S1 and S2 from the plurality of advertisement candidate scores according to a second preset mode; mapping the first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ]; acquiring a second score x set in [ M, N ], mapping x to [ S1, S2], and obtaining a first score S; and obtaining a plurality of target candidate advertisements with the relevance scores larger than or equal to S from a plurality of candidate advertisements of each video, and recommending the target candidate advertisements to the video. The recommended advertisement can improve the correlation between the advertisement and the video.
Description
Technical Field
The invention relates to the technical field of advertisement recommendation, in particular to an advertisement recommendation method and device.
Background
The videos can provide a wide information browsing path for users by delivering advertisements, and are influenced by advertisement exposure and delivery strategies, the relevance scores of a plurality of candidate advertisements and the videos in each video are dynamic and change with time (for example, the relevance scores of today and yesterday may be different), and the relevance scores of different candidate advertisements are different greatly in the same or different videos. In the prior art, the truncation scheme of the candidate advertisements is mainly used for comparing and judging the relevance scores of the candidate advertisements in all videos, and the candidate advertisements with the scores larger than a threshold value are taken as recommended advertisements to be delivered in the videos. But the inventor finds a problem in implementing the present invention that the correlation between the recommended advertisement and the video is low.
Disclosure of Invention
The invention provides an advertisement recommendation method and device, which aim to solve the problem that in the prior art, when advertisement recommendation is performed on videos, the relevance between recommended advertisements and the videos is low.
In order to solve the above problem, according to an aspect of the present invention, there is disclosed an advertisement recommendation method including:
respectively calculating a relevance score between each candidate advertisement and the video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video respectively corresponds to a plurality of candidate advertisements;
aiming at each video in the plurality of videos, respectively selecting the relevance score of a candidate advertisement from the relevance scores of a plurality of candidate advertisements of each video according to a first preset mode, and recording the relevance score as the advertisement candidate score of each video;
selecting two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of a plurality of videos for which advertisements need to be recommended according to a second preset mode, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2;
forming a first relevance score interval [ S1, S2] with the two target advertisement candidate scores S1 and S2 as endpoints;
mapping the first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ], wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, and M is less than N;
acquiring a second score x set in a second relevance score interval [ M, N ], mapping the second score x to the first relevance score interval [ S1, S2], and acquiring a first score S, wherein the second score x is greater than or equal to M and less than or equal to N, and the first score S is greater than or equal to a target advertisement candidate score S1 and less than or equal to a target advertisement candidate score S2;
and obtaining a plurality of target candidate advertisements with the relevance scores larger than or equal to the first score S from a plurality of candidate advertisements of each video, and recommending the target candidate advertisements to the video.
According to another aspect of the present invention, the present invention also discloses an advertisement recommendation apparatus, comprising:
the calculation module is used for calculating a relevance score between each candidate advertisement and the video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video corresponds to a plurality of candidate advertisements;
the first selection module is used for selecting the relevance score of a candidate advertisement from the relevance scores of the candidate advertisements of each video according to a first preset mode aiming at each video of the videos and recording the relevance score as the advertisement candidate score of each video;
the second selection module is used for selecting two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of a plurality of videos for which advertisements need to be recommended according to a second preset mode, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2;
a score interval generation module for constituting a first relevance score interval [ S1, S2] with the two target advertisement candidate scores S1 and S2 as endpoints;
a first mapping module for mapping the first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ], wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, and M is less than N;
a second mapping module, configured to obtain a second score x set in a second relevance score interval [ M, N ], map the second score x to the first relevance score interval [ S1, S2], and obtain a first score S, where the second score x is greater than or equal to M and less than or equal to N, and the first score S is greater than or equal to a target advertisement candidate score S1 and less than or equal to a target advertisement candidate score S2;
and the recommending module is used for acquiring a plurality of target candidate advertisements with the relevance scores larger than or equal to the first score S from a plurality of candidate advertisements of each video and recommending the target candidate advertisements to the video.
