CN108228804B - Method and device for updating label weight value of resource file - Google Patents
Method and device for updating label weight value of resource file Download PDFInfo
- Publication number
- CN108228804B CN108228804B CN201711483109.5A CN201711483109A CN108228804B CN 108228804 B CN108228804 B CN 108228804B CN 201711483109 A CN201711483109 A CN 201711483109A CN 108228804 B CN108228804 B CN 108228804B
- Authority
- CN
- China
- Prior art keywords
- label
- resource file
- probability
- weight
- target cluster
- 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
- G06F16/10—File systems; File servers
- G06F16/14—Details of searching files based on file metadata
- G06F16/148—File search processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
In the scheme, the probability that an object operating the resource file is interested in the resource under each label is calculated; then, updating the association weight of the resource file and each label according to the calculated corresponding probability; in this way, the association degree of each resource file and each tag can be adjusted according to the probability that the resource under each tag is interested in by the object operating the resource file, and the resource file and the tag are not manually corresponding, so that the association degree of the resource file and the tag can be improved.
Description
Technical Field
The invention relates to the technical field of internet, in particular to a method and a device for updating a weight value of a resource file label.
Background
With the development of the internet and the popularization of intelligent terminals, a large number of application programs are produced at the same time. In order to meet the needs of people, resources in the application programs are also increasingly abundant, for example, picture resources, audio resources, video resources and the like, and the video resources also include a large number of short video resources, such as uploaded resources of a second-taking application program, a beautiful-taking application program, a micro-vision application program and the like.
Due to the fact that resources are extremely rich, in the process of using the application program, a large amount of time is often consumed to search interested resources from dazzling resources, and the defect of low efficiency exists, and user experience is further affected.
In order to enable a user to find an interested resource as soon as possible, the resource is generally recommended to the user according to a tag of the resource, and then the user searches the interested resource from the recommended resource.
Disclosure of Invention
In view of the above problems, the present invention is provided to provide a method and an apparatus for updating a weight value of a tag of a resource file, which overcome the above problems or at least partially solve the above problems, so as to solve the defect of low association degree between the resource file and the tag in the prior art.
According to a first aspect of the present invention, there is provided a method for updating a resource file tag weight value, including: calculating the associated weight of the resource file and a label of each first-stage target cluster in an interest point network level topological graph, wherein the interest point network level topological graph comprises N-stage target clusters, the J-th stage target cluster comprises at least one J-1-stage target cluster, J is more than 1 and is not less than N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different; calculating the probability that the object operating the resource file is interested in the resource under each label; and updating the association weight of the resource file and each label according to the calculated corresponding probability.
According to a second aspect of the present invention, there is provided an apparatus for updating a weight value of a tag of a resource file, including: the system comprises a computing unit, a calculating unit and a judging unit, wherein the computing unit is used for computing the relevance weight of a resource file and a label of each first-stage target cluster in an interest point network hierarchical topological graph, the interest point network hierarchical topological graph comprises N-stage target clusters, the J-th target cluster comprises at least one J-1-stage target cluster, J is more than 1 and is not less than N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different; the computing unit is further used for computing the probability that the object operating the resource file is interested in the resource under each label; and the updating unit is used for updating the association weight of the resource file and each label according to the calculated corresponding probability.
In the embodiment of the invention, the probability that the object operating the resource file is interested in the resource under each label is calculated; then, updating the association weight of the resource file and each label according to the calculated corresponding probability; in this way, the association degree of each resource file and each tag can be adjusted according to the probability that the resource under each tag is interested in by the object operating the resource file, and the resource file and the tag are not manually corresponding, so that the association degree of the resource file and the tag can be improved.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 is a flowchart of updating a resource file tag weight value according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a proposed point-of-interest network hierarchy topology according to an embodiment of the invention;
fig. 3 is a schematic diagram of an apparatus for updating a weight value of a resource file tag according to an embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Fig. 1 schematically shows a flowchart of a method 10 for updating a resource file tag weight value according to an embodiment of the present invention. As shown in fig. 1, the method 10 may include steps 100, 110, 120, and 130.
Step 100: calculating the associated weight of the resource file and a label of each first-stage target cluster in an interest point network level topological graph, wherein the interest point network level topological graph comprises N-stage target clusters, the J-th stage target cluster comprises at least one J-1-stage target cluster, J is more than 1 and is not less than N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different;
step 110: calculating the probability that the object operating the resource file is interested in the resource under each label;
step 120: and updating the association weight of the resource file and each label according to the calculated corresponding probability.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: label 1, label 2, label 3, label 4, label 5, label 6, label 7, label 8, label 9, label 10, the second level target cluster has 4: b1, B2, B3 and B4, and the third-level target cluster has 2: c1 and C2, the fourth level target cluster has 1: d1, for video 1 in a1, first calculating the association weights of video 1 respectively associated with a1, a2, A3, a4, a5, A6, a7, A8, a9 and a10 to obtain 10 association weights, obtaining association weight 1, association weight 2, association weight 3, association weight 4, association weight 5, association weight 6, association weight 7, association weight 8, association weight 9 and association weight 10, calculating the probability that the user playing video 1 is interested in the resource under each label, such as obtaining probability 1, probability 2, probability 3, … … and probability 10, updating association weight 1 with probability 1, updating association weight 2 with probability 2, updating association weight 3 with probability 3, updating association weight 4, … … with probability 4 and updating association weight 10 with probability 10.
