CN110929052B - Multimedia resource recommendation method and device, electronic equipment and storage medium - Google Patents
Multimedia resource recommendation method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110929052B CN110929052B CN201911218998.1A CN201911218998A CN110929052B CN 110929052 B CN110929052 B CN 110929052B CN 201911218998 A CN201911218998 A CN 201911218998A CN 110929052 B CN110929052 B CN 110929052B
- Authority
- CN
- China
- Prior art keywords
- tag
- value
- candidate
- label
- target user
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/40—Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
- G06F16/43—Querying
- G06F16/435—Filtering based on additional data, e.g. user or group profiles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The embodiment of the invention provides a multimedia resource recommendation method, a multimedia resource recommendation device, electronic equipment and a storage medium, and relates to the technical field of Internet, wherein the multimedia resource recommendation method comprises the following steps: obtaining an interest value of a target user on a target label; determining a label value of each candidate label according to the interest value and the correlation degree of each candidate label of the at least two candidate labels and the target label; and recommending multimedia resources to the target user according to the label value of each candidate label. According to the multimedia resource recommendation method provided by the embodiment of the invention, multimedia resources are divided into different labels, and then the relevance of the labels is utilized, so that the interest attitude of a user on a certain type of labels is rapidly extended to other labels; the method can effectively reduce the exploration cost of multimedia resource recommendation, recommend the multimedia resources which are possibly interested by the user, and has high recommendation accuracy.
Description
Technical Field
The present invention relates to the field of internet technologies, and in particular, to a multimedia resource recommendation method and apparatus, an electronic device, and a storage medium.
Background
In order to avoid that a user spends a large amount of time and energy to search a target multimedia resource from a large amount of multimedia resources, the multimedia resources are generally classified, the preference of the user is reflected by grading the category of the multimedia resources, and then the multimedia resources can be recommended according to the preference of the user.
In the prior art, scores of various multimedia resource categories are independent from one another, so that the multimedia resources under specific high-score multimedia resource categories are often recommended to users according to historical multimedia resource usage records, the content of the multimedia resources is single, other categories of multimedia resources which are potentially interesting to the users are difficult to explore, and the accuracy of multimedia resource recommendation is poor.
Disclosure of Invention
The embodiment of the invention aims to provide a multimedia resource recommendation method, a multimedia resource recommendation device, electronic equipment and a storage medium, so as to extend the interest attitude of a user on a certain type of tags to other tags, recommend multimedia resources corresponding to other tags which are likely to be interested to the user, and improve the accuracy of multimedia resource recommendation for the user. The specific technical scheme is as follows:
in a first aspect of the present invention, there is provided a multimedia resource recommendation method applied to an electronic device, including:
obtaining an interest value of a target user to a target tag, wherein the target tag is a tag of a multimedia resource operated by the target user;
respectively determining the tag value of each candidate tag according to the interest value and the correlation degree between each candidate tag of the at least two candidate tags and the target tag, wherein the candidate tags are tags of multimedia resources;
and recommending multimedia resources to the target user according to the label value of each candidate label.
In a second aspect of the present invention, there is also provided a multimedia resource recommendation apparatus applied to an electronic device, including:
the system comprises a first acquisition module, a first processing module and a second processing module, wherein the first acquisition module is used for acquiring an interest value of a target user on a target tag, and the target tag is a tag of a multimedia resource operated by the target user;
a first determining module, configured to determine a tag value of each candidate tag according to the interest value and a correlation degree between each candidate tag of the at least two candidate tags and the target tag, where the candidate tag is a tag of a multimedia resource;
and the recommending module is used for recommending the multimedia resource to the target user according to the label value of each candidate label.
In yet another aspect of the present invention, there is also provided a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to execute any one of the above-mentioned multimedia resource recommendation methods.
In yet another aspect of the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform any of the above described multimedia resource recommendation methods.
According to the multimedia resource recommendation method provided by the embodiment of the invention, multimedia resources are divided into different labels, and then the relevance of the labels is utilized, so that the interest attitude of a user on a certain type of labels is rapidly extended to other labels; the defects of single recommended content and low accuracy caused by independent scoring of various multimedia resources in the prior art are overcome; the method can effectively reduce the exploration cost of multimedia resource recommendation, recommend the multimedia resources which are possibly interested by the user, and has high recommendation accuracy.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below.
FIG. 1 is a flowchart of a method for recommending multimedia resources according to an embodiment of the present invention;
FIG. 2 is a flow chart of determining relevance between tags in an embodiment of the present invention;
FIG. 3 is a flowchart illustrating an embodiment of obtaining an interest value of a target user in a target tag;
FIG. 4 is a flow chart of determining a tag value for each candidate tag in an embodiment of the present invention;
FIG. 5 is a flowchart illustrating recommending multimedia resources to a target user according to a tag value of each candidate tag according to an embodiment of the present invention;
fig. 6 is a flowchart of updating candidate tags and giving weights to the candidate tags for a target user according to an embodiment of the present invention;
FIG. 7 is a flowchart illustrating the determination of candidate tag recommendation scores based on user recommendation requests in an exemplary application embodiment of the present invention;
FIG. 8 is a schematic structural diagram of a multimedia resource recommendation apparatus according to an embodiment of the present invention;
FIG. 9 is a schematic diagram of another multimedia resource recommendation apparatus according to an embodiment of the present invention;
FIG. 10 is a schematic structural diagram of a first obtaining module according to an embodiment of the present invention;
FIG. 11 is a block diagram illustrating a first determining module according to an embodiment of the present invention;
FIG. 12 is a block diagram of a recommendation module in accordance with an embodiment of the present invention;
fig. 13 is a schematic structural diagram of an electronic device in an embodiment of the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.