Compared with the prior art, the invention has the following advantages:
according to the method, the advertisement candidate scores are set for the videos needing to be recommended, the relevance score intervals [ S1, S2] are generated, the [ S1, S2] are mapped to the relevance score intervals [ M, N ] with fixed values, therefore when advertisement recommendation is conducted, a user only needs to select the second score x from the relevance score intervals [ M, N ], the first score S mapped with the second score x can be obtained in [ S1, S2], finally, the candidate advertisements with the relevance scores larger than or equal to the threshold are selected to serve as the advertisements recommended to the videos with the first score S serving as the threshold, the candidate advertisements with high relevance with the videos can be recommended, the relevance between the recommended advertisements and the videos is improved, and the watching experience of the user on the advertisements in the videos is improved.
Drawings
FIG. 1 is a flow chart of steps of an embodiment of a method of advertisement recommendation of the present invention;
fig. 2 is a block diagram of an embodiment of an advertisement recommendation apparatus according to the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Referring to fig. 1, a flowchart illustrating steps of an embodiment of an advertisement recommendation method according to the present invention is shown, which may specifically include the following steps:
step 101, respectively calculating a relevance score between each candidate advertisement and a video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video respectively corresponds to a plurality of candidate advertisements;
for a plurality of videos needing advertisement recommendation, each video corresponds to a plurality of candidate advertisements, wherein the candidate advertisements can include advertisements played by the video in history and corresponding types of advertisements (used for future playing) determined according to click rates of the played advertisements.
Here, the relevance score between each candidate advertisement and the video to which the candidate advertisement belongs may be calculated for a plurality of videos, respectively, wherein the calculation model with the relevance score is trained in advance in the calculation of the relevance score, so that the relevance score between the candidate advertisement and the video to which the candidate advertisement belongs may be output for each candidate advertisement. Wherein the higher the score, the greater the correlation. For example, video a corresponds to 3 candidate ads, then the video a has 3 relevance scores for the 3 candidate ads.
However, for the same video, the relevance scores of the candidate advertisements of the video are constantly changing in different time periods, and for different videos, the relevance scores of the candidate advertisements are also constantly changing, so that if a uniform threshold is adopted to judge the candidate advertisements with large relevance scores to recommend, it is not reliable, and therefore, the embodiment of the invention provides the following steps:
step 102, aiming at each video in the plurality of videos, selecting a relevance score of a candidate advertisement from relevance scores of a plurality of candidate advertisements of each video according to a first preset mode, and recording the relevance score as an advertisement candidate score of each video;
for example, the number of videos to be recommended is 100, each video has 300 candidate advertisements, where, taking video a as an example, the relevance score corresponding to video a is the same as the number of candidate advertisements, that is, 300, and each candidate advertisement corresponds to a relevance score, then the relevance score of a candidate advertisement may be selected from the 300 relevance scores according to a first preset manner, and is recorded as the advertisement candidate score of video a. Similarly, the advertisement candidate scores of the videos other than the video a among the 100 videos may be determined in a manner similar to that of the video a and may be referred to each other. Thus, 100 videos have corresponding 100 advertisement candidate scores, where each video has a corresponding advertisement candidate score.
For the first preset mode, mathematical modes such as selecting the relevance score of a candidate advertisement according to the ranking of the relevance scores, selecting the relevance score of a candidate advertisement according to the normal distribution condition of the relevance scores, randomly extracting, selecting the relevance score of a candidate advertisement from the relevance scores, and the like are within the protection scope of the present invention.
103, selecting two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of a plurality of videos for which advertisements need to be recommended according to a second preset mode;
wherein the target advertisement candidate score S1 is less than the target advertisement candidate score S2;
that is, two advertisement candidate scores S1 and S2 are selected from the corresponding 100 advertisement candidate scores of the 100 videos in the second preset manner.