In the embodiment of the present invention, the operation manner may be different according to different types of resource files, for example, the operation may be playing according to an audio file or a video file, and the operation manner may be browsing according to a text file, although the above are only some examples, and are not limited herein.
In the embodiment of the present invention, when calculating the association weight associated with the label of each first-level target cluster in the network-level topology map of the resource file and the point of interest, optionally, the following method may be adopted:
constructing a first vector of the resource file;
and for each first-stage target cluster, constructing a second vector of the first-stage target cluster, and taking the distance value between the second vector and the first vector as the association weight of the resource file and the label of the first-stage target cluster.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: tag 1, tag 2, tag 3, tag 4, tag 5, tag 6, tag 7, tag 8, tag 9, and tag 10, where the association weight of video 1 associated with tag 1 is association weight 1, the association weight of video 1 associated with tag 2 is association weight 2, … …, and the association weight of video 1 associated with tag 10 is association weight 10, where when the vector constructed for video 1 is vector 1 and association weight 1 is calculated, vector 2 of a1 is constructed, and the distance between vector 1 and vector 2 is taken as association weight 1; when the association weight 2 is calculated, constructing a vector 3 of A2, and taking the distance between the vector 1 and the vector 3 as the association weight 2; when the association weight 3 is calculated, constructing a vector 4 of A3, and taking the distance between the vector 1 and the vector 4 as the association weight 3; … …, respectively; by analogy, when calculating the association weight 10, a vector 11 of a10 is constructed, and the distance between the vector 1 and the vector 11 is taken as the association weight 10.
In the embodiment of the present invention, when constructing the second vector of the first-stage target cluster, optionally, the following manner may be adopted:
constructing a third vector of each resource file in the first-level target cluster;
a second vector for the first level target cluster is determined from all third vectors for the first level target cluster.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, when constructing the second vector of the A1, firstly constructing a third vector 1 of a video 1, a third vector 2 of a video 4 and a third vector 3 of a video 8, and determining the second vector of the A1 according to the third vector 1, the third vector 2 and the third vector 3; when the second vector of A2 is constructed, a third vector 4 of video 2 and a third vector 5 of video 19 are constructed, and the second vector of A2 is determined according to the third vector 4 and the third vector 5; for A3, the second vector of A3 is determined directly from the third vector of the constructed video 20, and so on, for a10, the third vector 19 of video 12, the third vector 20 of video 15, and the second vector of a10 is determined from the third vector 19 and the third vector 20.
In the embodiment of the present invention, when calculating the probability that the object operating the resource file is interested in the resource under each label, optionally, the following method may be adopted:
aiming at any label, calculating the sub-probability that each object in the objects operating the resource file is interested in the resource under the label;
and taking the sum of the sub probabilities as the probability that the object operating the resource file is interested in the resource under the label.
For example, the topology map of the interest point network hierarchy includes 3 first-level target clusters, a1, a2 and A3, the associated weight of video 1 and a1 is 0.5, the associated weight of video 1 and a2 is 0.3, the associated weight of video 1 and A3 is 0.2, the number of users playing video 1 is 10, for a1, a sub-probability of interest of each of the 10 users playing video 1 for video at a1 is calculated, e.g., sub-probability 1 of interest of user 1 for video at a1, sub-probability 2 of interest of user 2 for video at a1, sub-probability 3 of interest of user 3 for video at a1, … …, sub-probability 10 of interest of user 10 for video at a1, and the sum of sub-probability 1, sub-probability 2, sub-probability 3, … … and sub-probability 10 is used as the probability of interest of video at a1 for video 1, a2 and a2 for a similar processes, and will not be described in detail herein.
In the embodiment of the present invention, in order to avoid a large difference between the sub probabilities and the total probability, normalization processing may be performed on the sub probabilities, so that the sub probabilities after the normalization processing are in the same interval, for example, the sub probabilities after the normalization processing are all in a range of [0,1], still taking the above example as an example, for example, in order to avoid a large difference between the sub probabilities after the normalization processing and the probabilities of interest of the user on the video under a1, the sub probabilities 1, 2, 3, … …, and 10 may be normalized, so that the sub probabilities after the normalization processing are in the same interval, for example, all in a range of [0,1 ].
The above is described by taking the number of users as 10 as an example, and of course, the number of users that can play the video 1 in practical application is much larger than 10, but the calculation principle is similar, and will not be described in detail here.
In the embodiment of the present invention, there are various ways to calculate the sub-probability that each object in the objects operating the resource file is interested in the resource under the label, and optionally, the following ways may be adopted:
and for each object, determining the associated weight of at least one resource file operated by the object and the label, and adding all the determined associated weights to obtain the sub-probability that the object is interested in the resource under the label.