As shown in fig. 1, a multimedia resource recommendation method provided in an embodiment of the present invention includes:
step S100, obtaining an interest value of a target user on a target label, wherein the target label is a label of a multimedia resource operated by the target user;
step S200, respectively determining a label value of each candidate label according to the interest value and the correlation degree of each candidate label of at least two candidate labels and the target label, wherein the candidate labels are labels of multimedia resources;
and step S300, recommending multimedia resources to the target user according to the label value of each candidate label.
The multimedia resources described in the embodiments of the present invention may be video, music, articles, pictures, and other types of resources, and the following description mainly takes video type multimedia resources as an example.
In this embodiment, videos may be classified in advance, each category of video corresponds to one tag, and the classification basis may have various choices, such as generation of tags for animals, gourmet food, travel, and the like, based on understanding of video content. In other possible embodiments, the corresponding tag may also be generated according to the video duration, the video publisher type, and the like.
It will be appreciated that the interest value may then be understood as the degree to which the user is interested in a certain type of video. In some embodiments, the interest value of a target user in a target tag may be determined according to operation behavior information generated when the target user operates a certain video and the target tag corresponding to the operated video. The operation behavior of the target user on the video can be watching, fast forwarding, reviewing, praise, commenting, sharing and the like.
In combination with practical applications, different operation behaviors often represent the interest level of a target user in a certain video, for example: the target user only watches the first video and watches and approves the second video; or the target user has fast-forwarded when watching the first video and has reviewed when watching the second video. Generally, the operation behavior may indicate that the target user is interested in the second video more than the first video. The above interest level, i.e. the interest value, can be represented by a score (or rating), for example, the score (or rating) of different operation behaviors can be obtained through a preset operation behavior-score (or rating) correspondence table, and a specific score determination manner is further described below.
The correlation degree between the labels can be obtained according to historical behavior data of a plurality of users, and can be represented by a specific numerical value. For example, as for the statistical finding of historical behavior data, if a large number of users like videos with travel tags and also like videos with food tags, the correlation between the travel tags and the food tags can be determined to be a higher value; for the same two tags, the correlation degree can be set to 1; the calculation of the correlation between the tags will be described in detail below. In addition, the candidate tag, that is, the tag having a correlation with the target tag, may include the target tag because the same tag may have a correlation therebetween.
The tag value is a basis for recommending videos to target users, and in practical application, videos can be recommended directly according to sorting or screening of the tag value, or the tag value can be used as one of factors influencing video recommendation results. In general, the higher the tag value, the more likely it is that its corresponding video is recommended. The label value may be represented by a numerical label score, or may be represented by a rank, priority, or the like.
In this embodiment, when the interest value of the target user for the target tag is obtained, not only the tag value of the target tag but also the tag value of a candidate tag different from the target tag may be adjusted, and the degree of adjustment of the tag value of the candidate tag is related to the correlation between the tags.
If the target user is interested in a certain category of videos, the probability that the target user is interested in videos of similar categories is higher, and the same is true. The relevance among different categories is calculated according to the existing historical data, so that the attitudes of the target user on the videos of other categories can be rapidly acquired according to the feedback of the target user on the videos of a certain category (the feedback can be embodied as the interest value of the target user on the tags of the videos of the category), unnecessary exploration is reduced in subsequent video recommendation, and recommendation which is more likely to be interested by the target user is provided.
Watching and agreeing to a tag with a target user 1 Video of (2) as an example; wherein, the label tag 1 Namely the target label, the candidate label comprises the label tag with higher correlation degree 1 Tag with tag 2 . The watching and praise behaviors indicate that the target user targets the tag 1 There is a high interest, and accordingly, these actions will generate a positive interest value for the tag 1 The tag value of (c) is raised. Due to the tag 2 Tag with tag 1 The relevance of (2) is higher, and the target user is at tag 2 The probability of interest is high, so the tag is marked according to the correlation degree between the two tags 2 The tag value of (a) is correspondingly promoted, and then the promotion will have a tag 2 The likelihood of video recommendation to the target user.
It is to be noted that the multimedia resource recommendation method provided in the embodiment of the present invention can be applied not only to video-type multimedia resource recommendation but also to music, articles, pictures, and other types of multimedia resource recommendation. For example, for music type multimedia resources, ancient style, popular style, classical style and other labels can be established according to music content, the correlation degree among the labels is established, and then the label values of other labels are determined according to the interest value of a target user in the target label.
According to the embodiment of the invention, multimedia resources are divided into different tags, and then the interest attitude of a user on a certain type of tags is rapidly extended to other tags by utilizing the relevance of the tags; the defects of single recommended content, low flexibility and accuracy caused by independent scoring of various multimedia resources in the prior art are overcome; the method can effectively reduce the exploration cost of multimedia resource recommendation, recommend the multimedia resources which are possibly interested by the user, and has higher recommendation flexibility and accuracy.
As shown in fig. 2, in this embodiment, before the step S100 obtains the interest value of the target user in the target tag, the method further includes:
step S410, obtaining operation behavior information of N multimedia resources, wherein N is an integer greater than 1;
step S420, determining a correlation degree of every two tags in the tags corresponding to the N multimedia resources according to the operation behavior information of each multimedia resource in the N multimedia resources.
In this embodiment, the obtained operation behavior information may be performed for a plurality of users, and the finally determined degree of correlation between each two tags mainly refers to the degree of correlation between each two different tags, and of course, the degree of correlation between the same two tags may be directly determined to be 1.