The second preset mode may be a mathematical mode that selects two candidate advertisement scores from the multiple advertisement candidate scores according to the ranking of the relevance scores, or selects two candidate advertisement scores from the multiple advertisement candidate scores according to the normal distribution condition of the relevance scores, or selects two candidate advertisement scores from the multiple advertisement candidate scores by a random extraction mode, and the like, which are all within the protection scope of the present invention.
Step 104, forming a first relevance score interval [ S1, S2] by taking the two target advertisement candidate scores S1 and S2 as endpoints;
step 105, mapping said first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ];
wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, and M is less than N;
where M and N are both two preset unified thresholds, the user may flexibly set in advance according to actual needs, for example, M is 1 and N is 10, or M is 0 and N is 1.
Step 106, acquiring a second score x set in a second relevance score interval [ M, N ], mapping the second score x to the first relevance score interval [ S1, S2], and obtaining a first score S;
the second fraction x is greater than or equal to M and less than or equal to N, and the first fraction S is greater than or equal to S1 and less than or equal to S2;
then once the values of M and N are set, a relevance score may be selected from M, N as a relevance threshold for recommending advertisements when recommending advertisements. For example, x is 1 in [0,1] is selected as a relevance threshold, and since the score intervals [ S1, S2] and [0,1] are mapped in advance, the first score S is S2, so that even if the relevance scores of the candidate advertisements change, so that [ S1, S2] change, the mapping relation between [ S1, S2] and [ M, N ] is not changed, and the values of the two endpoints of [ M, N ] are fixed, the relevance threshold, i.e., the first score S, of the embodiment of the present invention is flexible, and can change according to the changed relevance scores of the respective candidate advertisements, so that the adaptive adjustment of the relevance score threshold on the overall change trend of the relevance scores is realized, and the difference of the relevance scores of the candidate advertisements under the same video and different videos is taken into account.
And step 107, acquiring a plurality of target candidate advertisements with the relevance scores larger than or equal to the first score S from the candidate advertisements of each video, and recommending the target candidate advertisements to the video.
Finally, advertisements can be selected from the 300 candidate advertisements of each of the 100 videos and recommended to the corresponding video, for example, for the advertisement recommended by the video a, only the target candidate advertisement with the relevance score greater than or equal to the first score S needs to be obtained from the candidate advertisements of the video a as the advertisement recommended to the video a.
By means of the technical scheme of the above embodiment of the invention, the invention sets the advertisement candidate score aiming at the video needing to recommend the advertisement, and generates a correlation score interval [ S1, S2], and by mapping [ S1, S2] to a fixed-value correlation score interval [ M, N ], therefore, when the advertisement recommendation is carried out, the user only needs to select the second score x from the relevance score interval [ M, N ] to obtain the first score S mapped with the second score x in [ S1, S2], and finally, the candidate advertisements with the relevance scores larger than or equal to the threshold are selected as the advertisements recommended to the videos by taking the first score S as the threshold, the candidate advertisements with high relevance to the videos can be recommended, the relevance between the recommended advertisements and the videos is improved, and the watching experience of the users on the advertisements in the videos is improved.
In one embodiment, when step 102 is executed, the following steps may be implemented:
sorting the candidate advertisements of each of the plurality of videos in an order from high to low of the relevance scores; for example, video A (target video) has 300 candidate advertisements, and the 300 advertisements are ranked in order of their relevance scores to video A from high to low. The candidate advertisements of the other videos than video a in the 100 videos are sorted in the same manner as video a, and are not described herein again.
Aiming at the plurality of candidate advertisements sequenced by each video, obtaining the relevance score of the candidate advertisement ranked as the Kth, and recording the relevance score as the advertisement candidate score of a certain video currently calculated, wherein K is a positive integer and is a preset threshold value; for example, for video a, the candidate advertisement ranked as 100 th may be selected from the 300 ranked candidate advertisements, and the relevance score of the 100 th candidate advertisement may be obtained as the advertisement candidate score of video a.
And the other videos except the video A in the 100 videos also perform the same operation as the video A, so as to obtain the respective advertisement candidate scores.
Wherein the number of candidate advertisements for different videos may be the same or different.