Still taking the above example as an example, when calculating the sub-probability 1 of interest of the user 1 to the video under a1, determining the video played by the user 1, for example, the user 1 plays 10 videos: video 1, video 2, video 3, … … and video 10, wherein the associated weight of video 1 and A1 is a1, the associated weight of video 2 and A1 is a2, the associated weight of video 3 and A1 is A3 and … …, and the associated weight of video 10 and A1 is a10, the sum of a1, a2, A3, … … and a10 is taken as the sub-probability that the video under A1 is interested by the user 1.
Similarly, in order to avoid a large difference between the respective association weights and the sub-probabilities, the association weights may be normalized so that the respective association weights after the normalization process are in the same interval, for example, the association weights after the normalization process are all in [0,1], and still in the above example, for example, in order to avoid a1, a2, A3, … …, and a10 having a large difference between the effects of a1, a2, A3, … …, and a10 on the sub-probabilities of the user 1 that is interested in the video under a1, the association weights after the normalization process may be in the same interval, for example, all in [0,1 ].
In the embodiment of the present invention, when the association weight between the resource file and each tag is updated according to the calculated corresponding probability, optionally, the following manner may be adopted:
for each label, directly adopting the calculated corresponding probability to replace the associated weight of the resource file and the label; or
And for each label, replacing the association weight of the resource file and the label by the sum of the corresponding probability and the corresponding association weight obtained by calculation.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: tag 1, tag 2, tag 3, tag 4, tag 5, tag 6, tag 7, tag 8, tag 9 and tag 10, wherein the resource file is video 1 in A1, the association weights of video 1 and A1, A2, A3, A4, A5, A6, A7, A8, A9 and A10 are calculated first to obtain 10 association weights, association weight 1, association weight 2, association weight 3, association weight 4, association weight 5, association weight 6, association weight 7, association weight 8, association weight 9 and association weight 10, the probability of interest of the object playing video 1 to the video under tag 1, tag 2, tag 3, … … and tag 10 is calculated, such as obtaining probability 1, probability 2, probability 3, … … and probability 10, and the association weight 1 is directly replaced by probability 1, the association weight 2 is directly replaced by probability 2, the association weight 3 is directly replaced by probability 3, directly replacing the association weight 4 with a probability of 4, … …, and directly replacing the association weight 10 with a probability of 10; alternatively, the sum of the probability 1 and the association weight 1 is used as the new association weight 1 of the videos 1 and a1, the sum of the probability 2 and the association weight 2 is used as the new association weight 2 of the videos 1 and a2, the sum of the probability 3 and the association weight 3 is used as the new association weight 3, … … of the videos 1 and A3, and so on, the sum of the probability 10 and the association weight 10 is used as the new association weight 10 of the videos 1 and a 10.
It should be noted that, because the probability that the user is interested in the resource under the tag is used in the process of updating the associated weight value of the resource file and the tag, and the associated weight value of the resource file and the tag is used in the process of calculating the probability that the user is interested in the resource under the tag, the above process is an iterative process, and after the associated weight value is updated, the probability that the user is interested in the resource under the tag is further updated, and the process is repeated until the calculated associated weight value is stable.
In the embodiment of the invention, the resource file is a file in the network level topological graph of the interest point or a newly added file.
For example, the resource file is video 1 in a1 shown in fig. 2, or may be a file other than the video shown in fig. 2.
When recommending a resource file to an object operating the resource file, recommending the resource file according to the association weight between the resource file and each tag, and then updating the association weight between the resource file and each tag, the recommending manner may also change, for example, after updating the association weight between the resource file and each tag according to the calculated corresponding probability, recommending in the following manner:
and recommending the resource file to the object operating the resource file according to the updated association weight.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: label 1, label 2, label 3, label 4, label 5, label 6, label 7, label 8, label 9, label 10, the second level target cluster has 4: b1, B2, B3 and B4, the third polar target cluster has 2: c1 and C2, the fourth level target cluster has 1: d1, for video 1 in A1, calculating the associated weights of video 1 respectively associated with A1, A2, A3, A4, A5, A6, A7, A8, A9 and A10 to obtain 10 associated weights, wherein before the update of the associated weights, the video under the label 1 is recommended to the user playing video 1 according to the associated weight 1, the video under the label 2 is recommended to the user playing video 1 according to the associated weight 2, the video under the label 3 is recommended to the user playing video 1 according to the associated weight 3, and so on, the video under the label 10 is recommended to the user playing video 1 according to the associated weight 10, and the video under each label is recommended to the user playing video 1 according to the associated weight 10, calculating the probability that the user playing the video 1 is interested in the resource under each label, such as obtaining probability 1, probability 2, probability 3, … … and probability 10, replacing the associated weight 1 with probability 1, replacing the associated weight 2 with probability 2, replacing the associated weight 3 with probability 3, replacing the associated weight 4 with probability 4, … … and replacing the associated weight 10 with probability 10, at this time, when recommending the video to the user playing the video 1, recommending the video under the label 1 to the user playing the video 1 according to probability 1, recommending the video under the label 2 to the user playing the video 1 according to probability 2, recommending the video under the label 3 to the user playing the video 1 according to probability 3, and so on, recommending the video under the label 10 to the user playing the video 1 according to probability 10.