In practical applications, mapping rules may be designed for different play click behaviors according to all play click data of users on videos, and the mapping rules represent the degrees that the users like or dislike videos, for example, a more forward click play behavior may be represented by a higher score, and the score is controlled within a certain range, such as 0 to 5. The relationship between the labels and the degree is fitted by an FM (Factor Machine) algorithm, and the similarity between the labels is calculated by using the obtained embedding (transliteration is embedding, which is a mapping in nature) space, and the calculation method can be cosine similarity or euclidean distance.
For example, after a user clicks and watches a video with a certain label, a multidimensional vector is generated according to the characteristics of the user (such as gender, age group, watching place, and the like) and the characteristics of the label, and a score is given to the watching behavior of the user; since there are multiple users, and the same user will also generate different behaviors for different videos, multiple such multidimensional vectors and scores will be generated; and sending the multi-dimensional vectors and the scores into an FM algorithm to perform matrix operation to obtain a multi-dimensional matrix. The embedding algorithm is essentially to reduce the dimension of the multi-dimensional matrix, embed the features related to the tags in the multi-dimensional matrix into a lower-dimensional matrix (i.e., the above-mentioned embedding space), and then use cosine similarity or euclidean distance to calculate the similarity between the tags.
Of course, in some other possible embodiments, the FM algorithm may be replaced by an FFM (Field-aware Decomposition Machine) algorithm, and the embedding algorithm may be replaced by a PCA (Principal Component Analysis), an SVD (Singular Value Decomposition), and the like.
Optionally, the correlation degree of every two tags is the same and fixed for all users in a certain period. When a certain user acts on a video with a certain label, the label scores of the user on the candidate labels corresponding to the label are updated, the relevance of every two labels cannot be changed, and the label scores of the candidate labels corresponding to the label by other users cannot be influenced. Optionally, since the data of the user clicking the video may change continuously, the correlation between the two tags may be updated after a set time interval.
According to the embodiment, the corresponding label relevancy relation does not need to be established for each user, and the label relevancy calculation cost is reduced. In subsequent application, the label value of the target user to the candidate label is updated according to the relevance of the behavior of the target user to the video and the label, and further video recommendation which is more likely to be interested by the target user can be accurately provided.
The above describes the manner of determining the degree of tag correlation. In practical applications, a situation where a target user first uses an Application (APP) for providing a video resource is faced. At this time, for the target user, there is no history behavior record, and there is no information about the interest value of the user in a certain tag, and the recommended content of the first screen can be obtained based on whether there are two schemes for image:
the method comprises the steps of (1) recommending a first screen video according to an existing portrait; available portraits can be age, gender, region, income, occupation, installation application and the like, and a simple personalized ranking model can be trained based on the portraits;
recommending according to video heat in the pool without an image; the video exposure times, the play completion number, the praise number and the share number can be used for calculating the video popularity.
After the first screen recommended content is obtained, the target user can operate on a video with a certain target tag in the recommended content, and further the interest value of the target user on the target tag can be obtained according to the operation behavior.
In some possible embodiments, if there is a video with multiple tags, one main tag may be defined for each video, and when a target user operates one video, the interest value of the target user in the main tag of the video is obtained.
In other feasible embodiments, when the target user uses the APP for the first time, a plurality of preset tags can be displayed in the APP display interface for the target user to select, and the tag selected by the target user is used as the target tag and is given a preset interest value; or displaying the preset labels and the corresponding scoring options to the target user, taking the labels scored by the target user as the target labels, and directly taking the scores as interest values or converting the scores according to preset rules to obtain the interest values.
Due to the lack of sufficient historical play records, when video recommendation is performed on a new user (which can be understood as a target user who uses the APP for the first time or uses the APP for a short time), the APP needs to adopt a cold start recommendation strategy. In the embodiment of the invention, when the first screen is recommended, the portrait of a new user or video recommendation can be performed according to video heat in a pool; after the operation behavior of the new user on the video is obtained, the interest value of the new user on the corresponding label of the video can be further obtained, and the interest attitude (which can be embodied by the label value) of the user on other types of labels is obtained by combining the interest value and the correlation degree of each candidate label of the at least two candidate labels and the target label, so that the interest attitude of the new user on a certain type of labels is rapidly extended to other labels. The method can effectively reduce the exploration cost recommended by the new user, recommend videos which are more likely to be interested by the new user, and is more efficient in the interest exploration of the new user. After enough historical playing records are acquired at a later stage, the multimedia resource recommendation method can be adopted to recommend more types of videos to the target user. Of course, the multimedia resource recommendation method can also be used for recommending multimedia resources of types such as music, articles, pictures and the like, and can obtain the same or similar technical effects as the video recommendation scheme.
Optionally, as shown in fig. 3, the step S100 of obtaining the interest value of the target user in the target tag includes:
step S110, acquiring the operation behavior information of the target user on the multimedia resource;
step S120, acquiring a behavior feedback value of the operation behavior information of the target user on the multimedia resource according to a corresponding relation between preset operation behavior information and the behavior feedback value;
step S130, determining the behavior feedback value as an interest value of the target user in the target tag.
It can be understood that, in this embodiment, the target tag is a tag of a multimedia resource operated by the target user, and the operation behavior information is obtained for the multimedia resource operated by the target user, so that a behavior feedback value obtained according to the operation behavior information may be determined as an interest value of the target user in the target tag.
In the process of using the video resources, a user can have operation behaviors of watching, fast forwarding, reviewing, paying attention to, commenting, sharing, clicking dislikes, clicking reduction recommendation and the like, corresponding operation behavior information is generated, and the interest value can be determined according to the operation behavior information.