However, the K is a fixed threshold, and whether the total number of candidate advertisements is consistent among different videos, the advertisement candidate scores of the videos are all the relevance scores of the candidate advertisements ranked at the K are selected as the advertisement candidate scores of each video.
In addition, it should be noted that the value of K can be flexibly set according to the actual application scenario, for the a video website, the value of K is 100, and for the B video website, the value of K may be 150, which needs to be reasonably adjusted according to the specific application scenario.
In addition, if the total number of candidate advertisements of a plurality of videos that need to be recommended is not exactly the same, the value of K needs to be less than or equal to the sum of the number of candidate advertisements of the video with the fewest candidate advertisements.
And, in another embodiment, when step 103 is executed, it can be implemented as follows:
sequencing a plurality of videos needing to recommend advertisements according to the sequence of the advertisement candidate scores from low to high;
selecting two target videos (for example, 75% of the total number of the ordered videos) arranged at a first preset position (for example, 25% of the total number of the ordered videos) and a second preset position from the ordered videos; specifically, there are 100 videos, then the sorted videos a ranked at the 25 th position and the videos B ranked at the 75 th position may be selected.
It should be noted that the first preset position and the second preset position may be positions arranged in a predetermined percentage of the total number of videos in the above specific example, or positions arranged in the several (e.g., 8 th, 70 th) of the total number of videos.
Finally, two target advertisement candidate scores S1 and S2 for the two target videos are obtained, wherein the target advertisement candidate score S1 is less than the target advertisement candidate score S2.
Wherein, the predetermined percentage is used to determine the preset position to ensure that the advertisement candidate scores of the two taken target videos are distributed in a uniform manner in the advertisement candidate scores of the videos in the whole sequence, and the selected S1 and S2 are more capable of reflecting the correlation between the videos and the advertisements than the manner of directly using the number of the video sums. Moreover, the truncation threshold (namely the first score S is dynamically bound with the total number of the candidate advertisements, so that the recommendation quantity of the relevant advertisements is ensured) can be dynamically bound by adopting a mode of a preset percentage
In addition, it is to be noted that the percentages 25% and 75% of the sum of the video numbers are only specific examples as long as the sum of the two percentages is 1, that is, the percentages of the sum of the video numbers may also be 20% and 80%, and the like.
Optionally, in an embodiment, when step 107 is executed, the following manner may be adopted to implement:
determining a first type of video with an advertisement candidate score being a first score S from a plurality of videos needing to recommend advertisements;
wherein the first score S can be calculated according to the following formula:
S＝(S2-S1)*x+S1。
for example, [ S1, S2 ═ 0.2,0.6], [ M, N ═ 0,1], and x ═ 1, then S ═ 0.6 can be calculated from the above formula.
For example, if the number of videos for which advertisements are to be recommended is 100, then the above-described sorting of the embodiment may result in 100 videos sorted from low to high in terms of advertisement candidate scores, where the video with the high advertisement candidate score is ranked behind the video with the low advertisement candidate score. And each video has an advertisement candidate score, a video with an advertisement candidate score of 0.6 may be determined as a first type of video (e.g., a video ranked at position 75).
Then, selecting a plurality of second-class videos (namely, a plurality of videos arranged at 76 th to 100 th positions) arranged behind the first-class videos from the plurality of sequenced videos;
since the videos are ranked from low to high according to the advertisement candidate scores and the 300 candidate advertisements of each video are ranked from high to low according to the relevance scores, 1 st to K (for example, K is 100 mentioned above) ranked target candidate advertisements can be respectively obtained as advertisements recommended to the first type of video and the second type of video for the multiple candidate advertisements ranked after the first type of video and the multiple second type of video (wherein, since the relevance score of the K ranked candidate advertisement is the advertisement candidate score of the corresponding video, the advertisements ranked from 1 st to K are obtained here for recommendation, that is, the videos ranked from 75 th to 100 th, and the relevance scores of the candidate advertisements ranked from 1 st to K of each video are all greater than or equal to 0.6);
and selecting a plurality of third-class videos arranged before the first-class videos from the sorted videos (namely, 75 videos arranged in the 1 st to 74 th rows, wherein the relevance scores of the candidate advertisements arranged in the 1 st to 99 th rows of the 75 videos are more than 0.6 although the relevance scores of the candidate advertisements arranged in the 1 st to 99 th rows are less than 0.6); therefore, for a plurality of candidate advertisements ranked by a third type of video, traversing the relevance scores of each candidate advertisement in the order from high to low according to the relevance scores, and stopping traversing until the relevance scores are smaller than the first score S; and taking the candidate advertisements with the relevance scores larger than or equal to the first score S as the advertisements recommended to the third type of videos.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the illustrated order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments of the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the invention.