When recommending the resource file to the object operating the resource file according to the updated association weight, optionally, the following method may be adopted:
for each label, selecting a target object with the updated association weight corresponding to the label from the objects for operating the resource file; and recommending the resource file under the label to the selected target object.
For example, the topology of the network hierarchy of the interest point includes 3 first-level target clusters, a1, a2 and A3, the updated association weight of video 1 and a1 is 0.5, the updated association weight of video 1 and a2 is 0.3, the updated association weight of video 1 and A3 is 0.2, the number of people playing video 1 is 100 ten thousand, for a1, a video included in a1 is recommended to 50% of 100 ten thousand users, for a2, a video included in a2 is recommended to 30% of 100 ten thousand users, and for A3, a video included in A3 is recommended to 20% of 100 ten thousand users.
Referring to fig. 3, in an embodiment of the present invention, an apparatus 30 for updating a weight value of a resource file tag is further provided, including:
the computing unit 300 is configured to compute an association weight of a resource file associated with a label of each first-level target cluster in an interest point network hierarchical topological graph, where the interest point network hierarchical topological graph includes N-level target clusters, a J-level target cluster includes at least one J-1-level target cluster, J is greater than 1 and is greater than or equal to N, J is an integer, each first-level target cluster includes at least one resource file, one first-level target cluster corresponds to one label, and labels corresponding to any two different first-level target clusters are different;
the calculating unit 300 is configured to calculate a probability that an object operating the resource file is interested in the resource under each tag;
an updating unit 310, configured to update the association weight between the resource file and each tag according to the calculated corresponding probability.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: label 1, label 2, label 3, label 4, label 5, label 6, label 7, label 8, label 9, label 10, the second level target cluster has 4: b1, B2, B3 and B4, and the third-level target cluster has 2: c1 and C2, the fourth level target cluster has 1: d1, for video 1 in a1, first calculating the association weights of video 1 respectively associated with a1, a2, A3, a4, a5, A6, a7, A8, a9 and a10 to obtain 10 association weights, obtaining association weight 1, association weight 2, association weight 3, association weight 4, association weight 5, association weight 6, association weight 7, association weight 8, association weight 9 and association weight 10, calculating the probability that the user playing video 1 is interested in the resource under each label, such as obtaining probability 1, probability 2, probability 3, … … and probability 10, updating association weight 1 with probability 1, updating association weight 2 with probability 2, updating association weight 3 with probability 3, updating association weight 4, … … with probability 4 and updating association weight 10 with probability 10.
In the embodiment of the present invention, the operation manner may be different according to different types of resource files, for example, the operation may be playing according to an audio file or a video file, and the operation manner may be browsing according to a text file, although the above are only some examples, and are not limited herein.
In this embodiment of the present invention, optionally, the calculating unit 300 calculates an association weight between the resource file and the label of each first-level target cluster in the network-level topology map of the point of interest, including:
constructing a first vector of the resource file;
and for each first-stage target cluster, constructing a second vector of the first-stage target cluster, and taking the distance value between the second vector and the first vector as the association weight of the resource file and the label of the first-stage target cluster.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, wherein the labels corresponding to the first-level target clusters are respectively as follows: tag 1, tag 2, tag 3, tag 4, tag 5, tag 6, tag 7, tag 8, tag 9, and tag 10, where the association weight of video 1 associated with tag 1 is association weight 1, the association weight of video 1 associated with tag 2 is association weight 2, … …, and the association weight of video 1 associated with tag 10 is association weight 10, where when the vector constructed for video 1 is vector 1 and association weight 1 is calculated, vector 2 of a1 is constructed, and the distance between vector 1 and vector 2 is taken as association weight 1; when the association weight 2 is calculated, constructing a vector 3 of A2, and taking the distance between the vector 1 and the vector 3 as the association weight 2; when the association weight 3 is calculated, constructing a vector 4 of A3, and taking the distance between the vector 1 and the vector 4 as the association weight 3; … …, respectively; by analogy, when calculating the association weight 10, a vector 11 of a10 is constructed, and the distance between the vector 1 and the vector 11 is taken as the association weight 10.
In this embodiment of the present invention, the constructing the second vector of the first-stage target cluster by the computing unit 300 includes:
constructing a third vector of each resource file in the first-level target cluster;
a second vector for the first level target cluster is determined from all third vectors for the first level target cluster.