The operation behavior of the user can reflect the interest attitude of the user in the video, such as whether the user is interested, the interest degree and the like. The operation behaviors of the user can be divided into a positive operation behavior and a negative operation behavior: forward operation behaviors, such as watching, paying attention to, praise and the like, correspond to forward interest values, so that the tag values of tags corresponding to the videos can be promoted or maintained; negative operation behaviors, such as dislike clicking and recommendation reduction clicking, correspond to negative interest values, so that the tag values of the tags corresponding to the video can be reduced. Of course, alternatively, the influence of the positive operation behavior and the negative operation behavior on the tag value may also be reflected in the degree of increase of the tag value, for example, the positive operation behavior causes the degree of increase of the tag value to be higher than a threshold, and the negative operation behavior causes the degree of increase of the tag value to be lower than the threshold.
Optionally, for the forward operation behavior, different operation behaviors may also correspond to different interest values, for example, an interest value corresponding to a praise behavior may be higher than an interest value corresponding to a view behavior.
In the embodiment, the interest value of the target user to the target tag can be dynamically acquired according to the operation behavior information of the target user to the multimedia resource, so that the tag value of the candidate tag can be updated according to the use condition of the target user to the multimedia resource, and the accuracy of recommending the multimedia resource is improved.
The behavior feedback value can be represented by a digitized behavior feedback component, and the application of the embodiment in an actual application scenario is exemplified below. For the forward operation behaviors, the corresponding behavior feedback score is between 1 and 2; for negative-going operation behavior, the corresponding behavior feedback score is between 0-1. Because the operation behavior information and the specific operation behavior of the user are corresponding to each other, the correspondence between the preset operation behavior information and the behavior feedback value can also be represented by the correspondence between the operation behavior and the behavior feedback value, and the specific relationship is shown in table 1:
TABLE 1
| Operational behavior | Behavioral feedback score |
| Like a Chinese character' Zhi | 2 |
| Attention (front) | 1.8 |
| Share (positive) | 1.7 |
| Review comment (Zheng) | 1.5 |
| Playing time (positive) | Score(tm,length) |
| Quickly draw across (negative) | 0.9 |
| Dislike (negative) | 0.5 |
| Reduction recommendation (negative) | 0.5 |
In the Score (tm, length), length represents the total duration of the video, the unit is second(s), tm represents the playing duration of the current behavior of the user, the unit is second, the ratio of tm to length is represented by rate, and the Score (tm, length) represents a behavior feedback Score calculated according to the proportion of the playing duration of the user occupying the total duration of the video. Wherein a rate greater than 1 indicates that replay behavior has occurred.
Score (tm, length) is defined as shown in table 2:
TABLE 2
The embodiment provides a concept of behavior feedback score, designs a set of effective mapping rules, and maps abstract user operation behaviors into specific scores. Different behavior feedback scores are given to different user operation behaviors, so that the interest degree (interest value) of the target user on the operated multimedia resource is determined by distinguishing different operation behaviors of the target user, and the tag value of the target tag is reasonably updated.
In some possible implementations, the behavior feedback value may also be represented by a ranked behavior rating, such as a rating of "high", "medium", "low", or a rating of "primary", "secondary", "tertiary", etc., which extends the behavior rating to other candidate tags according to the degree of correlation between the tags.
Optionally, as shown in fig. 4, in step S200, determining a tag value of each candidate tag according to the interest value and a correlation between each candidate tag of the at least two candidate tags and the target tag, respectively, includes:
step S210, obtaining the historical label value S of the ith candidate label i And the correlation Rel of the ith candidate label and the target label ij ；
Step S220, updating the label value of the ith candidate label to S i ×(1+(S action -1)×Rel ij )；
Wherein S is action The historical tag value S is the behavior feedback value i Before updating the label value, the label value of the ith candidate label.
For the target user, each candidate tag may have an initial tag value(i.e., the above-mentioned historical tag value S) i ) The initial tag value may be configured uniformly, or may have been changed on the basis of a uniform configuration value according to the historical data corresponding to the target user. Therefore, the determination process of the tag value of each candidate tag may be a process of updating the tag value of the candidate tag for the target user.
In a possible embodiment, the correlation between the tags can be represented by a correlation matrix as shown in table 3, wherein the correlation between any two tags can be set between 0-1:
TABLE 3
| tag 1 | tag 2 | tag 3 | … | |
| tag 1 | 1 | Rel 12 | Rel 13 | … |
| tag 2 | Rel 21 | 1 | Rel 31 | … |
| tag 3 | Rel 31 | Rel 32 | 1 | … |
| … | … | … | … | … |
Wherein tag is defined i Indicates the ith candidate tag (e.g., tag) 1 Representing the 1 st candidate tag, and so on), hereinafter referred to as tag i ，Rel ij Presentation tag i Tag with tag j Correlation between (e.g. Rel) 12 Presentation tag 1 Tag with tag 2 And so on), i and j are both positive integers less than or equal to the number of candidate tags. When i = j, rel ij ＝1。
It is easy to understand that the target user pairs tag i The larger the tag value of (a), the higher the target user's liking for the tag.
In an initial state, the label values of all candidate labels of a target user are set to be 1, and when the target user acts on a certain video, the label values of the candidate notes of the target user are updated in real time. Specifically, after the user operation behavior occurs, the operation behavior information of the target user may be obtained by using a pingback (which may be translated into the on-site delivery data for obtaining the data such as display, click, etc., and may be obtained in real time or according to a history log), and the behavior corresponding to the operation behavior information may be obtained according to the corresponding relationship between the preset operation behavior information and the behavior feedback valueFeedback score S action (ii) a Then, the target tag of the video is taken out j Correlation Rel with all candidate tags ij (ii) a Finally, tag is matched with the target user i The tag value of (2) is updated, and the updating formula is as follows:
S i ←S i ×(1+(S action -1)×Rel ij )；
wherein, the step of going to step is to assign a value to the target user tag after the user operation behavior occurs i The tag value of (a) is updated. Due to S i The video recommendation method can be used for transmitting the favorite of the target user to all the related video tags, is favorable for recommending videos with the related video tags to the user subsequently, and provides recommendation flexibility.