Corresponding to the method provided by the embodiment of the present invention, referring to fig. 2, a block diagram of an advertisement recommendation apparatus according to an embodiment of the present invention is shown, which may specifically include the following modules:
the calculation module 21 is configured to calculate, for a plurality of videos for which advertisements need to be recommended, a relevance score between each candidate advertisement and the corresponding video, where each video corresponds to a plurality of candidate advertisements;
a first selecting module 22, configured to select, for each video in the multiple videos, a relevance score of a candidate advertisement from relevance scores of multiple candidate advertisements of each video according to a first preset manner, and record the relevance score as an advertisement candidate score of each video;
the second selecting module 23 is configured to select two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of a plurality of videos for which an advertisement needs to be recommended according to a second preset manner, where the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2;
a score interval generation module 24 for constituting a first relevance score interval [ S1, S2] with the two target advertisement candidate scores S1 and S2 as endpoints;
a first mapping module 25 for mapping said first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ], wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, M is less than N;
a second mapping module 26, configured to obtain a second score x set in a second relevance score interval [ M, N ], map the second score x to the first relevance score interval [ S1, S2], and obtain a first score S, where the second score x is greater than or equal to M and less than or equal to N, and the first score S is greater than or equal to the target advertisement candidate score S1 and less than or equal to the target advertisement candidate score S2;
and the recommending module 27 is configured to obtain, from the candidate advertisements of each video, a plurality of target candidate advertisements with relevance scores greater than or equal to the first score S and recommend the target candidate advertisements to the video.
Optionally, the first selecting module 22 includes:
a first ordering submodule, configured to order the candidate advertisements of each of the videos in an order from high to low according to the relevance scores;
the first obtaining sub-module is configured to obtain, for the multiple candidate advertisements ranked by each video, a relevance score of a candidate advertisement ranked as the kth, and record the relevance score as an advertisement candidate score of the target video, where K is a positive integer and K is a preset threshold.
Optionally, the second selecting module 23 includes:
the second sequencing submodule is used for sequencing a plurality of videos needing to recommend the advertisements according to the sequence of the candidate advertisement scores from low to high;
the first selection submodule is used for selecting two target videos which are arranged at a first preset position and a second preset position from the sequenced videos;
a second obtaining sub-module for obtaining two target advertisement candidate scores S1 and S2 of the two target videos, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2.
Optionally, the recommending module 27 includes:
the determining submodule is used for determining a first type of video with an advertisement candidate score being a first score S from a plurality of videos needing to recommend advertisements;
a second selection module, configured to select, from the sorted videos, a plurality of second-type videos arranged behind the first-type video;
the third acquisition module is used for respectively acquiring 1 st to K th ranked target candidate advertisements as advertisements recommended to the first type of videos and the second type of videos aiming at the candidate advertisements sequenced by the first type of videos and the second type of videos;
a third selecting module, configured to select, from the sorted videos, a plurality of third videos arranged before the first video;
the traversing module is used for traversing the relevance scores of the candidate advertisements according to the sequence of the relevance scores from high to low aiming at the candidate advertisements sequenced by the third type of videos, and stopping the traversal until the relevance scores are smaller than the first score S;
and the recommending submodule is used for taking the candidate advertisements with the relevance scores larger than or equal to the first score S obtained through traversal as the advertisements recommended to the third type of videos.