For example, the topology of the network level of interest is shown in fig. 2, the topology of the network level of interest includes 4 levels of target clusters, and the number of the first level of target clusters is 10: a1, A2, A3, A4, A5, A6, A7, A8, A9 and A10, when constructing the second vector of the A1, firstly constructing a third vector 1 of a video 1, a third vector 2 of a video 4 and a third vector 3 of a video 8, and determining the second vector of the A1 according to the third vector 1, the third vector 2 and the third vector 3; when the second vector of A2 is constructed, a third vector 4 of video 2 and a third vector 5 of video 19 are constructed, and the second vector of A2 is determined according to the third vector 4 and the third vector 5; for A3, the second vector of A3 is determined directly from the third vector of the constructed video 20, and so on, for a10, the third vector 19 of video 12, the third vector 20 of video 15, and the second vector of a10 is determined from the third vector 19 and the third vector 20.
In this embodiment of the present invention, the calculating unit 300 calculates the probability that the object operating the resource file is interested in the resource under each label, including:
aiming at any label, calculating the sub-probability that each object in the objects operating the resource file is interested in the resource under the label;
and taking the sum of the sub probabilities as the probability that the object operating the resource file is interested in the resource under the label.
For example, the topology map of the interest point network hierarchy includes 3 first-level target clusters, a1, a2 and A3, the associated weight of video 1 and a1 is 0.5, the associated weight of video 1 and a2 is 0.3, the associated weight of video 1 and A3 is 0.2, the number of users playing video 1 is 10, for a1, a sub-probability of interest of each of the 10 users playing video 1 for video at a1 is calculated, e.g., sub-probability 1 of interest of user 1 for video at a1, sub-probability 2 of interest of user 2 for video at a1, sub-probability 3 of interest of user 3 for video at a1, … …, sub-probability 10 of interest of user 10 for video at a1, and the sum of sub-probability 1, sub-probability 2, sub-probability 3, … … and sub-probability 10 is used as the probability of interest of video at a1 for video 1, a2 and a2 for a similar processes, and will not be described in detail herein.
The above is described by taking the number of users as 10 as an example, and of course, the number of users that can play the video 1 in practical application is much larger than 10, but the calculation principle is similar, and will not be described in detail here.
In this embodiment of the present invention, optionally, the calculating unit 300 calculates a sub-probability that each of the objects operating the resource file is interested in the resource under the label, including:
and for each object, determining the associated weight of at least one resource file operated by the object and the label, and adding all the determined associated weights to obtain the sub-probability that the object is interested in the resource under the label.
Still taking the above example as an example, when calculating the sub-probability 1 of interest of the user 1 to the video under a1, determining the video played by the user 1, for example, the user 1 plays 10 videos: video 1, video 2, video 3, … … and video 10, wherein the associated weight of video 1 and A1 is a1, the associated weight of video 2 and A1 is a2, the associated weight of video 3 and A1 is A3 and … …, and the associated weight of video 10 and A1 is a10, the sum of a1, a2, A3, … … and a10 is taken as the sub-probability that the video under A1 is interested by the user 1.
In this embodiment of the present invention, optionally, the updating unit 310 updates the association weight of the resource file and each tag according to the calculated corresponding probability, including:
for each label, directly adopting the calculated corresponding probability to replace the associated weight of the resource file and the label; or
And for each label, replacing the association weight of the resource file and the label by the sum of the corresponding probability and the corresponding association weight obtained by calculation.
In the embodiment of the present invention, optionally, the resource file is a file in the network level topology map of the point of interest or a newly added file.
In the embodiment of the present invention, after the association weight between the resource file and each tag is updated according to the calculated corresponding probability, the apparatus further includes a recommending unit 320, configured to recommend the resource file to an object that operates the resource file according to the updated association weight.
In this embodiment of the present invention, optionally, the recommending unit 320 recommends the resource file to the object operating the resource file according to the updated association weight, including:
for each label, selecting a target object with the updated association weight corresponding to the label from the objects for operating the resource file; and recommending the resource file under the label to the selected target object.
For example, the topology of the network hierarchy of the interest point includes 3 first-level target clusters, a1, a2 and A3, the updated association weight of video 1 and a1 is 0.5, the updated association weight of video 1 and a2 is 0.3, the updated association weight of video 1 and A3 is 0.2, the number of people playing video 1 is 100 ten thousand, for a1, a video included in a1 is recommended to 50% of 100 ten thousand users, for a2, a video included in a2 is recommended to 30% of 100 ten thousand users, and for A3, a video included in A3 is recommended to 20% of 100 ten thousand users.
In an embodiment of the present invention, a device for updating a weight value of a resource file tag is further provided, including:
one or more processors;
a memory;
a program stored in the memory, which when executed by the one or more processors, causes the processors to perform the method as described above.
In an embodiment of the present invention, a computer-readable storage medium is also provided, which stores a program, and when the program is executed by a processor, the program causes the processor to execute the method described above.
The invention also discloses:
a1, a method for updating a resource file label weight value, comprising:
calculating the associated weight of the resource file and a label of each first-stage target cluster in an interest point network level topological graph, wherein the interest point network level topological graph comprises N-stage target clusters, the J-th stage target cluster comprises at least one J-1-stage target cluster, J is more than 1 and is not less than N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different;
calculating the probability that the object operating the resource file is interested in the resource under each label;
and updating the association weight of the resource file and each label according to the calculated corresponding probability.