As shown in fig. 5, in this embodiment, the step S300 of recommending a multimedia resource to the target user according to the tag value of each candidate tag includes:
step S310, acquiring a use time sequence of the candidate tag by the target user;
step S320, according to the use time sequence, a weight is distributed to each candidate label, wherein the weight of the candidate label with the use time before is larger than or equal to the weight of the candidate label with the use time after;
step S330, determining a recommended value of each candidate label according to the label value of each candidate label and the weight of each candidate label;
step S340, recommending multimedia resources to the target user according to the recommended value of each candidate tag.
In this embodiment, the tag list experienced by each user may be recorded according to the occurrence sequence of the user operation behavior, and the weights may be assigned to the tags in sequence. For example, a list of videos watched by a target user in a certain period of time is recorded, tags are sorted according to the watching sequence of the videos and tags of the videos, and weights are respectively assigned to the tags according to the sequence. Of course, when the user operation behavior occurs, the labels experienced by the target user can also be directly recorded, and the weight is distributed according to the ordering of the labels in the record.
For example, the current target user pairs tag i Is divided into S i Obtaining tag by sequentially distributing weights i Has a weight of w i Then tag i Recommended value of
The calculation can be made according to the following formula:
for the final recommended value
And sequencing all the candidate tags to obtain a plurality of candidate tags with the highest scores, independently sequencing videos under all the candidate tags according to a set rule, respectively taking videos ranked in front, and outputting the videos according to the sequence of the candidate tags to obtain a final recall list so as to complete recommendation of the videos to a target user. The set rules may be sorted independently, such as by video popularity, by user portrait, by video provider priority, and so forth.
In this embodiment, the use sequence of the candidate tag by the target user is used as the basis for recommending the multimedia resource, so that the flexibility of recommending the multimedia resource can be further improved. For example, the corresponding multimedia resources may be recommended according to the tag preference of the target user in a certain period, or the multimedia resources corresponding to the tags that the target user has recently experienced may be arranged backwards, so as to increase the diversity of the recommendation.
In addition, in the embodiment, when the weight is assigned to each candidate tag, the weight of the candidate tag with the previous use time is greater than or equal to the weight of the candidate tag with the later use time.
For example, the assignment of weights for labels used by the target user is shown in table 4:
TABLE 4
| Label (R) | tag -1 | tag -2 | tag -3 | tag -4 | tag -5 | tag -6 |
| Weight of | 0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 0.95 |
tag -n N-th tag, w, indicating the most recent usage by the target user n Denoted as the weight corresponding to the most recently used nth tag. The weight of the label newly used by the target user is the minimum, the longer the label is, the closer the weight is to 1, and the weight of the label which is not used or the label of which the arrangement serial number exceeds the set threshold value is 1. In practical application, there may be different tags -n Each corresponding to a different candidate label, or a plurality of candidate labelsSame tag -n Corresponding to the same candidate tag, and in the latter case, the minimum value of the multiple weights corresponding to the candidate tag may be taken when calculating the recommendation value.
In this embodiment, even if the tag value of a certain candidate tag is high, if the target user has just recently viewed the video corresponding to the candidate tag, the weight of the candidate tag will be lower, and the final recommendation value of the candidate tag will also be reduced, so that the diversity of recommendations is increased, and the video of the candidate tag that has not been viewed is encouraged to be introduced.
As shown in fig. 6 and 7, the steps of the method for recommending multimedia resources according to the embodiment of the present invention for implementing a specific application of video recommendation are as follows:
step S10, obtaining operation behavior information of a target user on a first video through a pingback, wherein the first video is a video operated by the target user;
s11, acquiring corresponding behavior feedback scores according to the operation behavior information;
step S12, acquiring a target label corresponding to the first video;
s13, acquiring the correlation degree of the target label and the candidate label according to the label correlation degree matrix;
step S14, updating the label scores of the candidate labels aiming at the target user according to the behavior feedback scores and the correlation degrees;
step S15, according to the using time sequence of the target user to the candidate label, giving weight to the candidate label aiming at the target user;
step S16, acquiring a recommendation request of a target user;
s17, acquiring the final recommendation score of each candidate label according to the label score and the weight of the candidate label;
and S18, recalling the video according to the final recommendation score.
According to the specific application implementation mode, after the operation behavior information of the target user on the video corresponding to the target label is obtained, the label scores of the target label are updated, the label scores of a plurality of candidate labels are updated through the label similarity, the interest attitude of the target user on the target label can be extended to the candidate labels, unnecessary interest direction exploration can be reduced in subsequent recommendation, and video recommendation which is more likely to be interested by the target user is provided; meanwhile, the use time sequence of the candidate tags is taken as a consideration factor of video recommendation, so that the accuracy and the flexibility of video recommendation are improved.
As shown in fig. 8, an embodiment of the present invention further provides a multimedia resource recommendation apparatus, including:
a first obtaining module 510, configured to obtain an interest value of a target user for a target tag, where the target tag is a tag of a multimedia resource operated by the target user;
a first determining module 520, configured to determine a tag value of each candidate tag according to the interest value and a correlation degree between each candidate tag of the at least two candidate tags and the target tag, where the candidate tag is a tag of a multimedia resource;
a recommending module 530, configured to recommend a multimedia resource to the target user according to the tag value of each candidate tag.
Optionally, as shown in fig. 9, the apparatus further includes:
a second obtaining module 540, configured to obtain operation behavior information on N multimedia resources, where N is an integer greater than 1;
the second determining module 550 is configured to determine, according to the operation behavior information of each multimedia resource in the N multimedia resources, a correlation degree between every two tags in the tags corresponding to the N multimedia resources.