Optionally, the first score S is calculated according to the following formula:
S＝(S2-S1)*x+S1。
for the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
The advertisement recommendation method and the advertisement recommendation device provided by the invention are introduced in detail, and the principle and the implementation mode of the invention are explained by applying specific examples, and the description of the examples is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. An advertisement recommendation method, comprising:
respectively calculating a relevance score between each candidate advertisement and the video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video respectively corresponds to a plurality of candidate advertisements;
aiming at each video in the plurality of videos, respectively selecting the relevance score of a candidate advertisement from the relevance scores of a plurality of candidate advertisements of each video according to a first preset mode, and recording the relevance score as the advertisement candidate score of each video;
selecting two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of the videos for which advertisements need to be recommended according to a second preset mode, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2;
forming a first relevance score interval [ S1, S2] with the two target advertisement candidate scores S1 and S2 as endpoints;
mapping the first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ], wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, and M is less than N;
acquiring a second score x set in a second relevance score interval [ M, N ], mapping the second score x to the first relevance score interval [ S1, S2], and acquiring a first score S, wherein the second score x is greater than or equal to M and less than or equal to N, and the first score S is greater than or equal to a target advertisement candidate score S1 and less than or equal to a target advertisement candidate score S2;
and obtaining a plurality of target candidate advertisements with the relevance scores larger than or equal to the first score S from a plurality of candidate advertisements of each video, and recommending the target candidate advertisements to the video.
2. The method according to claim 1, wherein the step of selecting, for each video of the plurality of videos, a relevance score of a candidate advertisement from relevance scores of a plurality of candidate advertisements of the each video in a first preset manner, and recording the relevance score as the advertisement candidate score of the each video comprises:
sorting the candidate advertisements of each of the plurality of videos in an order from high to low of the relevance scores;
and aiming at a plurality of candidate advertisements sequenced by each video, obtaining the relevance score of the candidate advertisement ranked as the Kth, and recording the relevance score as the advertisement candidate score of the target video, wherein K is a positive integer and is a preset threshold value.
3. The method of claim 2, wherein the step of selecting two target advertisement candidate scores S1 and S2 from the advertisement candidate scores of the videos for which an advertisement needs to be recommended according to the second preset manner comprises:
sequencing a plurality of videos needing to recommend advertisements according to the sequence of the advertisement candidate scores from low to high;
selecting two target videos for arranging a first preset position and a second preset position from the sequenced videos;
two target advertisement candidate scores S1 and S2 for the two target videos are obtained, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2.
4. The method of claim 3, wherein the step of obtaining a plurality of target candidate advertisements with relevancy scores greater than or equal to the first score S from among the candidate advertisements of each video, and recommending the target candidate advertisements to the video comprises:
determining a first type of video with an advertisement candidate score being a first score S from a plurality of videos needing to recommend advertisements;
selecting a plurality of second-type videos arranged behind the first-type videos from the sequenced videos;
aiming at a plurality of candidate advertisements sequenced from a first type of video to a plurality of second type of videos, respectively acquiring 1 st to K th ranked target candidate advertisements as advertisements recommended to the first type of video and the second type of video;
selecting a plurality of third-class videos arranged in front of the first-class videos from the sequenced videos;
for a plurality of candidate advertisements sequenced by a third type of videos, traversing the relevance scores of the candidate advertisements in a sequence from high to low according to the relevance scores, and stopping traversing until the relevance scores are smaller than the first score S;
and taking the candidate advertisements with the relevance scores larger than or equal to the first score S as the advertisements recommended to the third type of videos.