A2, the method as in a1, calculating the association weight of the resource file and the label of each first-level target cluster in the topology map of the network hierarchy of the interest point, comprising:
constructing a first vector of the resource file;
and for each first-stage target cluster, constructing a second vector of the first-stage target cluster, and taking the distance value between the second vector and the first vector as the association weight of the resource file and the label of the first-stage target cluster.
A3, the method as described in a2, constructing a second vector of the first-stage target cluster, comprising:
constructing a third vector of each resource file in the first-level target cluster;
a second vector for the first level target cluster is determined from all third vectors for the first level target cluster.
A4, the method of A1, calculating a probability that an object operating the resource file is interested in the resource under each label, comprising:
aiming at any label, calculating the sub-probability that each object in the objects operating the resource file is interested in the resource under the label;
and taking the sum of the sub probabilities as the probability that the object operating the resource file is interested in the resource under the label.
A5, the method as in A4, calculating a sub-probability that each of the objects operating the resource file is interested in the resource under the label, comprising:
and for each object, determining the associated weight of at least one resource file operated by the object and the label, and adding all the determined associated weights to obtain the sub-probability that the object is interested in the resource under the label.
A6, the method according to any one of A1-A5, wherein the updating the associated weight of the resource file and each label according to the calculated corresponding probability comprises:
for each label, directly adopting the calculated corresponding probability to replace the associated weight of the resource file and the label; or
And for each label, replacing the association weight of the resource file and the label by the sum of the corresponding probability and the corresponding association weight obtained by calculation.
A7, the method as in A1, the resource file is a file in the topology of the network level of the point of interest or a newly added file.
A8, the method of A1, further comprising, after updating the associated weight of the resource file and each of the tags according to the calculated corresponding probability:
and recommending the resource file to the object operating the resource file according to the updated association weight.
A9, the method of A8, recommending the resource file to the object operating the resource file according to the updated association weight, including:
for each label, selecting a target object with the updated association weight corresponding to the label from the objects for operating the resource file; and recommending the resource file under the label to the selected target object.
A10, an apparatus for updating a resource file label weight value, comprising:
the system comprises a computing unit, a calculating unit and a judging unit, wherein the computing unit is used for computing the relevance weight of a resource file and a label of each first-stage target cluster in an interest point network hierarchical topological graph, the interest point network hierarchical topological graph comprises N-stage target clusters, the J-th target cluster comprises at least one J-1-stage target cluster, J is more than 1 and is not less than N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different;
the computing unit is used for computing the probability that the object operating the resource file is interested in the resource under each label;
and the updating unit is used for updating the association weight of the resource file and each label according to the calculated corresponding probability.
A11, the apparatus as described in a10, the computing unit calculating an associated weight of the resource file associated with the label of each first-level target cluster in the point of interest network-level topology graph, including:
constructing a first vector of the resource file;
and for each first-stage target cluster, constructing a second vector of the first-stage target cluster, and taking the distance value between the second vector and the first vector as the association weight of the resource file and the label of the first-stage target cluster.
A12, the apparatus as in a11, the computing unit constructing a second vector of the first-stage target cluster, including:
constructing a third vector of each resource file in the first-level target cluster;
a second vector for the first level target cluster is determined from all third vectors for the first level target cluster.
A13, the apparatus of A10, the calculating unit calculating the probability that the object operating the resource file is interested in the resource under each label, comprising:
aiming at any label, calculating the sub-probability that each object in the objects operating the resource file is interested in the resource under the label;
and taking the sum of the sub probabilities as the probability that the object operating the resource file is interested in the resource under the label.
A14, the apparatus as in A13, the calculating unit calculating a sub probability that each of the objects operating the resource file is interested in the resource under the label, comprising:
and for each object, determining the associated weight of at least one resource file operated by the object and the label, and adding all the determined associated weights to obtain the sub-probability that the object is interested in the resource under the label.
A15, the apparatus of any one of A10-A14, the updating unit updating the association weight of the resource file and each label according to the calculated corresponding probability, comprising:
for each label, directly adopting the calculated corresponding probability to replace the associated weight of the resource file and the label; or
And for each label, replacing the association weight of the resource file and the label by the sum of the corresponding probability and the corresponding association weight obtained by calculation.
A16, the apparatus as in A10, the resource file being a file in the topology of the network hierarchy of points of interest or a newly added file.
A17, the apparatus of A10, further comprising a recommending unit for recommending the resource file to the object operating the resource file according to the updated association weight after updating the association weight of the resource file and each label according to the calculated corresponding probability.
A18, the apparatus of a17, the recommending unit recommending the resource file to the object operating the resource file according to the updated association weight, the recommending unit including:
for each label, selecting a target object with the updated association weight corresponding to the label from the objects for operating the resource file; and recommending the resource file under the label to the selected target object.
A19, an apparatus for updating a resource file label weight value, comprising:
one or more processors;
a memory;
a program stored in the memory, which when executed by the one or more processors, causes the processors to perform a method as any one of a1-a9 recites.
A20, a computer readable storage medium storing a program which, when executed by a processor, causes the processor to carry out the method of any one of a1-a 9.