Optionally, as shown in fig. 10, the first obtaining module 510 includes:
a first obtaining unit 511, configured to obtain operation behavior information of a target user on a multimedia resource;
a second obtaining unit 512, configured to obtain a behavior feedback value of the operation behavior information of the target user on the multimedia resource according to a corresponding relationship between preset operation behavior information and the behavior feedback value;
a first determining unit 513, configured to determine the behavior feedback value as an interest value of the target user in a target tag.
Optionally, as shown in fig. 11, the first determining module 520 includes:
a third obtaining unit 521 for obtaining the historical tag value S of the ith candidate tag i And the correlation Rel of the ith candidate label and the target label ij ；
An updating unit 522 for updating the label value of the ith candidate label to S i ×(1+(S action -1)×Rel ij )；
Wherein S is action The historical tag value S is the behavior feedback value i Before updating the label value, the label value of the ith candidate label.
Optionally, as shown in fig. 12, the recommending module 530 includes:
a fourth obtaining unit 531, configured to obtain a use time sequence of the candidate tag by the target user;
an assigning unit 532, configured to assign a weight to each candidate tag according to the usage time sequence, where the weight of a candidate tag with a usage time before is greater than or equal to the weight of a candidate tag with a usage time after;
a second determining unit 533, configured to determine a recommended value of each candidate tag according to a product of the tag value of each candidate tag and the weight of each candidate tag;
and the recommending unit 534 is configured to recommend the multimedia resource to the target user according to the recommended value of each candidate tag.
It should be noted that the multimedia resource recommendation apparatus provided in the embodiments of the present invention is an apparatus capable of executing the multimedia resource recommendation method, and therefore all embodiments of the multimedia resource recommendation method are applicable to the multimedia resource recommendation apparatus and can achieve the same or similar beneficial effects.
An embodiment of the present invention further provides an electronic device, as shown in fig. 13, including a processor 601, a communication interface 602, a memory 603, and a communication bus 604, where the processor 601, the communication interface 602, and the memory 603 complete mutual communication through the communication bus 604;
a memory 603 for storing a computer program;
the processor 601 is configured to implement the following steps when executing the program stored in the memory 603:
obtaining an interest value of a target user on a target label, wherein the target label is a label of a multimedia resource operated by the target user;
respectively determining the tag value of each candidate tag according to the interest value and the correlation degree between each candidate tag of the at least two candidate tags and the target tag, wherein the candidate tags are tags of multimedia resources;
and recommending multimedia resources to the target user according to the label value of each candidate label.
Optionally, before obtaining the interest value of the target user on the target tag, the following steps may also be implemented:
acquiring operation behavior information of N multimedia resources, wherein N is an integer greater than 1;
and determining the correlation degree of every two tags in the tags corresponding to the N multimedia resources according to the operation behavior information of each multimedia resource in the N multimedia resources.
Optionally, the obtaining of the interest value of the target user in the target tag includes:
acquiring the operation behavior information of a target user on the multimedia resource;
acquiring a behavior feedback value of the operation behavior information of the target user on the multimedia resource according to a corresponding relation between preset operation behavior information and the behavior feedback value;
and determining the behavior feedback value as the interest value of the target user in the target tag.
Optionally, the determining, according to the interest value and the correlation between each candidate tag of the at least two candidate tags and the target tag, the tag value of each candidate tag includes:
obtaining the historical label value S of the ith candidate label i And the correlation Rel of the ith candidate label and the target label ij ；
Updating the label value of the ith candidate label to S i ×(1+(S action -1)×Rel ij )；
Wherein S is action The historical tag value S is the behavior feedback value i Before updating the label value, the label value of the ith candidate label.
Optionally, the recommending multimedia resources to the target user according to the tag value of each candidate tag includes:
acquiring a use time sequence of a target user on the candidate label;
according to the use time sequence, each candidate label is assigned with a weight, wherein the weight of the candidate label with the use time before is larger than or equal to the weight of the candidate label with the use time after;
respectively determining the recommended value of each candidate label according to the product of the label value of each candidate label and the weight of each candidate label;
and recommending the multimedia resources to the target user according to the recommended value of each candidate tag.
The communication bus mentioned in the above terminal may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this is not intended to represent only one bus or type of bus.
The communication interface is used for communication between the terminal and other equipment.
The Memory may include a Random Access Memory (RAM), and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
In another embodiment of the present invention, there is also provided a computer-readable storage medium, which stores instructions that, when executed on a computer, cause the computer to execute the multimedia resource recommendation method described in any one of the above embodiments.
In yet another embodiment of the present invention, there is also provided a computer program product containing instructions which, when run on a computer, cause the computer to perform the method for multimedia resource recommendation described in any of the above embodiments.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid State Disk (SSD)), among others.
It is 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 apparatus 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 apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.