5. The method of claim 1, wherein the first score, S, is calculated according to the following equation:
S＝(S2-S1)*x+S1。
6. an advertisement recommendation apparatus, comprising:
the calculation module is used for calculating a relevance score between each candidate advertisement and the video to which the candidate advertisement belongs aiming at a plurality of videos of which the advertisement needs to be recommended, wherein each video corresponds to a plurality of candidate advertisements;
the first selection module is used for selecting the relevance score of a candidate advertisement from the relevance scores of the candidate advertisements of each video according to a first preset mode aiming at each video of the videos and recording the relevance score as the advertisement candidate score of each video;
the second selection module is used for selecting two target advertisement candidate scores S1 and S2 from a plurality of advertisement candidate scores of a plurality of videos for which advertisements need to be recommended according to a second preset mode, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2;
a score interval generation module for constituting a first relevance score interval [ S1, S2] with the two target advertisement candidate scores S1 and S2 as endpoints;
a first mapping module for mapping the first relevance score interval [ S1, S2] to a second relevance score interval [ M, N ], wherein M is a positive number greater than or equal to 0, N is a positive number greater than 0, and M is less than N;
a second mapping module, configured to obtain a second score x set in a second relevance score interval [ M, N ], map the second score x to the first relevance score interval [ S1, S2], and obtain a first score S, where the second score x is greater than or equal to M and less than or equal to N, and the first score S is greater than or equal to a target advertisement candidate score S1 and less than or equal to a target advertisement candidate score S2;
and the recommending module is used for acquiring a plurality of target candidate advertisements with the relevance scores larger than or equal to the first score S from a plurality of candidate advertisements of each video and recommending the target candidate advertisements to the video.
7. The apparatus of claim 6, wherein the first selecting module comprises:
a first ordering submodule, configured to order the candidate advertisements of each of the videos in an order from high to low according to the relevance scores;
the first obtaining sub-module is used for obtaining the relevance scores of the candidate advertisements ranked as the Kth according to the candidate advertisements ranked by each video, and recording the relevance scores as the advertisement candidate scores of the target video, wherein K is a positive integer and is a preset threshold value.
8. The apparatus of claim 7, wherein the second selecting module comprises:
the second sequencing submodule is used for sequencing a plurality of videos needing to recommend the advertisements according to the sequence of the candidate advertisement scores from low to high;
the first selection submodule is used for selecting two target videos which are arranged at a first preset position and a second preset position from the sequenced videos;
a second obtaining sub-module for obtaining two target advertisement candidate scores S1 and S2 of the two target videos, wherein the target advertisement candidate score S1 is smaller than the target advertisement candidate score S2.
9. The apparatus of claim 8, wherein the recommendation module comprises:
the determining submodule is used for determining a first type of video with an advertisement candidate score being a first score S from a plurality of videos needing to recommend advertisements;
a second selection module, configured to select, from the sorted videos, a plurality of second-type videos arranged behind the first-type video;
the third acquisition module is used for respectively acquiring 1 st to K th ranked target candidate advertisements as advertisements recommended to the first type of videos and the second type of videos aiming at the candidate advertisements sequenced by the first type of videos and the second type of videos;
a third selecting module, configured to select, from the sorted videos, a plurality of third videos arranged before the first video;
the traversing module is used for traversing the relevance scores of the candidate advertisements according to the sequence of the relevance scores from high to low aiming at the candidate advertisements sequenced by the third type of videos, and stopping the traversal until the relevance scores are smaller than the first score S;
and the recommending submodule is used for taking the candidate advertisements with the relevance scores larger than or equal to the first score S obtained through traversal as the advertisements recommended to the third type of videos.