The methods and apparatus provided herein are not inherently related to any particular computer, virtual machine system, or other apparatus. Various general purpose systems may also be used with the teachings herein. The required structure for constructing such a device will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules in the apparatus of an embodiment may be adaptively changed and disposed in one or more apparatuses other than the embodiment. Several modules of embodiments may be combined into one module or unit or assembly and furthermore they may be divided into a plurality of sub-modules or sub-units or sub-assemblies. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or modules are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments may be used in any combination.
Various apparatus embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functionality of some or all of the modules in an apparatus according to embodiments of the present invention. The present invention may also be embodied as apparatus programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
Claims (10)
1. A method for updating a resource file label weight value comprises the following steps:
calculating the associated weight of the resource file and a label of each first-stage target cluster in an interest point network hierarchical topological graph, wherein the interest point network hierarchical topological graph comprises N-stage target clusters, the J-th stage target cluster comprises at least one J-1 stage target cluster, J is more than 1 and less than or equal to N, J is an integer, each first-stage target cluster comprises at least one resource file, one first-stage target cluster corresponds to one label, and the labels corresponding to any two different first-stage target clusters are different;
calculating the probability that the object operating the resource file is interested in the resource under each label;
and updating the association weight of the resource file and each label according to the calculated corresponding probability.
2. The method of claim 1, calculating an association weight associated with the resource file and the label of each first-level target cluster in the point of interest network-level topology graph, comprising:
constructing a first vector of the resource file;
and for each first-stage target cluster, constructing a second vector of the first-stage target cluster, and taking the distance value between the second vector and the first vector as the association weight of the resource file and the label of the first-stage target cluster.
3. The method of claim 2, constructing a second vector for the first-stage target cluster, comprising:
constructing a third vector of each resource file in the first-level target cluster;
a second vector for the first level target cluster is determined from all third vectors for the first level target cluster.
4. The method of claim 1, calculating a probability that an object operating the resource file is interested in the resource under each of the tags, comprising:
aiming at any label, calculating the sub-probability that each object in the objects operating the resource file is interested in the resource under the label;
and taking the sum of the sub probabilities as the probability that the object operating the resource file is interested in the resource under the label.
5. The method of claim 4, calculating a sub-probability that each of the objects operating on the resource file is interested in the resource under the label, comprising:
and for each object, determining the associated weight of at least one resource file operated by the object and the label, and adding all the determined associated weights to obtain the sub-probability that the object is interested in the resource under the label.
6. The method of any of claims 1-5, updating the association weight of the resource file with each of the tags according to the calculated corresponding probability, comprising:
for each label, directly adopting the calculated corresponding probability to replace the associated weight of the resource file and the label; or
And for each label, replacing the association weight of the resource file and the label by the sum of the corresponding probability and the corresponding association weight obtained by calculation.
7. The method of claim 1, after updating the association weight of the resource file with each of the tags according to the calculated corresponding probability, the method further comprising:
and recommending the resource file to the object operating the resource file according to the updated association weight.
8. An apparatus for updating a resource file tag weight value, comprising:
the system comprises a computing unit, a calculating unit and a processing unit, wherein the computing unit is used for computing the relevance weight of a resource file and a label of each first-level target cluster in an interest point network hierarchical topological graph, the interest point network hierarchical topological graph comprises N-level target clusters, the J-level target cluster comprises at least one J-1-level target cluster, J is more than 1 and less than or equal to N, J is an integer, each first-level target cluster comprises at least one resource file, one first-level target cluster corresponds to one label, and the labels corresponding to any two different first-level target clusters are different;
the computing unit is used for computing the probability that the object operating the resource file is interested in the resource under each label;
and the updating unit is used for updating the association weight of the resource file and each label according to the calculated corresponding probability.
9. An apparatus for updating a resource file tag weight value, comprising:
one or more processors;
a memory;
a program stored in the memory, which when executed by the one or more processors, causes the processors to perform the method of any one of claims 2-7.
10. A computer-readable storage medium storing a program which, when executed by a processor, causes the processor to perform the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711483109.5A CN108228804B (en) | 2017-12-29 | 2017-12-29 | Method and device for updating label weight value of resource file |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711483109.5A CN108228804B (en) | 2017-12-29 | 2017-12-29 | Method and device for updating label weight value of resource file |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108228804A CN108228804A (en) | 2018-06-29 |
| CN108228804B true CN108228804B (en) | 2020-12-11 |
Family
ID=62647182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711483109.5A Active CN108228804B (en) | 2017-12-29 | 2017-12-29 | Method and device for updating label weight value of resource file |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108228804B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101477554A (en) * | 2009-01-16 | 2009-07-08 | 西安电子科技大学 | User interest based personalized meta search engine and search result processing method |
| CN101770520A (en) * | 2010-03-05 | 2010-07-07 | 南京邮电大学 | User interest modeling method based on user browsing behavior |
| CN102713905A (en) * | 2010-01-08 | 2012-10-03 | 瑞典爱立信有限公司 | A method and apparatus for social tagging of media files |
| CN103853792A (en) * | 2012-12-07 | 2014-06-11 | 中兴通讯股份有限公司 | Automatic image semantic annotation method and system |
| CN104469430A (en) * | 2014-12-24 | 2015-03-25 | 武汉泰迪智慧科技有限公司 | Video recommending method and system based on context and group combination |
| CN105069172A (en) * | 2015-09-09 | 2015-11-18 | 中国人民大学 | Interest label generating method |
| CN105589917A (en) * | 2015-09-17 | 2016-05-18 | 广州市动景计算机科技有限公司 | Method and device for analyzing log information of browser |
| CN105847985A (en) * | 2016-03-30 | 2016-08-10 | 乐视控股（北京）有限公司 | Video recommendation method and device |
| CN106230911A (en) * | 2016-07-25 | 2016-12-14 | 腾讯科技（深圳）有限公司 | A kind of played data recommends method, interest tags to determine method and relevant device |
| CN107222570A (en) * | 2017-07-10 | 2017-09-29 | 江西博瑞彤芸科技有限公司 | A kind of health information method for pushing based on user's behaviour in service |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4169063B2 (en) * | 2006-04-06 | 2008-10-22 | ソニー株式会社 | Data processing apparatus, data processing method, and program |
-
2017
- 2017-12-29 CN CN201711483109.5A patent/CN108228804B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101477554A (en) * | 2009-01-16 | 2009-07-08 | 西安电子科技大学 | User interest based personalized meta search engine and search result processing method |
| CN102713905A (en) * | 2010-01-08 | 2012-10-03 | 瑞典爱立信有限公司 | A method and apparatus for social tagging of media files |
| CN101770520A (en) * | 2010-03-05 | 2010-07-07 | 南京邮电大学 | User interest modeling method based on user browsing behavior |
| CN103853792A (en) * | 2012-12-07 | 2014-06-11 | 中兴通讯股份有限公司 | Automatic image semantic annotation method and system |
| CN104469430A (en) * | 2014-12-24 | 2015-03-25 | 武汉泰迪智慧科技有限公司 | Video recommending method and system based on context and group combination |
| CN105069172A (en) * | 2015-09-09 | 2015-11-18 | 中国人民大学 | Interest label generating method |
| CN105589917A (en) * | 2015-09-17 | 2016-05-18 | 广州市动景计算机科技有限公司 | Method and device for analyzing log information of browser |
| CN105847985A (en) * | 2016-03-30 | 2016-08-10 | 乐视控股（北京）有限公司 | Video recommendation method and device |
| CN106230911A (en) * | 2016-07-25 | 2016-12-14 | 腾讯科技（深圳）有限公司 | A kind of played data recommends method, interest tags to determine method and relevant device |
| CN107222570A (en) * | 2017-07-10 | 2017-09-29 | 江西博瑞彤芸科技有限公司 | A kind of health information method for pushing based on user's behaviour in service |
Non-Patent Citations (2)
| Title |
|---|
| "一种基于标签概率相关性的微博推荐方法";张迪 等;《计算机工程与科学》;20170915;第39卷(第9期);第1742-1748页 * |
| "贝叶斯概率LSA模型权重更新算法";曾广平;《计算机工程与应用》;20090721;第45卷(第21期);第88-90页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108228804A (en) | 2018-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110457581B (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN108228758B (en) | Text classification method and device | |
| US20160188725A1 (en) | Method and System for Enhanced Content Recommendation | |
| US9864951B1 (en) | Randomized latent feature learning | |
| CN104809154A (en) | Method and device for recommending information | |
| US9299098B2 (en) | Systems for generating a global product taxonomy | |
| US20100312726A1 (en) | Feature vector clustering | |
| US10911384B2 (en) | Staggered notification by affinity to promote positive discussion | |
| CN107729578B (en) | Music recommendation method and device | |
| CN105898362A (en) | Video content retrieval method and device | |
| CN105095414A (en) | Method and apparatus used for predicting network search volume | |
| WO2017168967A1 (en) | Device for determining data analysis method candidate | |
| CN111046188A (en) | User preference degree determining method and device, electronic equipment and readable storage medium | |
| CN105719164A (en) | Paid multimedia resource recommending method and paid multimedia resource recommending device | |
| CN112199600A (en) | Target object identification method and device | |
| CN110515631B (en) | Method for generating application installation data packet, server and computer storage medium | |
| CN112182281B (en) | Audio recommendation method, device and storage medium | |
| CN109509051B (en) | Article recommendation method and device | |
| CN108647227B (en) | Recommendation method and device | |
| CN108228804B (en) | Method and device for updating label weight value of resource file | |
| CN103457944A (en) | Method and device for pushing information and server | |
| US9348922B2 (en) | Ranking channels in search | |
| CN113705683B (en) | Training method and device of recommendation model, electronic equipment and storage medium | |
| CN103970813A (en) | Multimedia content searching method and system | |
| CN108197239B (en) | Method and device for generating point of interest network topological graph |
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 |