Claims (8)
1. A multimedia resource recommendation method is applied to electronic equipment, and is characterized by comprising the following steps:
obtaining an interest value of a target user to a target tag, wherein the target tag is a tag of a multimedia resource operated by the target user;
respectively determining the tag value of each candidate tag according to the interest value and the correlation degree between each candidate tag of the at least two candidate tags and the target tag, wherein the candidate tags are tags of multimedia resources;
recommending multimedia resources to the target user according to the label value of each candidate label;
wherein, the recommending multimedia resources to the target user according to the label value of each candidate label comprises:
acquiring a use time sequence of a target user for the candidate tag;
according to the use time sequence, each candidate label is assigned with a weight, wherein the weight of the candidate label with the use time before is larger than or equal to the weight of the candidate label with the use time after;
respectively determining the recommended value of each candidate label according to the product of the label value of each candidate label and the weight of each candidate label;
recommending multimedia resources to the target user according to the recommended value of each candidate tag;
the obtaining of the interest value of the target user in the target tag includes:
acquiring the operation behavior information of a target user on the multimedia resource;
acquiring a behavior feedback value of the operation behavior information of the target user on the multimedia resource according to a corresponding relation between preset operation behavior information and the behavior feedback value;
determining the behavior feedback value as an interest value of the target user in a target tag;
the operation behavior information comprises positive operation information corresponding to the positive operation behavior and negative operation information corresponding to the negative operation behavior;
the interest value determined based on the positive operation information is used for promoting or keeping the label value of the candidate label, and the interest value determined based on the negative operation information is used for reducing the label value of the candidate label;
alternatively, the first and second electrodes may be,
the interest value determined based on the positive operation information is used to boost the tag value of the candidate tag by a degree above a threshold, while the interest value determined based on the negative operation information is also used to boost the tag value of the candidate tag by a degree below the threshold.
2. The method of claim 1, wherein before obtaining the interest value of the target user in the target tag, the method further comprises:
acquiring operation behavior information of N multimedia resources, wherein N is an integer greater than 1;
and determining the correlation degree of every two tags in the tags corresponding to the N multimedia resources according to the operation behavior information of each multimedia resource in the N multimedia resources.
3. The method of claim 1, wherein the determining the tag value of each of the at least two candidate tags according to the interest value and the correlation between the target tag and each of the candidate tags comprises:
obtaining the historical label value S of the ith candidate label i And the correlation Rel of the ith candidate label and the target label ij ；
Updating the label value of the ith candidate label to S i ×(1+(S action -1)×Rel ij )；
Wherein S is action The historical tag value S is the behavior feedback value i Before updating the label value, the label value of the ith candidate label.
4. A multimedia resource recommendation device applied to electronic equipment is characterized by comprising:
the system comprises a first acquisition module, a first processing module and a second processing module, wherein the first acquisition module is used for acquiring an interest value of a target user on a target tag, and the target tag is a tag of a multimedia resource operated by the target user;
a first determining module, configured to determine a tag value of each candidate tag according to the interest value and a correlation between each candidate tag of the at least two candidate tags and the target tag, where the candidate tag is a tag of a multimedia resource;
the recommending module is used for recommending multimedia resources to the target user according to the label value of each candidate label;
wherein, the recommendation module comprises:
the fourth acquisition unit is used for acquiring the use time sequence of the candidate tag by the target user;
the distribution unit is used for distributing a weight to each candidate label according to the use time sequence, wherein the weight of the candidate label with the use time before is more than or equal to the weight of the candidate label with the use time after;
a second determining unit, configured to determine a recommended value of each candidate tag according to a product of the tag value of each candidate tag and the weight of each candidate tag;
the recommending unit is used for recommending multimedia resources to the target user according to the recommended value of each candidate tag;
wherein, the first obtaining module comprises:
the first acquisition unit is used for acquiring the operation behavior information of a target user on the multimedia resource;
the second obtaining unit is used for obtaining the behavior feedback value of the operation behavior information of the target user on the multimedia resource according to the corresponding relation between the preset operation behavior information and the behavior feedback value;
a first determining unit, configured to determine the behavior feedback value as an interest value of the target user in a target tag;
the operation behavior information comprises positive operation information corresponding to the positive operation behavior and negative operation information corresponding to the negative operation behavior;
the interest value determined based on the positive operation information is used for promoting or keeping the label value of the candidate label, and the interest value determined based on the negative operation information is used for reducing the label value of the candidate label;
alternatively, the first and second electrodes may be,
the interest value determined based on the positive operation information is used to boost the tag value of the candidate tag by a degree above a threshold, while the interest value determined based on the negative operation information is also used to boost the tag value of the candidate tag by a degree below the threshold.
5. The apparatus of claim 4, further comprising:
the second acquisition module is used for acquiring the operation behavior information of N multimedia resources, wherein N is an integer greater than 1;
and the second determining module is used for determining the correlation degree of every two tags in the tags corresponding to the N multimedia resources according to the operation behavior information of each multimedia resource in the N multimedia resources.
6. The apparatus of claim 4, wherein the first determining module comprises:
a third obtaining unit for obtaining the historical label value S of the ith candidate label i And the correlation Rel of the ith candidate label and the target label ij ；
An updating unit for updating the label value of the ith candidate label to S i ×(1+(S action -1)×Rel ij )；
Wherein S is action The historical tag value S is the behavior feedback value i Before updating the label value, the label value of the ith candidate label.
7. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any one of claims 1 to 3 when executing a program stored in the memory.
8. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911218998.1A CN110929052B (en) | 2019-12-03 | 2019-12-03 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911218998.1A CN110929052B (en) | 2019-12-03 | 2019-12-03 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110929052A CN110929052A (en) | 2020-03-27 |
| CN110929052B true CN110929052B (en) | 2023-04-18 |
Family
ID=69847267
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911218998.1A Active CN110929052B (en) | 2019-12-03 | 2019-12-03 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110929052B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113535991B (en) * | 2020-04-14 | 2024-01-26 | 北京达佳互联信息技术有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
| CN113569126A (en) * | 2020-04-29 | 2021-10-29 | 北京达佳互联信息技术有限公司 | Multimedia file recommendation method and device and server |
| CN111651669A (en) * | 2020-05-20 | 2020-09-11 | 拉扎斯网络科技（上海）有限公司 | Information recommendation method and device, electronic equipment and computer-readable storage medium |
| CN111639199A (en) * | 2020-06-04 | 2020-09-08 | 北京声智科技有限公司 | Multimedia file recommendation method, device, server and storage medium |
| CN111723289B (en) * | 2020-06-08 | 2024-02-02 | 北京声智科技有限公司 | Information recommendation method and device |
| CN111708901B (en) * | 2020-06-19 | 2023-10-13 | 腾讯科技（深圳）有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
| CN111881666B (en) * | 2020-07-09 | 2021-08-17 | 腾讯科技（深圳）有限公司 | Information processing method, device, equipment and storage medium |
| CN112000841B (en) * | 2020-07-29 | 2023-09-26 | 北京达佳互联信息技术有限公司 | Electronic tag data processing method and device, electronic equipment and storage medium |
| CN112333210B (en) * | 2021-01-04 | 2022-03-29 | 视联动力信息技术股份有限公司 | Method and equipment for realizing data communication function of video network |
| CN113157951A (en) * | 2021-03-26 | 2021-07-23 | 北京达佳互联信息技术有限公司 | Multimedia resource processing method, device, server and storage medium |
| CN113672746B (en) * | 2021-07-27 | 2024-03-26 | 北京达佳互联信息技术有限公司 | Multimedia resource recommendation method and device, electronic equipment and storage medium |
| CN113836415A (en) * | 2021-09-24 | 2021-12-24 | 北京字节跳动网络技术有限公司 | Information recommendation method, device, medium and equipment |
| CN114501151B (en) * | 2022-01-24 | 2024-02-23 | 青岛聚看云科技有限公司 | Display equipment and media asset recommendation method |
| CN114302242B (en) * | 2022-01-25 | 2023-10-31 | 聚好看科技股份有限公司 | Media asset recommendation method, display equipment and server |
| CN115659027B (en) * | 2022-10-28 | 2023-06-20 | 广州彩蛋文化传媒有限公司 | Recommendation method and system based on short video data tag and cloud platform |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102541921A (en) * | 2010-12-24 | 2012-07-04 | 华东师范大学 | Control method and device for recommending resources through tag extension |
| CN104537115A (en) * | 2015-01-21 | 2015-04-22 | 北京字节跳动科技有限公司 | Method and device for exploring user interests |
| CN106055617A (en) * | 2016-05-26 | 2016-10-26 | 乐视控股（北京）有限公司 | Data pushing method and device |
| WO2017101317A1 (en) * | 2015-12-14 | 2017-06-22 | 乐视控股（北京）有限公司 | Method and apparatus for displaying intelligent recommendations on different terminals |
| CN107341245A (en) * | 2017-07-06 | 2017-11-10 | 广州优视网络科技有限公司 | Data processing method, device and server |
| CN109697629A (en) * | 2018-11-15 | 2019-04-30 | 平安科技（深圳）有限公司 | Product data method for pushing and device, storage medium, computer equipment |
-
2019
- 2019-12-03 CN CN201911218998.1A patent/CN110929052B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102541921A (en) * | 2010-12-24 | 2012-07-04 | 华东师范大学 | Control method and device for recommending resources through tag extension |
| CN104537115A (en) * | 2015-01-21 | 2015-04-22 | 北京字节跳动科技有限公司 | Method and device for exploring user interests |
| WO2017101317A1 (en) * | 2015-12-14 | 2017-06-22 | 乐视控股（北京）有限公司 | Method and apparatus for displaying intelligent recommendations on different terminals |
| CN106055617A (en) * | 2016-05-26 | 2016-10-26 | 乐视控股（北京）有限公司 | Data pushing method and device |
| CN107341245A (en) * | 2017-07-06 | 2017-11-10 | 广州优视网络科技有限公司 | Data processing method, device and server |
| CN109697629A (en) * | 2018-11-15 | 2019-04-30 | 平安科技（深圳）有限公司 | Product data method for pushing and device, storage medium, computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110929052A (en) | 2020-03-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110929052B (en) | Multimedia resource recommendation method and device, electronic equipment and storage medium | |
| CN109885770B (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CA2710238C (en) | Video quality measures | |
| US8468083B1 (en) | Advertisement topic diversification and ranking | |
| RU2725659C2 (en) | Method and system for evaluating data on user-element interactions | |
| US8893012B1 (en) | Visual indicator based on relative rating of content item | |
| WO2018121700A1 (en) | Method and device for recommending application information based on installed application, terminal device, and storage medium | |
| CN106708817B (en) | Information searching method and device | |
| EP1505521A2 (en) | Setting user preferences for an electronic program guide | |
| CN111767429B (en) | Video recommendation method and device and electronic equipment | |
| US20150310392A1 (en) | Job recommendation engine using a browsing history | |
| US20230089961A1 (en) | Optimizing content distribution using a model | |
| CN102132301A (en) | System and method for online advertising using user social information | |
| US20170323230A1 (en) | Evaluating keyword performance | |
| US20110131093A1 (en) | System and method for optimizing selection of online advertisements | |
| JP6334348B2 (en) | Distribution apparatus, distribution method, and distribution program | |
| CN111291217A (en) | Content recommendation method and device, electronic equipment and computer readable medium | |
| US20160275569A1 (en) | Method and system for advertisement coordination | |
| US20160267551A1 (en) | Method and system for advertisement coordination | |
| WO2018042179A1 (en) | Method and system for providing content | |
| TW202335511A (en) | System, method and computer-readable medium for recommending streaming data | |
| US20230334261A1 (en) | Methods, systems, and media for identifying relevant content | |
| CN112818231A (en) | Information delivery method and device, electronic equipment and storage medium | |
| CN107368533B (en) | Content item recommendation method and device and electronic equipment | |
| US20160267531A1 (en) | Method and system for advertisement coordination |
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 |