10. The apparatus of claim 6, wherein the first score, S, is calculated according to the following equation:
S＝(S2-S1)*x+S1。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710533587.6A CN107404657B (en) | 2017-07-03 | 2017-07-03 | Advertisement recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710533587.6A CN107404657B (en) | 2017-07-03 | 2017-07-03 | Advertisement recommendation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107404657A CN107404657A (en) | 2017-11-28 |
| CN107404657B true CN107404657B (en) | 2020-01-03 |
Family
ID=60404801
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710533587.6A Active CN107404657B (en) | 2017-07-03 | 2017-07-03 | Advertisement recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107404657B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110327628A (en) * | 2019-05-24 | 2019-10-15 | 北京金山安全软件有限公司 | Method and device for querying ranking and electronic equipment |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7647242B2 (en) * | 2003-09-30 | 2010-01-12 | Google, Inc. | Increasing a number of relevant advertisements using a relaxed match |
| US7594244B2 (en) * | 2003-11-12 | 2009-09-22 | Koninklijke Philips Electronics N.V. | Program recommendation system |
| US20080033810A1 (en) * | 2006-08-02 | 2008-02-07 | Yahoo! Inc. | System and method for forecasting the performance of advertisements using fuzzy systems |
| US20080270378A1 (en) * | 2007-04-24 | 2008-10-30 | Nokia Corporation | Method, Apparatus and Computer Program Product for Determining Relevance and/or Ambiguity in a Search System |
| CN101650731A (en) * | 2009-08-31 | 2010-02-17 | 浙江大学 | Method for generating suggested keywords of sponsored search advertisement based on user feedback |
| CN102572501A (en) * | 2010-12-23 | 2012-07-11 | 华东师范大学 | Video quality evaluation method and device capable of taking network performance and video self-owned characteristics into account |
| CN103209342B (en) * | 2013-04-01 | 2016-06-01 | 电子科技大学 | A kind of introduce video popularity and the collaborative filtered recommendation method of user's interests change |
| CN105608125B (en) * | 2015-12-15 | 2020-11-06 | 腾讯科技（深圳）有限公司 | Information processing method and server |
| CN106250442A (en) * | 2016-07-26 | 2016-12-21 | 新疆大学 | The feature selection approach of a kind of network security data and system |
| CN106570165B (en) * | 2016-11-07 | 2019-09-13 | 北京航空航天大学 | A kind of content based video retrieval system method and device |
| CN106844446A (en) * | 2016-12-16 | 2017-06-13 | 飞狐信息技术（天津）有限公司 | Video methods of marking, device and video system based on user's viewing behavior |
-
2017
- 2017-07-03 CN CN201710533587.6A patent/CN107404657B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN107404657A (en) | 2017-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10789634B2 (en) | Personalized recommendation method and system, and computer-readable record medium | |
| US8799306B2 (en) | Recommendation of search keywords based on indication of user intention | |
| CN106547767B (en) | Method and device for determining video cover picture | |
| US8903834B2 (en) | Recommending groups of items based on item ranks | |
| CN107832437B (en) | Audio/video pushing method, device, equipment and storage medium | |
| US10129596B2 (en) | Adaptive row selection | |
| US9594758B2 (en) | Method and apparatuses for enabling recommendations | |
| CN112364202B (en) | Video recommendation method and device and electronic equipment | |
| CN104053023B (en) | A kind of method and device of determining video similarity | |
| CN106649647B (en) | Search result ordering method and device based on artificial intelligence | |
| CN108028962A (en) | Video service condition information is handled to launch advertisement | |
| CN105589895B (en) | Resource ranking data generation method and device | |
| CN104484406B (en) | The dispensing analysis method and device of recommending data | |
| CN104239552B (en) | Generation association keyword, the method and system that association keyword is provided | |
| US11216518B2 (en) | Systems and methods of providing recommendations of content items | |
| CN106326297B (en) | Application program recommendation method and device | |
| KR20160113685A (en) | Method, apparatus and system for content recommendation | |
| CN105512156B (en) | Click model generation method and device | |
| CN111339416A (en) | Heat recall method and device, electronic equipment and storage medium | |
| CN106156351B (en) | Multimedia resource recommendation information generation method and device | |
| CN104967690A (en) | Information push method and device | |
| US11244014B2 (en) | System and method for enhancing exploration of data items | |
| CN106326388A (en) | Method and device for processing information | |
| WO2018148750A1 (en) | Method and system for re-aggregation and optimization of media | |
| CN109213933B (en) | Content item recommendation method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |